diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,93329 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.999024683021982,
+  "eval_steps": 500,
+  "global_step": 13328,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00030009753169780176,
+      "grad_norm": 53.75,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.5659,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006001950633956035,
+      "grad_norm": 35.5,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.3985,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009002925950934053,
+      "grad_norm": 54.75,
+      "learning_rate": 1.2e-05,
+      "loss": 1.477,
+      "step": 3
+    },
+    {
+      "epoch": 0.001200390126791207,
+      "grad_norm": 24.875,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.4813,
+      "step": 4
+    },
+    {
+      "epoch": 0.001500487658489009,
+      "grad_norm": 8.4375,
+      "learning_rate": 2e-05,
+      "loss": 1.2471,
+      "step": 5
+    },
+    {
+      "epoch": 0.0018005851901868107,
+      "grad_norm": 33.75,
+      "learning_rate": 2.4e-05,
+      "loss": 1.3291,
+      "step": 6
+    },
+    {
+      "epoch": 0.0021006827218846124,
+      "grad_norm": 6.875,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.2212,
+      "step": 7
+    },
+    {
+      "epoch": 0.002400780253582414,
+      "grad_norm": 24.75,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.1696,
+      "step": 8
+    },
+    {
+      "epoch": 0.0027008777852802163,
+      "grad_norm": 5.875,
+      "learning_rate": 3.6e-05,
+      "loss": 1.0113,
+      "step": 9
+    },
+    {
+      "epoch": 0.003000975316978018,
+      "grad_norm": 5.1875,
+      "learning_rate": 4e-05,
+      "loss": 1.1206,
+      "step": 10
+    },
+    {
+      "epoch": 0.0033010728486758197,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.0603,
+      "step": 11
+    },
+    {
+      "epoch": 0.0036011703803736214,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.8e-05,
+      "loss": 1.0618,
+      "step": 12
+    },
+    {
+      "epoch": 0.003901267912071423,
+      "grad_norm": 6.25,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.0808,
+      "step": 13
+    },
+    {
+      "epoch": 0.004201365443769225,
+      "grad_norm": 6.3125,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.1451,
+      "step": 14
+    },
+    {
+      "epoch": 0.004501462975467027,
+      "grad_norm": 7.125,
+      "learning_rate": 6e-05,
+      "loss": 1.134,
+      "step": 15
+    },
+    {
+      "epoch": 0.004801560507164828,
+      "grad_norm": 5.71875,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.0657,
+      "step": 16
+    },
+    {
+      "epoch": 0.00510165803886263,
+      "grad_norm": 12.375,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.2109,
+      "step": 17
+    },
+    {
+      "epoch": 0.0054017555705604325,
+      "grad_norm": 9.125,
+      "learning_rate": 7.2e-05,
+      "loss": 1.202,
+      "step": 18
+    },
+    {
+      "epoch": 0.005701853102258234,
+      "grad_norm": 6.46875,
+      "learning_rate": 7.6e-05,
+      "loss": 1.0517,
+      "step": 19
+    },
+    {
+      "epoch": 0.006001950633956036,
+      "grad_norm": 14.5625,
+      "learning_rate": 8e-05,
+      "loss": 1.1727,
+      "step": 20
+    },
+    {
+      "epoch": 0.006302048165653837,
+      "grad_norm": 7.03125,
+      "learning_rate": 8.4e-05,
+      "loss": 1.3076,
+      "step": 21
+    },
+    {
+      "epoch": 0.006602145697351639,
+      "grad_norm": 6.21875,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.1167,
+      "step": 22
+    },
+    {
+      "epoch": 0.0069022432290494415,
+      "grad_norm": 7.15625,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.2436,
+      "step": 23
+    },
+    {
+      "epoch": 0.007202340760747243,
+      "grad_norm": 7.375,
+      "learning_rate": 9.6e-05,
+      "loss": 1.3456,
+      "step": 24
+    },
+    {
+      "epoch": 0.007502438292445045,
+      "grad_norm": 7.3125,
+      "learning_rate": 0.0001,
+      "loss": 1.2576,
+      "step": 25
+    },
+    {
+      "epoch": 0.007802535824142846,
+      "grad_norm": 7.65625,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 1.3817,
+      "step": 26
+    },
+    {
+      "epoch": 0.008102633355840648,
+      "grad_norm": 8.5625,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 1.3565,
+      "step": 27
+    },
+    {
+      "epoch": 0.00840273088753845,
+      "grad_norm": 7.625,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 1.3091,
+      "step": 28
+    },
+    {
+      "epoch": 0.008702828419236253,
+      "grad_norm": 9.875,
+      "learning_rate": 0.000116,
+      "loss": 1.3855,
+      "step": 29
+    },
+    {
+      "epoch": 0.009002925950934054,
+      "grad_norm": 7.375,
+      "learning_rate": 0.00012,
+      "loss": 1.3624,
+      "step": 30
+    },
+    {
+      "epoch": 0.009303023482631855,
+      "grad_norm": 7.09375,
+      "learning_rate": 0.000124,
+      "loss": 1.4512,
+      "step": 31
+    },
+    {
+      "epoch": 0.009603121014329656,
+      "grad_norm": 5.84375,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 1.306,
+      "step": 32
+    },
+    {
+      "epoch": 0.00990321854602746,
+      "grad_norm": 6.90625,
+      "learning_rate": 0.000132,
+      "loss": 1.428,
+      "step": 33
+    },
+    {
+      "epoch": 0.01020331607772526,
+      "grad_norm": 6.9375,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 1.5248,
+      "step": 34
+    },
+    {
+      "epoch": 0.010503413609423062,
+      "grad_norm": 6.15625,
+      "learning_rate": 0.00014,
+      "loss": 1.3499,
+      "step": 35
+    },
+    {
+      "epoch": 0.010803511141120865,
+      "grad_norm": 5.84375,
+      "learning_rate": 0.000144,
+      "loss": 1.6045,
+      "step": 36
+    },
+    {
+      "epoch": 0.011103608672818666,
+      "grad_norm": 8.1875,
+      "learning_rate": 0.000148,
+      "loss": 1.4634,
+      "step": 37
+    },
+    {
+      "epoch": 0.011403706204516468,
+      "grad_norm": 8.125,
+      "learning_rate": 0.000152,
+      "loss": 1.5927,
+      "step": 38
+    },
+    {
+      "epoch": 0.011703803736214269,
+      "grad_norm": 6.75,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 1.6547,
+      "step": 39
+    },
+    {
+      "epoch": 0.012003901267912072,
+      "grad_norm": 7.09375,
+      "learning_rate": 0.00016,
+      "loss": 1.571,
+      "step": 40
+    },
+    {
+      "epoch": 0.012303998799609873,
+      "grad_norm": 6.28125,
+      "learning_rate": 0.000164,
+      "loss": 1.6022,
+      "step": 41
+    },
+    {
+      "epoch": 0.012604096331307674,
+      "grad_norm": 7.5625,
+      "learning_rate": 0.000168,
+      "loss": 1.6619,
+      "step": 42
+    },
+    {
+      "epoch": 0.012904193863005477,
+      "grad_norm": 6.90625,
+      "learning_rate": 0.000172,
+      "loss": 2.023,
+      "step": 43
+    },
+    {
+      "epoch": 0.013204291394703279,
+      "grad_norm": 6.375,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 1.8176,
+      "step": 44
+    },
+    {
+      "epoch": 0.01350438892640108,
+      "grad_norm": 5.9375,
+      "learning_rate": 0.00018,
+      "loss": 1.5446,
+      "step": 45
+    },
+    {
+      "epoch": 0.013804486458098883,
+      "grad_norm": 5.0,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 1.3001,
+      "step": 46
+    },
+    {
+      "epoch": 0.014104583989796684,
+      "grad_norm": 5.84375,
+      "learning_rate": 0.000188,
+      "loss": 1.4936,
+      "step": 47
+    },
+    {
+      "epoch": 0.014404681521494486,
+      "grad_norm": 7.0,
+      "learning_rate": 0.000192,
+      "loss": 1.7527,
+      "step": 48
+    },
+    {
+      "epoch": 0.014704779053192287,
+      "grad_norm": 6.4375,
+      "learning_rate": 0.000196,
+      "loss": 1.5695,
+      "step": 49
+    },
+    {
+      "epoch": 0.01500487658489009,
+      "grad_norm": 5.5,
+      "learning_rate": 0.0002,
+      "loss": 1.6462,
+      "step": 50
+    },
+    {
+      "epoch": 0.015304974116587891,
+      "grad_norm": 6.4375,
+      "learning_rate": 0.00019998493749058592,
+      "loss": 1.6091,
+      "step": 51
+    },
+    {
+      "epoch": 0.015605071648285692,
+      "grad_norm": 6.78125,
+      "learning_rate": 0.0001999698749811719,
+      "loss": 1.6734,
+      "step": 52
+    },
+    {
+      "epoch": 0.015905169179983494,
+      "grad_norm": 5.625,
+      "learning_rate": 0.0001999548124717578,
+      "loss": 1.5841,
+      "step": 53
+    },
+    {
+      "epoch": 0.016205266711681297,
+      "grad_norm": 5.59375,
+      "learning_rate": 0.00019993974996234372,
+      "loss": 1.727,
+      "step": 54
+    },
+    {
+      "epoch": 0.0165053642433791,
+      "grad_norm": 5.5625,
+      "learning_rate": 0.00019992468745292968,
+      "loss": 1.7715,
+      "step": 55
+    },
+    {
+      "epoch": 0.0168054617750769,
+      "grad_norm": 5.0625,
+      "learning_rate": 0.0001999096249435156,
+      "loss": 1.726,
+      "step": 56
+    },
+    {
+      "epoch": 0.017105559306774702,
+      "grad_norm": 5.53125,
+      "learning_rate": 0.00019989456243410154,
+      "loss": 1.8394,
+      "step": 57
+    },
+    {
+      "epoch": 0.017405656838472505,
+      "grad_norm": 5.96875,
+      "learning_rate": 0.00019987949992468745,
+      "loss": 1.713,
+      "step": 58
+    },
+    {
+      "epoch": 0.017705754370170305,
+      "grad_norm": 5.25,
+      "learning_rate": 0.0001998644374152734,
+      "loss": 1.8177,
+      "step": 59
+    },
+    {
+      "epoch": 0.018005851901868108,
+      "grad_norm": 6.03125,
+      "learning_rate": 0.00019984937490585933,
+      "loss": 1.8893,
+      "step": 60
+    },
+    {
+      "epoch": 0.018305949433565907,
+      "grad_norm": 5.78125,
+      "learning_rate": 0.00019983431239644524,
+      "loss": 1.7885,
+      "step": 61
+    },
+    {
+      "epoch": 0.01860604696526371,
+      "grad_norm": 4.5,
+      "learning_rate": 0.00019981924988703118,
+      "loss": 1.7431,
+      "step": 62
+    },
+    {
+      "epoch": 0.018906144496961513,
+      "grad_norm": 5.25,
+      "learning_rate": 0.00019980418737761713,
+      "loss": 1.8099,
+      "step": 63
+    },
+    {
+      "epoch": 0.019206242028659313,
+      "grad_norm": 4.53125,
+      "learning_rate": 0.00019978912486820304,
+      "loss": 1.8839,
+      "step": 64
+    },
+    {
+      "epoch": 0.019506339560357116,
+      "grad_norm": 5.15625,
+      "learning_rate": 0.000199774062358789,
+      "loss": 1.6072,
+      "step": 65
+    },
+    {
+      "epoch": 0.01980643709205492,
+      "grad_norm": 5.4375,
+      "learning_rate": 0.00019975899984937492,
+      "loss": 1.8463,
+      "step": 66
+    },
+    {
+      "epoch": 0.02010653462375272,
+      "grad_norm": 4.84375,
+      "learning_rate": 0.00019974393733996083,
+      "loss": 2.0039,
+      "step": 67
+    },
+    {
+      "epoch": 0.02040663215545052,
+      "grad_norm": 4.75,
+      "learning_rate": 0.0001997288748305468,
+      "loss": 1.8937,
+      "step": 68
+    },
+    {
+      "epoch": 0.020706729687148324,
+      "grad_norm": 4.9375,
+      "learning_rate": 0.0001997138123211327,
+      "loss": 1.782,
+      "step": 69
+    },
+    {
+      "epoch": 0.021006827218846124,
+      "grad_norm": 5.8125,
+      "learning_rate": 0.00019969874981171863,
+      "loss": 2.0115,
+      "step": 70
+    },
+    {
+      "epoch": 0.021306924750543927,
+      "grad_norm": 4.71875,
+      "learning_rate": 0.00019968368730230457,
+      "loss": 1.8325,
+      "step": 71
+    },
+    {
+      "epoch": 0.02160702228224173,
+      "grad_norm": 5.15625,
+      "learning_rate": 0.0001996686247928905,
+      "loss": 1.882,
+      "step": 72
+    },
+    {
+      "epoch": 0.02190711981393953,
+      "grad_norm": 4.40625,
+      "learning_rate": 0.00019965356228347645,
+      "loss": 2.0092,
+      "step": 73
+    },
+    {
+      "epoch": 0.022207217345637333,
+      "grad_norm": 4.65625,
+      "learning_rate": 0.00019963849977406236,
+      "loss": 1.7513,
+      "step": 74
+    },
+    {
+      "epoch": 0.022507314877335136,
+      "grad_norm": 5.1875,
+      "learning_rate": 0.0001996234372646483,
+      "loss": 1.9309,
+      "step": 75
+    },
+    {
+      "epoch": 0.022807412409032935,
+      "grad_norm": 6.5,
+      "learning_rate": 0.00019960837475523424,
+      "loss": 1.9402,
+      "step": 76
+    },
+    {
+      "epoch": 0.023107509940730738,
+      "grad_norm": 5.1875,
+      "learning_rate": 0.00019959331224582015,
+      "loss": 1.9941,
+      "step": 77
+    },
+    {
+      "epoch": 0.023407607472428538,
+      "grad_norm": 4.40625,
+      "learning_rate": 0.0001995782497364061,
+      "loss": 2.0266,
+      "step": 78
+    },
+    {
+      "epoch": 0.02370770500412634,
+      "grad_norm": 4.625,
+      "learning_rate": 0.00019956318722699203,
+      "loss": 1.8835,
+      "step": 79
+    },
+    {
+      "epoch": 0.024007802535824144,
+      "grad_norm": 4.71875,
+      "learning_rate": 0.00019954812471757795,
+      "loss": 2.0939,
+      "step": 80
+    },
+    {
+      "epoch": 0.024307900067521943,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.0001995330622081639,
+      "loss": 1.7804,
+      "step": 81
+    },
+    {
+      "epoch": 0.024607997599219746,
+      "grad_norm": 4.03125,
+      "learning_rate": 0.00019951799969874983,
+      "loss": 1.573,
+      "step": 82
+    },
+    {
+      "epoch": 0.02490809513091755,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.00019950293718933574,
+      "loss": 1.8337,
+      "step": 83
+    },
+    {
+      "epoch": 0.02520819266261535,
+      "grad_norm": 4.34375,
+      "learning_rate": 0.00019948787467992168,
+      "loss": 1.9067,
+      "step": 84
+    },
+    {
+      "epoch": 0.025508290194313152,
+      "grad_norm": 4.34375,
+      "learning_rate": 0.00019947281217050762,
+      "loss": 1.7579,
+      "step": 85
+    },
+    {
+      "epoch": 0.025808387726010955,
+      "grad_norm": 5.1875,
+      "learning_rate": 0.00019945774966109354,
+      "loss": 1.908,
+      "step": 86
+    },
+    {
+      "epoch": 0.026108485257708754,
+      "grad_norm": 3.75,
+      "learning_rate": 0.00019944268715167948,
+      "loss": 1.758,
+      "step": 87
+    },
+    {
+      "epoch": 0.026408582789406557,
+      "grad_norm": 4.28125,
+      "learning_rate": 0.00019942762464226542,
+      "loss": 1.9805,
+      "step": 88
+    },
+    {
+      "epoch": 0.02670868032110436,
+      "grad_norm": 4.25,
+      "learning_rate": 0.00019941256213285133,
+      "loss": 1.8596,
+      "step": 89
+    },
+    {
+      "epoch": 0.02700877785280216,
+      "grad_norm": 4.5625,
+      "learning_rate": 0.00019939749962343727,
+      "loss": 2.0147,
+      "step": 90
+    },
+    {
+      "epoch": 0.027308875384499963,
+      "grad_norm": 4.6875,
+      "learning_rate": 0.0001993824371140232,
+      "loss": 2.0307,
+      "step": 91
+    },
+    {
+      "epoch": 0.027608972916197766,
+      "grad_norm": 4.40625,
+      "learning_rate": 0.00019936737460460915,
+      "loss": 1.7792,
+      "step": 92
+    },
+    {
+      "epoch": 0.027909070447895565,
+      "grad_norm": 3.984375,
+      "learning_rate": 0.00019935231209519506,
+      "loss": 1.9743,
+      "step": 93
+    },
+    {
+      "epoch": 0.02820916797959337,
+      "grad_norm": 4.15625,
+      "learning_rate": 0.000199337249585781,
+      "loss": 2.0801,
+      "step": 94
+    },
+    {
+      "epoch": 0.028509265511291168,
+      "grad_norm": 4.25,
+      "learning_rate": 0.00019932218707636694,
+      "loss": 1.8371,
+      "step": 95
+    },
+    {
+      "epoch": 0.02880936304298897,
+      "grad_norm": 3.703125,
+      "learning_rate": 0.00019930712456695286,
+      "loss": 1.8714,
+      "step": 96
+    },
+    {
+      "epoch": 0.029109460574686774,
+      "grad_norm": 3.9375,
+      "learning_rate": 0.0001992920620575388,
+      "loss": 1.7964,
+      "step": 97
+    },
+    {
+      "epoch": 0.029409558106384574,
+      "grad_norm": 4.25,
+      "learning_rate": 0.00019927699954812474,
+      "loss": 1.9042,
+      "step": 98
+    },
+    {
+      "epoch": 0.029709655638082377,
+      "grad_norm": 3.921875,
+      "learning_rate": 0.00019926193703871065,
+      "loss": 2.0471,
+      "step": 99
+    },
+    {
+      "epoch": 0.03000975316978018,
+      "grad_norm": 3.703125,
+      "learning_rate": 0.0001992468745292966,
+      "loss": 1.8343,
+      "step": 100
+    },
+    {
+      "epoch": 0.03030985070147798,
+      "grad_norm": 3.640625,
+      "learning_rate": 0.00019923181201988253,
+      "loss": 1.8421,
+      "step": 101
+    },
+    {
+      "epoch": 0.030609948233175782,
+      "grad_norm": 3.765625,
+      "learning_rate": 0.00019921674951046844,
+      "loss": 2.033,
+      "step": 102
+    },
+    {
+      "epoch": 0.030910045764873585,
+      "grad_norm": 3.90625,
+      "learning_rate": 0.00019920168700105438,
+      "loss": 2.0294,
+      "step": 103
+    },
+    {
+      "epoch": 0.031210143296571385,
+      "grad_norm": 4.5,
+      "learning_rate": 0.00019918662449164033,
+      "loss": 1.9534,
+      "step": 104
+    },
+    {
+      "epoch": 0.031510240828269184,
+      "grad_norm": 3.9375,
+      "learning_rate": 0.00019917156198222624,
+      "loss": 2.0067,
+      "step": 105
+    },
+    {
+      "epoch": 0.03181033835996699,
+      "grad_norm": 3.984375,
+      "learning_rate": 0.00019915649947281218,
+      "loss": 1.9351,
+      "step": 106
+    },
+    {
+      "epoch": 0.03211043589166479,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00019914143696339812,
+      "loss": 1.7907,
+      "step": 107
+    },
+    {
+      "epoch": 0.03241053342336259,
+      "grad_norm": 3.84375,
+      "learning_rate": 0.00019912637445398403,
+      "loss": 1.9125,
+      "step": 108
+    },
+    {
+      "epoch": 0.032710630955060396,
+      "grad_norm": 3.984375,
+      "learning_rate": 0.00019911131194456997,
+      "loss": 1.9038,
+      "step": 109
+    },
+    {
+      "epoch": 0.0330107284867582,
+      "grad_norm": 3.5625,
+      "learning_rate": 0.0001990962494351559,
+      "loss": 2.016,
+      "step": 110
+    },
+    {
+      "epoch": 0.033310826018455995,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.00019908118692574185,
+      "loss": 1.8636,
+      "step": 111
+    },
+    {
+      "epoch": 0.0336109235501538,
+      "grad_norm": 3.46875,
+      "learning_rate": 0.00019906612441632777,
+      "loss": 1.7806,
+      "step": 112
+    },
+    {
+      "epoch": 0.0339110210818516,
+      "grad_norm": 3.890625,
+      "learning_rate": 0.0001990510619069137,
+      "loss": 1.9326,
+      "step": 113
+    },
+    {
+      "epoch": 0.034211118613549404,
+      "grad_norm": 4.84375,
+      "learning_rate": 0.00019903599939749965,
+      "loss": 1.8132,
+      "step": 114
+    },
+    {
+      "epoch": 0.03451121614524721,
+      "grad_norm": 3.90625,
+      "learning_rate": 0.00019902093688808556,
+      "loss": 1.907,
+      "step": 115
+    },
+    {
+      "epoch": 0.03481131367694501,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00019900587437867147,
+      "loss": 1.9876,
+      "step": 116
+    },
+    {
+      "epoch": 0.035111411208642807,
+      "grad_norm": 3.640625,
+      "learning_rate": 0.00019899081186925744,
+      "loss": 1.9407,
+      "step": 117
+    },
+    {
+      "epoch": 0.03541150874034061,
+      "grad_norm": 3.453125,
+      "learning_rate": 0.00019897574935984335,
+      "loss": 2.0497,
+      "step": 118
+    },
+    {
+      "epoch": 0.03571160627203841,
+      "grad_norm": 3.703125,
+      "learning_rate": 0.0001989606868504293,
+      "loss": 2.0436,
+      "step": 119
+    },
+    {
+      "epoch": 0.036011703803736216,
+      "grad_norm": 3.65625,
+      "learning_rate": 0.00019894562434101523,
+      "loss": 1.8777,
+      "step": 120
+    },
+    {
+      "epoch": 0.03631180133543402,
+      "grad_norm": 3.921875,
+      "learning_rate": 0.00019893056183160115,
+      "loss": 1.7188,
+      "step": 121
+    },
+    {
+      "epoch": 0.036611898867131815,
+      "grad_norm": 3.828125,
+      "learning_rate": 0.0001989154993221871,
+      "loss": 1.9898,
+      "step": 122
+    },
+    {
+      "epoch": 0.03691199639882962,
+      "grad_norm": 3.78125,
+      "learning_rate": 0.00019890043681277303,
+      "loss": 1.7972,
+      "step": 123
+    },
+    {
+      "epoch": 0.03721209393052742,
+      "grad_norm": 5.0625,
+      "learning_rate": 0.00019888537430335894,
+      "loss": 2.0151,
+      "step": 124
+    },
+    {
+      "epoch": 0.037512191462225224,
+      "grad_norm": 3.28125,
+      "learning_rate": 0.00019887031179394488,
+      "loss": 1.8565,
+      "step": 125
+    },
+    {
+      "epoch": 0.03781228899392303,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.00019885524928453082,
+      "loss": 1.7748,
+      "step": 126
+    },
+    {
+      "epoch": 0.03811238652562083,
+      "grad_norm": 3.53125,
+      "learning_rate": 0.00019884018677511676,
+      "loss": 2.1386,
+      "step": 127
+    },
+    {
+      "epoch": 0.038412484057318626,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.00019882512426570268,
+      "loss": 1.8583,
+      "step": 128
+    },
+    {
+      "epoch": 0.03871258158901643,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.0001988100617562886,
+      "loss": 1.6611,
+      "step": 129
+    },
+    {
+      "epoch": 0.03901267912071423,
+      "grad_norm": 3.609375,
+      "learning_rate": 0.00019879499924687456,
+      "loss": 1.8367,
+      "step": 130
+    },
+    {
+      "epoch": 0.039312776652412035,
+      "grad_norm": 3.515625,
+      "learning_rate": 0.00019877993673746047,
+      "loss": 1.6843,
+      "step": 131
+    },
+    {
+      "epoch": 0.03961287418410984,
+      "grad_norm": 3.25,
+      "learning_rate": 0.00019876487422804638,
+      "loss": 1.8017,
+      "step": 132
+    },
+    {
+      "epoch": 0.03991297171580764,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00019874981171863235,
+      "loss": 1.7009,
+      "step": 133
+    },
+    {
+      "epoch": 0.04021306924750544,
+      "grad_norm": 3.96875,
+      "learning_rate": 0.00019873474920921826,
+      "loss": 2.0435,
+      "step": 134
+    },
+    {
+      "epoch": 0.04051316677920324,
+      "grad_norm": 3.6875,
+      "learning_rate": 0.00019871968669980418,
+      "loss": 1.9264,
+      "step": 135
+    },
+    {
+      "epoch": 0.04081326431090104,
+      "grad_norm": 3.46875,
+      "learning_rate": 0.00019870462419039014,
+      "loss": 1.8442,
+      "step": 136
+    },
+    {
+      "epoch": 0.041113361842598846,
+      "grad_norm": 3.671875,
+      "learning_rate": 0.00019868956168097606,
+      "loss": 1.853,
+      "step": 137
+    },
+    {
+      "epoch": 0.04141345937429665,
+      "grad_norm": 3.375,
+      "learning_rate": 0.000198674499171562,
+      "loss": 1.8712,
+      "step": 138
+    },
+    {
+      "epoch": 0.041713556905994445,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.0001986594366621479,
+      "loss": 1.6733,
+      "step": 139
+    },
+    {
+      "epoch": 0.04201365443769225,
+      "grad_norm": 3.53125,
+      "learning_rate": 0.00019864437415273385,
+      "loss": 2.0395,
+      "step": 140
+    },
+    {
+      "epoch": 0.04231375196939005,
+      "grad_norm": 3.671875,
+      "learning_rate": 0.0001986293116433198,
+      "loss": 1.9471,
+      "step": 141
+    },
+    {
+      "epoch": 0.042613849501087854,
+      "grad_norm": 3.78125,
+      "learning_rate": 0.0001986142491339057,
+      "loss": 1.8903,
+      "step": 142
+    },
+    {
+      "epoch": 0.04291394703278566,
+      "grad_norm": 4.3125,
+      "learning_rate": 0.00019859918662449164,
+      "loss": 1.9424,
+      "step": 143
+    },
+    {
+      "epoch": 0.04321404456448346,
+      "grad_norm": 3.546875,
+      "learning_rate": 0.00019858412411507759,
+      "loss": 1.9033,
+      "step": 144
+    },
+    {
+      "epoch": 0.043514142096181256,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.0001985690616056635,
+      "loss": 1.7218,
+      "step": 145
+    },
+    {
+      "epoch": 0.04381423962787906,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.00019855399909624947,
+      "loss": 1.7865,
+      "step": 146
+    },
+    {
+      "epoch": 0.04411433715957686,
+      "grad_norm": 3.21875,
+      "learning_rate": 0.00019853893658683538,
+      "loss": 1.8312,
+      "step": 147
+    },
+    {
+      "epoch": 0.044414434691274665,
+      "grad_norm": 3.625,
+      "learning_rate": 0.0001985238740774213,
+      "loss": 1.8489,
+      "step": 148
+    },
+    {
+      "epoch": 0.04471453222297247,
+      "grad_norm": 4.3125,
+      "learning_rate": 0.00019850881156800726,
+      "loss": 2.0047,
+      "step": 149
+    },
+    {
+      "epoch": 0.04501462975467027,
+      "grad_norm": 3.25,
+      "learning_rate": 0.00019849374905859317,
+      "loss": 1.9503,
+      "step": 150
+    },
+    {
+      "epoch": 0.04531472728636807,
+      "grad_norm": 3.5625,
+      "learning_rate": 0.00019847868654917909,
+      "loss": 1.782,
+      "step": 151
+    },
+    {
+      "epoch": 0.04561482481806587,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.00019846362403976503,
+      "loss": 1.8253,
+      "step": 152
+    },
+    {
+      "epoch": 0.04591492234976367,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.00019844856153035097,
+      "loss": 2.0763,
+      "step": 153
+    },
+    {
+      "epoch": 0.046215019881461476,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.0001984334990209369,
+      "loss": 2.0554,
+      "step": 154
+    },
+    {
+      "epoch": 0.04651511741315928,
+      "grad_norm": 3.53125,
+      "learning_rate": 0.00019841843651152282,
+      "loss": 1.8726,
+      "step": 155
+    },
+    {
+      "epoch": 0.046815214944857075,
+      "grad_norm": 3.53125,
+      "learning_rate": 0.00019840337400210876,
+      "loss": 1.844,
+      "step": 156
+    },
+    {
+      "epoch": 0.04711531247655488,
+      "grad_norm": 3.78125,
+      "learning_rate": 0.0001983883114926947,
+      "loss": 1.9448,
+      "step": 157
+    },
+    {
+      "epoch": 0.04741541000825268,
+      "grad_norm": 3.484375,
+      "learning_rate": 0.00019837324898328061,
+      "loss": 1.7888,
+      "step": 158
+    },
+    {
+      "epoch": 0.047715507539950484,
+      "grad_norm": 3.625,
+      "learning_rate": 0.00019835818647386655,
+      "loss": 1.833,
+      "step": 159
+    },
+    {
+      "epoch": 0.04801560507164829,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.0001983431239644525,
+      "loss": 1.9482,
+      "step": 160
+    },
+    {
+      "epoch": 0.04831570260334609,
+      "grad_norm": 3.375,
+      "learning_rate": 0.0001983280614550384,
+      "loss": 1.7975,
+      "step": 161
+    },
+    {
+      "epoch": 0.048615800135043886,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00019831299894562435,
+      "loss": 1.7426,
+      "step": 162
+    },
+    {
+      "epoch": 0.04891589766674169,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.0001982979364362103,
+      "loss": 1.729,
+      "step": 163
+    },
+    {
+      "epoch": 0.04921599519843949,
+      "grad_norm": 3.390625,
+      "learning_rate": 0.0001982828739267962,
+      "loss": 1.6922,
+      "step": 164
+    },
+    {
+      "epoch": 0.049516092730137296,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00019826781141738214,
+      "loss": 1.7636,
+      "step": 165
+    },
+    {
+      "epoch": 0.0498161902618351,
+      "grad_norm": 3.171875,
+      "learning_rate": 0.00019825274890796808,
+      "loss": 1.7565,
+      "step": 166
+    },
+    {
+      "epoch": 0.0501162877935329,
+      "grad_norm": 2.875,
+      "learning_rate": 0.000198237686398554,
+      "loss": 1.8834,
+      "step": 167
+    },
+    {
+      "epoch": 0.0504163853252307,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00019822262388913994,
+      "loss": 1.7744,
+      "step": 168
+    },
+    {
+      "epoch": 0.0507164828569285,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00019820756137972588,
+      "loss": 2.0087,
+      "step": 169
+    },
+    {
+      "epoch": 0.051016580388626304,
+      "grad_norm": 3.828125,
+      "learning_rate": 0.0001981924988703118,
+      "loss": 1.8857,
+      "step": 170
+    },
+    {
+      "epoch": 0.05131667792032411,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00019817743636089773,
+      "loss": 1.8271,
+      "step": 171
+    },
+    {
+      "epoch": 0.05161677545202191,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00019816237385148367,
+      "loss": 1.952,
+      "step": 172
+    },
+    {
+      "epoch": 0.051916872983719706,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.0001981473113420696,
+      "loss": 1.9371,
+      "step": 173
+    },
+    {
+      "epoch": 0.05221697051541751,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00019813224883265552,
+      "loss": 1.7685,
+      "step": 174
+    },
+    {
+      "epoch": 0.05251706804711531,
+      "grad_norm": 3.328125,
+      "learning_rate": 0.00019811718632324146,
+      "loss": 1.8866,
+      "step": 175
+    },
+    {
+      "epoch": 0.052817165578813115,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.0001981021238138274,
+      "loss": 1.7749,
+      "step": 176
+    },
+    {
+      "epoch": 0.05311726311051092,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00019808706130441332,
+      "loss": 1.9592,
+      "step": 177
+    },
+    {
+      "epoch": 0.05341736064220872,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00019807199879499926,
+      "loss": 1.7798,
+      "step": 178
+    },
+    {
+      "epoch": 0.05371745817390652,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001980569362855852,
+      "loss": 1.8398,
+      "step": 179
+    },
+    {
+      "epoch": 0.05401755570560432,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001980418737761711,
+      "loss": 1.8683,
+      "step": 180
+    },
+    {
+      "epoch": 0.05431765323730212,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00019802681126675705,
+      "loss": 1.779,
+      "step": 181
+    },
+    {
+      "epoch": 0.054617750768999926,
+      "grad_norm": 3.34375,
+      "learning_rate": 0.000198011748757343,
+      "loss": 1.7244,
+      "step": 182
+    },
+    {
+      "epoch": 0.05491784830069773,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001979966862479289,
+      "loss": 1.7576,
+      "step": 183
+    },
+    {
+      "epoch": 0.05521794583239553,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00019798162373851485,
+      "loss": 1.8803,
+      "step": 184
+    },
+    {
+      "epoch": 0.05551804336409333,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00019796656122910079,
+      "loss": 1.7587,
+      "step": 185
+    },
+    {
+      "epoch": 0.05581814089579113,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.0001979514987196867,
+      "loss": 1.8982,
+      "step": 186
+    },
+    {
+      "epoch": 0.056118238427488934,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00019793643621027264,
+      "loss": 1.5534,
+      "step": 187
+    },
+    {
+      "epoch": 0.05641833595918674,
+      "grad_norm": 3.765625,
+      "learning_rate": 0.00019792137370085858,
+      "loss": 1.7182,
+      "step": 188
+    },
+    {
+      "epoch": 0.05671843349088454,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.0001979063111914445,
+      "loss": 1.8385,
+      "step": 189
+    },
+    {
+      "epoch": 0.057018531022582336,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00019789124868203043,
+      "loss": 1.7305,
+      "step": 190
+    },
+    {
+      "epoch": 0.05731862855428014,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00019787618617261637,
+      "loss": 1.9204,
+      "step": 191
+    },
+    {
+      "epoch": 0.05761872608597794,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001978611236632023,
+      "loss": 1.921,
+      "step": 192
+    },
+    {
+      "epoch": 0.057918823617675745,
+      "grad_norm": 3.625,
+      "learning_rate": 0.00019784606115378823,
+      "loss": 1.5947,
+      "step": 193
+    },
+    {
+      "epoch": 0.05821892114937355,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00019783099864437417,
+      "loss": 1.7543,
+      "step": 194
+    },
+    {
+      "epoch": 0.05851901868107135,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.0001978159361349601,
+      "loss": 1.8031,
+      "step": 195
+    },
+    {
+      "epoch": 0.05881911621276915,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00019780087362554602,
+      "loss": 1.6785,
+      "step": 196
+    },
+    {
+      "epoch": 0.05911921374446695,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00019778581111613196,
+      "loss": 2.0096,
+      "step": 197
+    },
+    {
+      "epoch": 0.05941931127616475,
+      "grad_norm": 3.140625,
+      "learning_rate": 0.0001977707486067179,
+      "loss": 1.8979,
+      "step": 198
+    },
+    {
+      "epoch": 0.059719408807862556,
+      "grad_norm": 3.484375,
+      "learning_rate": 0.00019775568609730381,
+      "loss": 1.7718,
+      "step": 199
+    },
+    {
+      "epoch": 0.06001950633956036,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.00019774062358788975,
+      "loss": 1.8331,
+      "step": 200
+    },
+    {
+      "epoch": 0.06031960387125816,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.0001977255610784757,
+      "loss": 1.7295,
+      "step": 201
+    },
+    {
+      "epoch": 0.06061970140295596,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.0001977104985690616,
+      "loss": 2.0634,
+      "step": 202
+    },
+    {
+      "epoch": 0.06091979893465376,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019769543605964755,
+      "loss": 1.769,
+      "step": 203
+    },
+    {
+      "epoch": 0.061219896466351564,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001976803735502335,
+      "loss": 1.86,
+      "step": 204
+    },
+    {
+      "epoch": 0.06151999399804937,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.0001976653110408194,
+      "loss": 1.7321,
+      "step": 205
+    },
+    {
+      "epoch": 0.06182009152974717,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00019765024853140534,
+      "loss": 1.7122,
+      "step": 206
+    },
+    {
+      "epoch": 0.062120189061444966,
+      "grad_norm": 3.265625,
+      "learning_rate": 0.00019763518602199128,
+      "loss": 1.7178,
+      "step": 207
+    },
+    {
+      "epoch": 0.06242028659314277,
+      "grad_norm": 2.640625,
+      "learning_rate": 0.00019762012351257722,
+      "loss": 1.697,
+      "step": 208
+    },
+    {
+      "epoch": 0.06272038412484057,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00019760506100316314,
+      "loss": 1.8518,
+      "step": 209
+    },
+    {
+      "epoch": 0.06302048165653837,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00019758999849374905,
+      "loss": 1.8185,
+      "step": 210
+    },
+    {
+      "epoch": 0.06332057918823618,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00019757493598433502,
+      "loss": 1.7672,
+      "step": 211
+    },
+    {
+      "epoch": 0.06362067671993397,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00019755987347492093,
+      "loss": 1.6654,
+      "step": 212
+    },
+    {
+      "epoch": 0.06392077425163178,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00019754481096550684,
+      "loss": 1.9504,
+      "step": 213
+    },
+    {
+      "epoch": 0.06422087178332958,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.0001975297484560928,
+      "loss": 1.7497,
+      "step": 214
+    },
+    {
+      "epoch": 0.06452096931502739,
+      "grad_norm": 3.296875,
+      "learning_rate": 0.00019751468594667872,
+      "loss": 1.8027,
+      "step": 215
+    },
+    {
+      "epoch": 0.06482106684672519,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00019749962343726464,
+      "loss": 1.8369,
+      "step": 216
+    },
+    {
+      "epoch": 0.06512116437842298,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.0001974845609278506,
+      "loss": 1.7718,
+      "step": 217
+    },
+    {
+      "epoch": 0.06542126191012079,
+      "grad_norm": 3.0625,
+      "learning_rate": 0.00019746949841843652,
+      "loss": 2.0664,
+      "step": 218
+    },
+    {
+      "epoch": 0.06572135944181859,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.00019745443590902246,
+      "loss": 1.6568,
+      "step": 219
+    },
+    {
+      "epoch": 0.0660214569735164,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001974393733996084,
+      "loss": 1.8081,
+      "step": 220
+    },
+    {
+      "epoch": 0.0663215545052142,
+      "grad_norm": 3.484375,
+      "learning_rate": 0.0001974243108901943,
+      "loss": 1.7602,
+      "step": 221
+    },
+    {
+      "epoch": 0.06662165203691199,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00019740924838078025,
+      "loss": 1.6876,
+      "step": 222
+    },
+    {
+      "epoch": 0.0669217495686098,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00019739418587136616,
+      "loss": 1.8168,
+      "step": 223
+    },
+    {
+      "epoch": 0.0672218471003076,
+      "grad_norm": 2.5,
+      "learning_rate": 0.0001973791233619521,
+      "loss": 1.7603,
+      "step": 224
+    },
+    {
+      "epoch": 0.0675219446320054,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00019736406085253805,
+      "loss": 1.9572,
+      "step": 225
+    },
+    {
+      "epoch": 0.0678220421637032,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00019734899834312396,
+      "loss": 1.8395,
+      "step": 226
+    },
+    {
+      "epoch": 0.068122139695401,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00019733393583370993,
+      "loss": 1.8317,
+      "step": 227
+    },
+    {
+      "epoch": 0.06842223722709881,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.00019731887332429584,
+      "loss": 1.6134,
+      "step": 228
+    },
+    {
+      "epoch": 0.0687223347587966,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00019730381081488175,
+      "loss": 1.7573,
+      "step": 229
+    },
+    {
+      "epoch": 0.06902243229049441,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00019728874830546772,
+      "loss": 1.7313,
+      "step": 230
+    },
+    {
+      "epoch": 0.06932252982219221,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00019727368579605363,
+      "loss": 1.878,
+      "step": 231
+    },
+    {
+      "epoch": 0.06962262735389002,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019725862328663955,
+      "loss": 1.8936,
+      "step": 232
+    },
+    {
+      "epoch": 0.06992272488558782,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.0001972435607772255,
+      "loss": 1.7405,
+      "step": 233
+    },
+    {
+      "epoch": 0.07022282241728561,
+      "grad_norm": 3.765625,
+      "learning_rate": 0.00019722849826781143,
+      "loss": 1.7309,
+      "step": 234
+    },
+    {
+      "epoch": 0.07052291994898342,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00019721343575839737,
+      "loss": 1.7067,
+      "step": 235
+    },
+    {
+      "epoch": 0.07082301748068122,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00019719837324898328,
+      "loss": 1.9431,
+      "step": 236
+    },
+    {
+      "epoch": 0.07112311501237903,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00019718331073956922,
+      "loss": 1.8008,
+      "step": 237
+    },
+    {
+      "epoch": 0.07142321254407683,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00019716824823015516,
+      "loss": 1.8448,
+      "step": 238
+    },
+    {
+      "epoch": 0.07172331007577462,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00019715318572074107,
+      "loss": 1.8422,
+      "step": 239
+    },
+    {
+      "epoch": 0.07202340760747243,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00019713812321132701,
+      "loss": 1.934,
+      "step": 240
+    },
+    {
+      "epoch": 0.07232350513917023,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00019712306070191295,
+      "loss": 1.7717,
+      "step": 241
+    },
+    {
+      "epoch": 0.07262360267086804,
+      "grad_norm": 4.8125,
+      "learning_rate": 0.00019710799819249887,
+      "loss": 1.845,
+      "step": 242
+    },
+    {
+      "epoch": 0.07292370020256583,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001970929356830848,
+      "loss": 1.8812,
+      "step": 243
+    },
+    {
+      "epoch": 0.07322379773426363,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00019707787317367075,
+      "loss": 1.7325,
+      "step": 244
+    },
+    {
+      "epoch": 0.07352389526596144,
+      "grad_norm": 2.875,
+      "learning_rate": 0.00019706281066425666,
+      "loss": 1.7831,
+      "step": 245
+    },
+    {
+      "epoch": 0.07382399279765924,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001970477481548426,
+      "loss": 1.7124,
+      "step": 246
+    },
+    {
+      "epoch": 0.07412409032935705,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00019703268564542854,
+      "loss": 2.0327,
+      "step": 247
+    },
+    {
+      "epoch": 0.07442418786105484,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00019701762313601446,
+      "loss": 1.76,
+      "step": 248
+    },
+    {
+      "epoch": 0.07472428539275265,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001970025606266004,
+      "loss": 1.6563,
+      "step": 249
+    },
+    {
+      "epoch": 0.07502438292445045,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00019698749811718634,
+      "loss": 1.7377,
+      "step": 250
+    },
+    {
+      "epoch": 0.07532448045614824,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00019697243560777225,
+      "loss": 1.761,
+      "step": 251
+    },
+    {
+      "epoch": 0.07562457798784605,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.0001969573730983582,
+      "loss": 1.7535,
+      "step": 252
+    },
+    {
+      "epoch": 0.07592467551954385,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00019694231058894413,
+      "loss": 1.7543,
+      "step": 253
+    },
+    {
+      "epoch": 0.07622477305124166,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00019692724807953007,
+      "loss": 1.8524,
+      "step": 254
+    },
+    {
+      "epoch": 0.07652487058293946,
+      "grad_norm": 3.796875,
+      "learning_rate": 0.00019691218557011598,
+      "loss": 1.9389,
+      "step": 255
+    },
+    {
+      "epoch": 0.07682496811463725,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00019689712306070192,
+      "loss": 1.8954,
+      "step": 256
+    },
+    {
+      "epoch": 0.07712506564633506,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00019688206055128786,
+      "loss": 1.7184,
+      "step": 257
+    },
+    {
+      "epoch": 0.07742516317803286,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00019686699804187378,
+      "loss": 1.7767,
+      "step": 258
+    },
+    {
+      "epoch": 0.07772526070973067,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00019685193553245972,
+      "loss": 1.7224,
+      "step": 259
+    },
+    {
+      "epoch": 0.07802535824142846,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00019683687302304566,
+      "loss": 1.7314,
+      "step": 260
+    },
+    {
+      "epoch": 0.07832545577312626,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00019682181051363157,
+      "loss": 1.8375,
+      "step": 261
+    },
+    {
+      "epoch": 0.07862555330482407,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001968067480042175,
+      "loss": 1.9115,
+      "step": 262
+    },
+    {
+      "epoch": 0.07892565083652187,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00019679168549480345,
+      "loss": 1.8343,
+      "step": 263
+    },
+    {
+      "epoch": 0.07922574836821968,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00019677662298538936,
+      "loss": 1.7809,
+      "step": 264
+    },
+    {
+      "epoch": 0.07952584589991747,
+      "grad_norm": 2.8125,
+      "learning_rate": 0.0001967615604759753,
+      "loss": 1.7773,
+      "step": 265
+    },
+    {
+      "epoch": 0.07982594343161528,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00019674649796656125,
+      "loss": 1.9381,
+      "step": 266
+    },
+    {
+      "epoch": 0.08012604096331308,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019673143545714716,
+      "loss": 1.8356,
+      "step": 267
+    },
+    {
+      "epoch": 0.08042613849501087,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.0001967163729477331,
+      "loss": 1.6324,
+      "step": 268
+    },
+    {
+      "epoch": 0.08072623602670868,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00019670131043831904,
+      "loss": 1.7117,
+      "step": 269
+    },
+    {
+      "epoch": 0.08102633355840648,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00019668624792890495,
+      "loss": 1.9915,
+      "step": 270
+    },
+    {
+      "epoch": 0.08132643109010429,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001966711854194909,
+      "loss": 1.836,
+      "step": 271
+    },
+    {
+      "epoch": 0.08162652862180209,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00019665612291007683,
+      "loss": 1.7106,
+      "step": 272
+    },
+    {
+      "epoch": 0.08192662615349988,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00019664106040066277,
+      "loss": 1.773,
+      "step": 273
+    },
+    {
+      "epoch": 0.08222672368519769,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001966259978912487,
+      "loss": 1.6944,
+      "step": 274
+    },
+    {
+      "epoch": 0.08252682121689549,
+      "grad_norm": 21.0,
+      "learning_rate": 0.00019661093538183463,
+      "loss": 1.7898,
+      "step": 275
+    },
+    {
+      "epoch": 0.0828269187485933,
+      "grad_norm": 19.125,
+      "learning_rate": 0.00019659587287242057,
+      "loss": 1.9784,
+      "step": 276
+    },
+    {
+      "epoch": 0.0831270162802911,
+      "grad_norm": 4.3125,
+      "learning_rate": 0.00019658081036300648,
+      "loss": 1.929,
+      "step": 277
+    },
+    {
+      "epoch": 0.08342711381198889,
+      "grad_norm": 6.5625,
+      "learning_rate": 0.00019656574785359242,
+      "loss": 1.931,
+      "step": 278
+    },
+    {
+      "epoch": 0.0837272113436867,
+      "grad_norm": 3.09375,
+      "learning_rate": 0.00019655068534417836,
+      "loss": 1.7778,
+      "step": 279
+    },
+    {
+      "epoch": 0.0840273088753845,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00019653562283476427,
+      "loss": 1.7492,
+      "step": 280
+    },
+    {
+      "epoch": 0.0843274064070823,
+      "grad_norm": 56.0,
+      "learning_rate": 0.00019652056032535021,
+      "loss": 2.2216,
+      "step": 281
+    },
+    {
+      "epoch": 0.0846275039387801,
+      "grad_norm": 4.1875,
+      "learning_rate": 0.00019650549781593615,
+      "loss": 2.1483,
+      "step": 282
+    },
+    {
+      "epoch": 0.08492760147047791,
+      "grad_norm": 16.375,
+      "learning_rate": 0.00019649043530652207,
+      "loss": 1.8711,
+      "step": 283
+    },
+    {
+      "epoch": 0.08522769900217571,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.000196475372797108,
+      "loss": 1.9292,
+      "step": 284
+    },
+    {
+      "epoch": 0.0855277965338735,
+      "grad_norm": 16.0,
+      "learning_rate": 0.00019646031028769395,
+      "loss": 2.6547,
+      "step": 285
+    },
+    {
+      "epoch": 0.08582789406557131,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00019644524777827986,
+      "loss": 1.8063,
+      "step": 286
+    },
+    {
+      "epoch": 0.08612799159726911,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.0001964301852688658,
+      "loss": 1.8171,
+      "step": 287
+    },
+    {
+      "epoch": 0.08642808912896692,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00019641512275945174,
+      "loss": 1.9172,
+      "step": 288
+    },
+    {
+      "epoch": 0.08672818666066472,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00019640006025003768,
+      "loss": 1.6614,
+      "step": 289
+    },
+    {
+      "epoch": 0.08702828419236251,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001963849977406236,
+      "loss": 1.7848,
+      "step": 290
+    },
+    {
+      "epoch": 0.08732838172406032,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00019636993523120954,
+      "loss": 1.7673,
+      "step": 291
+    },
+    {
+      "epoch": 0.08762847925575812,
+      "grad_norm": 3.25,
+      "learning_rate": 0.00019635487272179548,
+      "loss": 2.2057,
+      "step": 292
+    },
+    {
+      "epoch": 0.08792857678745593,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001963398102123814,
+      "loss": 1.7921,
+      "step": 293
+    },
+    {
+      "epoch": 0.08822867431915372,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001963247477029673,
+      "loss": 1.8208,
+      "step": 294
+    },
+    {
+      "epoch": 0.08852877185085152,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00019630968519355327,
+      "loss": 1.8308,
+      "step": 295
+    },
+    {
+      "epoch": 0.08882886938254933,
+      "grad_norm": 5.34375,
+      "learning_rate": 0.00019629462268413918,
+      "loss": 2.2499,
+      "step": 296
+    },
+    {
+      "epoch": 0.08912896691424713,
+      "grad_norm": 3.140625,
+      "learning_rate": 0.0001962795601747251,
+      "loss": 1.6618,
+      "step": 297
+    },
+    {
+      "epoch": 0.08942906444594494,
+      "grad_norm": 4.25,
+      "learning_rate": 0.00019626449766531106,
+      "loss": 2.149,
+      "step": 298
+    },
+    {
+      "epoch": 0.08972916197764273,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019624943515589698,
+      "loss": 1.7057,
+      "step": 299
+    },
+    {
+      "epoch": 0.09002925950934054,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00019623437264648292,
+      "loss": 1.6305,
+      "step": 300
+    },
+    {
+      "epoch": 0.09032935704103834,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00019621931013706886,
+      "loss": 1.574,
+      "step": 301
+    },
+    {
+      "epoch": 0.09062945457273613,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00019620424762765477,
+      "loss": 1.7159,
+      "step": 302
+    },
+    {
+      "epoch": 0.09092955210443394,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001961891851182407,
+      "loss": 1.5729,
+      "step": 303
+    },
+    {
+      "epoch": 0.09122964963613174,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00019617412260882665,
+      "loss": 1.6326,
+      "step": 304
+    },
+    {
+      "epoch": 0.09152974716782955,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019615906009941256,
+      "loss": 1.6142,
+      "step": 305
+    },
+    {
+      "epoch": 0.09182984469952735,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001961439975899985,
+      "loss": 1.8829,
+      "step": 306
+    },
+    {
+      "epoch": 0.09212994223122514,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00019612893508058442,
+      "loss": 1.7353,
+      "step": 307
+    },
+    {
+      "epoch": 0.09243003976292295,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00019611387257117039,
+      "loss": 1.7764,
+      "step": 308
+    },
+    {
+      "epoch": 0.09273013729462075,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001960988100617563,
+      "loss": 1.809,
+      "step": 309
+    },
+    {
+      "epoch": 0.09303023482631856,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001960837475523422,
+      "loss": 1.8253,
+      "step": 310
+    },
+    {
+      "epoch": 0.09333033235801635,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019606868504292818,
+      "loss": 1.608,
+      "step": 311
+    },
+    {
+      "epoch": 0.09363042988971415,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001960536225335141,
+      "loss": 1.5565,
+      "step": 312
+    },
+    {
+      "epoch": 0.09393052742141196,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001960385600241,
+      "loss": 1.8403,
+      "step": 313
+    },
+    {
+      "epoch": 0.09423062495310976,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00019602349751468597,
+      "loss": 1.7102,
+      "step": 314
+    },
+    {
+      "epoch": 0.09453072248480757,
+      "grad_norm": 8.0625,
+      "learning_rate": 0.0001960084350052719,
+      "loss": 1.9664,
+      "step": 315
+    },
+    {
+      "epoch": 0.09483082001650536,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00019599337249585783,
+      "loss": 1.6402,
+      "step": 316
+    },
+    {
+      "epoch": 0.09513091754820317,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00019597830998644374,
+      "loss": 1.7076,
+      "step": 317
+    },
+    {
+      "epoch": 0.09543101507990097,
+      "grad_norm": 4.3125,
+      "learning_rate": 0.00019596324747702968,
+      "loss": 1.8879,
+      "step": 318
+    },
+    {
+      "epoch": 0.09573111261159876,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019594818496761562,
+      "loss": 1.756,
+      "step": 319
+    },
+    {
+      "epoch": 0.09603121014329657,
+      "grad_norm": 17.625,
+      "learning_rate": 0.00019593312245820153,
+      "loss": 1.8216,
+      "step": 320
+    },
+    {
+      "epoch": 0.09633130767499437,
+      "grad_norm": 4.03125,
+      "learning_rate": 0.00019591805994878747,
+      "loss": 1.6627,
+      "step": 321
+    },
+    {
+      "epoch": 0.09663140520669218,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019590299743937341,
+      "loss": 1.7324,
+      "step": 322
+    },
+    {
+      "epoch": 0.09693150273838998,
+      "grad_norm": 4.40625,
+      "learning_rate": 0.00019588793492995933,
+      "loss": 1.6513,
+      "step": 323
+    },
+    {
+      "epoch": 0.09723160027008777,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019587287242054527,
+      "loss": 1.5848,
+      "step": 324
+    },
+    {
+      "epoch": 0.09753169780178558,
+      "grad_norm": 4.0625,
+      "learning_rate": 0.0001958578099111312,
+      "loss": 1.7972,
+      "step": 325
+    },
+    {
+      "epoch": 0.09783179533348338,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019584274740171712,
+      "loss": 1.8432,
+      "step": 326
+    },
+    {
+      "epoch": 0.09813189286518119,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001958276848923031,
+      "loss": 1.6936,
+      "step": 327
+    },
+    {
+      "epoch": 0.09843199039687898,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.000195812622382889,
+      "loss": 1.6162,
+      "step": 328
+    },
+    {
+      "epoch": 0.09873208792857678,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00019579755987347492,
+      "loss": 1.7071,
+      "step": 329
+    },
+    {
+      "epoch": 0.09903218546027459,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.00019578249736406086,
+      "loss": 1.7842,
+      "step": 330
+    },
+    {
+      "epoch": 0.09933228299197239,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.0001957674348546468,
+      "loss": 1.8356,
+      "step": 331
+    },
+    {
+      "epoch": 0.0996323805236702,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001957523723452327,
+      "loss": 1.6637,
+      "step": 332
+    },
+    {
+      "epoch": 0.099932478055368,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019573730983581865,
+      "loss": 1.7073,
+      "step": 333
+    },
+    {
+      "epoch": 0.1002325755870658,
+      "grad_norm": 7.25,
+      "learning_rate": 0.0001957222473264046,
+      "loss": 1.958,
+      "step": 334
+    },
+    {
+      "epoch": 0.1005326731187636,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00019570718481699053,
+      "loss": 1.7528,
+      "step": 335
+    },
+    {
+      "epoch": 0.1008327706504614,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019569212230757644,
+      "loss": 1.6439,
+      "step": 336
+    },
+    {
+      "epoch": 0.1011328681821592,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00019567705979816238,
+      "loss": 1.7725,
+      "step": 337
+    },
+    {
+      "epoch": 0.101432965713857,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019566199728874832,
+      "loss": 1.6004,
+      "step": 338
+    },
+    {
+      "epoch": 0.10173306324555481,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00019564693477933424,
+      "loss": 1.78,
+      "step": 339
+    },
+    {
+      "epoch": 0.10203316077725261,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019563187226992018,
+      "loss": 1.7048,
+      "step": 340
+    },
+    {
+      "epoch": 0.1023332583089504,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019561680976050612,
+      "loss": 1.6206,
+      "step": 341
+    },
+    {
+      "epoch": 0.10263335584064821,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019560174725109203,
+      "loss": 1.6979,
+      "step": 342
+    },
+    {
+      "epoch": 0.10293345337234601,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00019558668474167797,
+      "loss": 1.7589,
+      "step": 343
+    },
+    {
+      "epoch": 0.10323355090404382,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001955716222322639,
+      "loss": 1.7095,
+      "step": 344
+    },
+    {
+      "epoch": 0.10353364843574162,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00019555655972284982,
+      "loss": 1.8464,
+      "step": 345
+    },
+    {
+      "epoch": 0.10383374596743941,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019554149721343577,
+      "loss": 1.5749,
+      "step": 346
+    },
+    {
+      "epoch": 0.10413384349913722,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001955264347040217,
+      "loss": 1.7944,
+      "step": 347
+    },
+    {
+      "epoch": 0.10443394103083502,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00019551137219460762,
+      "loss": 1.7216,
+      "step": 348
+    },
+    {
+      "epoch": 0.10473403856253283,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00019549630968519356,
+      "loss": 1.5322,
+      "step": 349
+    },
+    {
+      "epoch": 0.10503413609423062,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.0001954812471757795,
+      "loss": 1.6603,
+      "step": 350
+    },
+    {
+      "epoch": 0.10533423362592843,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001954661846663654,
+      "loss": 1.7032,
+      "step": 351
+    },
+    {
+      "epoch": 0.10563433115762623,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00019545112215695135,
+      "loss": 1.609,
+      "step": 352
+    },
+    {
+      "epoch": 0.10593442868932403,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.0001954360596475373,
+      "loss": 1.6667,
+      "step": 353
+    },
+    {
+      "epoch": 0.10623452622102184,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019542099713812323,
+      "loss": 1.6547,
+      "step": 354
+    },
+    {
+      "epoch": 0.10653462375271963,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00019540593462870915,
+      "loss": 1.7228,
+      "step": 355
+    },
+    {
+      "epoch": 0.10683472128441744,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001953908721192951,
+      "loss": 1.6998,
+      "step": 356
+    },
+    {
+      "epoch": 0.10713481881611524,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019537580960988103,
+      "loss": 1.7126,
+      "step": 357
+    },
+    {
+      "epoch": 0.10743491634781303,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00019536074710046694,
+      "loss": 1.7248,
+      "step": 358
+    },
+    {
+      "epoch": 0.10773501387951084,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00019534568459105288,
+      "loss": 1.7882,
+      "step": 359
+    },
+    {
+      "epoch": 0.10803511141120864,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00019533062208163882,
+      "loss": 1.7691,
+      "step": 360
+    },
+    {
+      "epoch": 0.10833520894290645,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019531555957222473,
+      "loss": 1.5793,
+      "step": 361
+    },
+    {
+      "epoch": 0.10863530647460425,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00019530049706281067,
+      "loss": 1.767,
+      "step": 362
+    },
+    {
+      "epoch": 0.10893540400630204,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00019528543455339661,
+      "loss": 1.533,
+      "step": 363
+    },
+    {
+      "epoch": 0.10923550153799985,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00019527037204398253,
+      "loss": 1.8839,
+      "step": 364
+    },
+    {
+      "epoch": 0.10953559906969765,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00019525530953456847,
+      "loss": 1.5946,
+      "step": 365
+    },
+    {
+      "epoch": 0.10983569660139546,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001952402470251544,
+      "loss": 1.8419,
+      "step": 366
+    },
+    {
+      "epoch": 0.11013579413309325,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019522518451574032,
+      "loss": 1.7132,
+      "step": 367
+    },
+    {
+      "epoch": 0.11043589166479106,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00019521012200632626,
+      "loss": 1.7913,
+      "step": 368
+    },
+    {
+      "epoch": 0.11073598919648886,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001951950594969122,
+      "loss": 1.8396,
+      "step": 369
+    },
+    {
+      "epoch": 0.11103608672818666,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019517999698749814,
+      "loss": 1.8348,
+      "step": 370
+    },
+    {
+      "epoch": 0.11133618425988447,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019516493447808406,
+      "loss": 1.8373,
+      "step": 371
+    },
+    {
+      "epoch": 0.11163628179158226,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00019514987196867,
+      "loss": 1.7057,
+      "step": 372
+    },
+    {
+      "epoch": 0.11193637932328007,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00019513480945925594,
+      "loss": 1.7544,
+      "step": 373
+    },
+    {
+      "epoch": 0.11223647685497787,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00019511974694984185,
+      "loss": 1.5365,
+      "step": 374
+    },
+    {
+      "epoch": 0.11253657438667566,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019510468444042776,
+      "loss": 1.6592,
+      "step": 375
+    },
+    {
+      "epoch": 0.11283667191837347,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019508962193101373,
+      "loss": 1.669,
+      "step": 376
+    },
+    {
+      "epoch": 0.11313676945007127,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00019507455942159964,
+      "loss": 1.8414,
+      "step": 377
+    },
+    {
+      "epoch": 0.11343686698176908,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00019505949691218556,
+      "loss": 1.8678,
+      "step": 378
+    },
+    {
+      "epoch": 0.11373696451346688,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019504443440277152,
+      "loss": 1.8424,
+      "step": 379
+    },
+    {
+      "epoch": 0.11403706204516467,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019502937189335744,
+      "loss": 1.6625,
+      "step": 380
+    },
+    {
+      "epoch": 0.11433715957686248,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00019501430938394338,
+      "loss": 1.8434,
+      "step": 381
+    },
+    {
+      "epoch": 0.11463725710856028,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00019499924687452932,
+      "loss": 1.8069,
+      "step": 382
+    },
+    {
+      "epoch": 0.11493735464025809,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019498418436511523,
+      "loss": 1.8178,
+      "step": 383
+    },
+    {
+      "epoch": 0.11523745217195588,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019496912185570117,
+      "loss": 1.7506,
+      "step": 384
+    },
+    {
+      "epoch": 0.1155375497036537,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.0001949540593462871,
+      "loss": 1.7893,
+      "step": 385
+    },
+    {
+      "epoch": 0.11583764723535149,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019493899683687302,
+      "loss": 1.6928,
+      "step": 386
+    },
+    {
+      "epoch": 0.11613774476704929,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00019492393432745897,
+      "loss": 1.5413,
+      "step": 387
+    },
+    {
+      "epoch": 0.1164378422987471,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00019490887181804488,
+      "loss": 1.8343,
+      "step": 388
+    },
+    {
+      "epoch": 0.11673793983044489,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00019489380930863085,
+      "loss": 1.4793,
+      "step": 389
+    },
+    {
+      "epoch": 0.1170380373621427,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019487874679921676,
+      "loss": 1.672,
+      "step": 390
+    },
+    {
+      "epoch": 0.1173381348938405,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00019486368428980267,
+      "loss": 1.5806,
+      "step": 391
+    },
+    {
+      "epoch": 0.1176382324255383,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019484862178038864,
+      "loss": 1.7068,
+      "step": 392
+    },
+    {
+      "epoch": 0.1179383299572361,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00019483355927097455,
+      "loss": 1.4543,
+      "step": 393
+    },
+    {
+      "epoch": 0.1182384274889339,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019481849676156047,
+      "loss": 1.803,
+      "step": 394
+    },
+    {
+      "epoch": 0.11853852502063171,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019480343425214643,
+      "loss": 1.4074,
+      "step": 395
+    },
+    {
+      "epoch": 0.1188386225523295,
+      "grad_norm": 3.640625,
+      "learning_rate": 0.00019478837174273235,
+      "loss": 2.0418,
+      "step": 396
+    },
+    {
+      "epoch": 0.1191387200840273,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001947733092333183,
+      "loss": 1.6686,
+      "step": 397
+    },
+    {
+      "epoch": 0.11943881761572511,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00019475824672390423,
+      "loss": 1.737,
+      "step": 398
+    },
+    {
+      "epoch": 0.11973891514742291,
+      "grad_norm": 7.28125,
+      "learning_rate": 0.00019474318421449014,
+      "loss": 2.0573,
+      "step": 399
+    },
+    {
+      "epoch": 0.12003901267912072,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019472812170507608,
+      "loss": 1.6827,
+      "step": 400
+    },
+    {
+      "epoch": 0.12033911021081851,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.000194713059195662,
+      "loss": 1.6224,
+      "step": 401
+    },
+    {
+      "epoch": 0.12063920774251632,
+      "grad_norm": 8.8125,
+      "learning_rate": 0.00019469799668624793,
+      "loss": 2.3057,
+      "step": 402
+    },
+    {
+      "epoch": 0.12093930527421412,
+      "grad_norm": 3.8125,
+      "learning_rate": 0.00019468293417683387,
+      "loss": 2.3373,
+      "step": 403
+    },
+    {
+      "epoch": 0.12123940280591192,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001946678716674198,
+      "loss": 1.7069,
+      "step": 404
+    },
+    {
+      "epoch": 0.12153950033760973,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019465280915800573,
+      "loss": 1.587,
+      "step": 405
+    },
+    {
+      "epoch": 0.12183959786930752,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00019463774664859167,
+      "loss": 1.5961,
+      "step": 406
+    },
+    {
+      "epoch": 0.12213969540100533,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00019462268413917758,
+      "loss": 1.6449,
+      "step": 407
+    },
+    {
+      "epoch": 0.12243979293270313,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019460762162976355,
+      "loss": 1.7905,
+      "step": 408
+    },
+    {
+      "epoch": 0.12273989046440092,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00019459255912034946,
+      "loss": 1.754,
+      "step": 409
+    },
+    {
+      "epoch": 0.12303998799609873,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00019457749661093538,
+      "loss": 2.0521,
+      "step": 410
+    },
+    {
+      "epoch": 0.12334008552779653,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00019456243410152132,
+      "loss": 1.8169,
+      "step": 411
+    },
+    {
+      "epoch": 0.12364018305949434,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019454737159210726,
+      "loss": 1.5915,
+      "step": 412
+    },
+    {
+      "epoch": 0.12394028059119214,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.00019453230908269317,
+      "loss": 2.2229,
+      "step": 413
+    },
+    {
+      "epoch": 0.12424037812288993,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001945172465732791,
+      "loss": 1.7073,
+      "step": 414
+    },
+    {
+      "epoch": 0.12454047565458774,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00019450218406386505,
+      "loss": 1.5338,
+      "step": 415
+    },
+    {
+      "epoch": 0.12484057318628554,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.000194487121554451,
+      "loss": 1.7315,
+      "step": 416
+    },
+    {
+      "epoch": 0.12514067071798335,
+      "grad_norm": 3.421875,
+      "learning_rate": 0.0001944720590450369,
+      "loss": 1.8115,
+      "step": 417
+    },
+    {
+      "epoch": 0.12544076824968114,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00019445699653562284,
+      "loss": 1.6631,
+      "step": 418
+    },
+    {
+      "epoch": 0.12574086578137894,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019444193402620878,
+      "loss": 1.8004,
+      "step": 419
+    },
+    {
+      "epoch": 0.12604096331307674,
+      "grad_norm": 16.875,
+      "learning_rate": 0.0001944268715167947,
+      "loss": 2.2407,
+      "step": 420
+    },
+    {
+      "epoch": 0.12634106084477456,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00019441180900738064,
+      "loss": 1.6495,
+      "step": 421
+    },
+    {
+      "epoch": 0.12664115837647236,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00019439674649796658,
+      "loss": 1.7246,
+      "step": 422
+    },
+    {
+      "epoch": 0.12694125590817015,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001943816839885525,
+      "loss": 1.6171,
+      "step": 423
+    },
+    {
+      "epoch": 0.12724135343986795,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00019436662147913843,
+      "loss": 1.6409,
+      "step": 424
+    },
+    {
+      "epoch": 0.12754145097156575,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00019435155896972437,
+      "loss": 1.7021,
+      "step": 425
+    },
+    {
+      "epoch": 0.12784154850326357,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019433649646031028,
+      "loss": 1.7784,
+      "step": 426
+    },
+    {
+      "epoch": 0.12814164603496137,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019432143395089623,
+      "loss": 1.7024,
+      "step": 427
+    },
+    {
+      "epoch": 0.12844174356665916,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019430637144148217,
+      "loss": 1.8126,
+      "step": 428
+    },
+    {
+      "epoch": 0.12874184109835696,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00019429130893206808,
+      "loss": 1.6615,
+      "step": 429
+    },
+    {
+      "epoch": 0.12904193863005478,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019427624642265402,
+      "loss": 1.7268,
+      "step": 430
+    },
+    {
+      "epoch": 0.12934203616175258,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019426118391323996,
+      "loss": 1.4514,
+      "step": 431
+    },
+    {
+      "epoch": 0.12964213369345037,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019424612140382587,
+      "loss": 1.4657,
+      "step": 432
+    },
+    {
+      "epoch": 0.12994223122514817,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001942310588944118,
+      "loss": 1.8722,
+      "step": 433
+    },
+    {
+      "epoch": 0.13024232875684597,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019421599638499775,
+      "loss": 1.4007,
+      "step": 434
+    },
+    {
+      "epoch": 0.1305424262885438,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.0001942009338755837,
+      "loss": 1.5361,
+      "step": 435
+    },
+    {
+      "epoch": 0.13084252382024159,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001941858713661696,
+      "loss": 1.7993,
+      "step": 436
+    },
+    {
+      "epoch": 0.13114262135193938,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00019417080885675555,
+      "loss": 1.5655,
+      "step": 437
+    },
+    {
+      "epoch": 0.13144271888363718,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001941557463473415,
+      "loss": 1.4928,
+      "step": 438
+    },
+    {
+      "epoch": 0.13174281641533497,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001941406838379274,
+      "loss": 1.7321,
+      "step": 439
+    },
+    {
+      "epoch": 0.1320429139470328,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019412562132851334,
+      "loss": 1.6269,
+      "step": 440
+    },
+    {
+      "epoch": 0.1323430114787306,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019411055881909928,
+      "loss": 1.8111,
+      "step": 441
+    },
+    {
+      "epoch": 0.1326431090104284,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.0001940954963096852,
+      "loss": 1.4506,
+      "step": 442
+    },
+    {
+      "epoch": 0.13294320654212619,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00019408043380027113,
+      "loss": 1.9013,
+      "step": 443
+    },
+    {
+      "epoch": 0.13324330407382398,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00019406537129085707,
+      "loss": 1.4885,
+      "step": 444
+    },
+    {
+      "epoch": 0.1335434016055218,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.000194050308781443,
+      "loss": 1.5648,
+      "step": 445
+    },
+    {
+      "epoch": 0.1338434991372196,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00019403524627202893,
+      "loss": 1.7881,
+      "step": 446
+    },
+    {
+      "epoch": 0.1341435966689174,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00019402018376261487,
+      "loss": 1.4731,
+      "step": 447
+    },
+    {
+      "epoch": 0.1344436942006152,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00019400512125320078,
+      "loss": 1.6967,
+      "step": 448
+    },
+    {
+      "epoch": 0.134743791732313,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019399005874378672,
+      "loss": 1.6821,
+      "step": 449
+    },
+    {
+      "epoch": 0.1350438892640108,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019397499623437266,
+      "loss": 1.6326,
+      "step": 450
+    },
+    {
+      "epoch": 0.1353439867957086,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001939599337249586,
+      "loss": 1.5639,
+      "step": 451
+    },
+    {
+      "epoch": 0.1356440843274064,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019394487121554452,
+      "loss": 1.5422,
+      "step": 452
+    },
+    {
+      "epoch": 0.1359441818591042,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019392980870613046,
+      "loss": 1.7708,
+      "step": 453
+    },
+    {
+      "epoch": 0.136244279390802,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001939147461967164,
+      "loss": 1.7786,
+      "step": 454
+    },
+    {
+      "epoch": 0.13654437692249982,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001938996836873023,
+      "loss": 1.6996,
+      "step": 455
+    },
+    {
+      "epoch": 0.13684447445419762,
+      "grad_norm": 2.3125,
+      "learning_rate": 0.00019388462117788825,
+      "loss": 1.4751,
+      "step": 456
+    },
+    {
+      "epoch": 0.1371445719858954,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001938695586684742,
+      "loss": 1.7451,
+      "step": 457
+    },
+    {
+      "epoch": 0.1374446695175932,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001938544961590601,
+      "loss": 1.57,
+      "step": 458
+    },
+    {
+      "epoch": 0.137744767049291,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00019383943364964602,
+      "loss": 1.5973,
+      "step": 459
+    },
+    {
+      "epoch": 0.13804486458098883,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00019382437114023198,
+      "loss": 1.8124,
+      "step": 460
+    },
+    {
+      "epoch": 0.13834496211268663,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001938093086308179,
+      "loss": 1.9279,
+      "step": 461
+    },
+    {
+      "epoch": 0.13864505964438442,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00019379424612140384,
+      "loss": 1.567,
+      "step": 462
+    },
+    {
+      "epoch": 0.13894515717608222,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00019377918361198978,
+      "loss": 1.6679,
+      "step": 463
+    },
+    {
+      "epoch": 0.13924525470778004,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001937641211025757,
+      "loss": 1.5801,
+      "step": 464
+    },
+    {
+      "epoch": 0.13954535223947784,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019374905859316163,
+      "loss": 1.6882,
+      "step": 465
+    },
+    {
+      "epoch": 0.13984544977117563,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00019373399608374757,
+      "loss": 1.7908,
+      "step": 466
+    },
+    {
+      "epoch": 0.14014554730287343,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019371893357433348,
+      "loss": 1.7243,
+      "step": 467
+    },
+    {
+      "epoch": 0.14044564483457123,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019370387106491943,
+      "loss": 1.7183,
+      "step": 468
+    },
+    {
+      "epoch": 0.14074574236626905,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019368880855550537,
+      "loss": 1.7466,
+      "step": 469
+    },
+    {
+      "epoch": 0.14104583989796685,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001936737460460913,
+      "loss": 1.6636,
+      "step": 470
+    },
+    {
+      "epoch": 0.14134593742966464,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00019365868353667722,
+      "loss": 1.7036,
+      "step": 471
+    },
+    {
+      "epoch": 0.14164603496136244,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019364362102726313,
+      "loss": 1.5873,
+      "step": 472
+    },
+    {
+      "epoch": 0.14194613249306023,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.0001936285585178491,
+      "loss": 1.9267,
+      "step": 473
+    },
+    {
+      "epoch": 0.14224623002475806,
+      "grad_norm": 6.15625,
+      "learning_rate": 0.000193613496008435,
+      "loss": 1.6562,
+      "step": 474
+    },
+    {
+      "epoch": 0.14254632755645585,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00019359843349902093,
+      "loss": 1.7047,
+      "step": 475
+    },
+    {
+      "epoch": 0.14284642508815365,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.0001935833709896069,
+      "loss": 1.7173,
+      "step": 476
+    },
+    {
+      "epoch": 0.14314652261985145,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001935683084801928,
+      "loss": 1.6671,
+      "step": 477
+    },
+    {
+      "epoch": 0.14344662015154924,
+      "grad_norm": 7.0,
+      "learning_rate": 0.00019355324597077875,
+      "loss": 1.6172,
+      "step": 478
+    },
+    {
+      "epoch": 0.14374671768324707,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001935381834613647,
+      "loss": 1.5094,
+      "step": 479
+    },
+    {
+      "epoch": 0.14404681521494486,
+      "grad_norm": 240.0,
+      "learning_rate": 0.0001935231209519506,
+      "loss": 1.748,
+      "step": 480
+    },
+    {
+      "epoch": 0.14434691274664266,
+      "grad_norm": 1568.0,
+      "learning_rate": 0.00019350805844253654,
+      "loss": 4.3148,
+      "step": 481
+    },
+    {
+      "epoch": 0.14464701027834045,
+      "grad_norm": 350.0,
+      "learning_rate": 0.00019349299593312245,
+      "loss": 7.3686,
+      "step": 482
+    },
+    {
+      "epoch": 0.14494710781003825,
+      "grad_norm": 76.0,
+      "learning_rate": 0.0001934779334237084,
+      "loss": 7.4777,
+      "step": 483
+    },
+    {
+      "epoch": 0.14524720534173607,
+      "grad_norm": 38.0,
+      "learning_rate": 0.00019346287091429433,
+      "loss": 6.1912,
+      "step": 484
+    },
+    {
+      "epoch": 0.14554730287343387,
+      "grad_norm": 39.25,
+      "learning_rate": 0.00019344780840488025,
+      "loss": 4.5296,
+      "step": 485
+    },
+    {
+      "epoch": 0.14584740040513167,
+      "grad_norm": 21.5,
+      "learning_rate": 0.0001934327458954662,
+      "loss": 3.1263,
+      "step": 486
+    },
+    {
+      "epoch": 0.14614749793682946,
+      "grad_norm": 72.0,
+      "learning_rate": 0.00019341768338605213,
+      "loss": 2.1498,
+      "step": 487
+    },
+    {
+      "epoch": 0.14644759546852726,
+      "grad_norm": 10.4375,
+      "learning_rate": 0.00019340262087663804,
+      "loss": 2.4546,
+      "step": 488
+    },
+    {
+      "epoch": 0.14674769300022508,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.000193387558367224,
+      "loss": 1.891,
+      "step": 489
+    },
+    {
+      "epoch": 0.14704779053192288,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00019337249585780992,
+      "loss": 1.8553,
+      "step": 490
+    },
+    {
+      "epoch": 0.14734788806362067,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00019335743334839584,
+      "loss": 1.7628,
+      "step": 491
+    },
+    {
+      "epoch": 0.14764798559531847,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001933423708389818,
+      "loss": 1.7328,
+      "step": 492
+    },
+    {
+      "epoch": 0.14794808312701627,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00019332730832956772,
+      "loss": 1.6661,
+      "step": 493
+    },
+    {
+      "epoch": 0.1482481806587141,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019331224582015363,
+      "loss": 1.5965,
+      "step": 494
+    },
+    {
+      "epoch": 0.1485482781904119,
+      "grad_norm": 4.59375,
+      "learning_rate": 0.00019329718331073957,
+      "loss": 1.6053,
+      "step": 495
+    },
+    {
+      "epoch": 0.14884837572210968,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001932821208013255,
+      "loss": 1.6873,
+      "step": 496
+    },
+    {
+      "epoch": 0.14914847325380748,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019326705829191145,
+      "loss": 1.5406,
+      "step": 497
+    },
+    {
+      "epoch": 0.1494485707855053,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019325199578249736,
+      "loss": 1.6759,
+      "step": 498
+    },
+    {
+      "epoch": 0.1497486683172031,
+      "grad_norm": 3.140625,
+      "learning_rate": 0.0001932369332730833,
+      "loss": 1.6577,
+      "step": 499
+    },
+    {
+      "epoch": 0.1500487658489009,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019322187076366924,
+      "loss": 1.6675,
+      "step": 500
+    },
+    {
+      "epoch": 0.1503488633805987,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00019320680825425516,
+      "loss": 1.6049,
+      "step": 501
+    },
+    {
+      "epoch": 0.1506489609122965,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001931917457448411,
+      "loss": 1.6966,
+      "step": 502
+    },
+    {
+      "epoch": 0.1509490584439943,
+      "grad_norm": 8.4375,
+      "learning_rate": 0.00019317668323542704,
+      "loss": 2.0639,
+      "step": 503
+    },
+    {
+      "epoch": 0.1512491559756921,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00019316162072601295,
+      "loss": 1.547,
+      "step": 504
+    },
+    {
+      "epoch": 0.1515492535073899,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019314655821659892,
+      "loss": 1.6564,
+      "step": 505
+    },
+    {
+      "epoch": 0.1518493510390877,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00019313149570718483,
+      "loss": 1.6156,
+      "step": 506
+    },
+    {
+      "epoch": 0.1521494485707855,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00019311643319777074,
+      "loss": 1.4015,
+      "step": 507
+    },
+    {
+      "epoch": 0.15244954610248332,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019310137068835669,
+      "loss": 1.5436,
+      "step": 508
+    },
+    {
+      "epoch": 0.15274964363418111,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00019308630817894263,
+      "loss": 1.8201,
+      "step": 509
+    },
+    {
+      "epoch": 0.1530497411658789,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00019307124566952854,
+      "loss": 1.7368,
+      "step": 510
+    },
+    {
+      "epoch": 0.1533498386975767,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019305618316011448,
+      "loss": 1.5794,
+      "step": 511
+    },
+    {
+      "epoch": 0.1536499362292745,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00019304112065070042,
+      "loss": 1.6935,
+      "step": 512
+    },
+    {
+      "epoch": 0.15395003376097233,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019302605814128633,
+      "loss": 1.5433,
+      "step": 513
+    },
+    {
+      "epoch": 0.15425013129267012,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019301099563187227,
+      "loss": 1.6764,
+      "step": 514
+    },
+    {
+      "epoch": 0.15455022882436792,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001929959331224582,
+      "loss": 1.7757,
+      "step": 515
+    },
+    {
+      "epoch": 0.15485032635606572,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00019298087061304415,
+      "loss": 1.5875,
+      "step": 516
+    },
+    {
+      "epoch": 0.1551504238877635,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00019296580810363007,
+      "loss": 1.4456,
+      "step": 517
+    },
+    {
+      "epoch": 0.15545052141946133,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.000192950745594216,
+      "loss": 1.5678,
+      "step": 518
+    },
+    {
+      "epoch": 0.15575061895115913,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00019293568308480195,
+      "loss": 1.5371,
+      "step": 519
+    },
+    {
+      "epoch": 0.15605071648285693,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00019292062057538786,
+      "loss": 1.7181,
+      "step": 520
+    },
+    {
+      "epoch": 0.15635081401455472,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001929055580659738,
+      "loss": 1.6222,
+      "step": 521
+    },
+    {
+      "epoch": 0.15665091154625252,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00019289049555655974,
+      "loss": 1.7834,
+      "step": 522
+    },
+    {
+      "epoch": 0.15695100907795034,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019287543304714565,
+      "loss": 1.4747,
+      "step": 523
+    },
+    {
+      "epoch": 0.15725110660964814,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001928603705377316,
+      "loss": 1.6354,
+      "step": 524
+    },
+    {
+      "epoch": 0.15755120414134594,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019284530802831753,
+      "loss": 1.827,
+      "step": 525
+    },
+    {
+      "epoch": 0.15785130167304373,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019283024551890345,
+      "loss": 1.6644,
+      "step": 526
+    },
+    {
+      "epoch": 0.15815139920474153,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.0001928151830094894,
+      "loss": 1.619,
+      "step": 527
+    },
+    {
+      "epoch": 0.15845149673643935,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00019280012050007533,
+      "loss": 1.6275,
+      "step": 528
+    },
+    {
+      "epoch": 0.15875159426813715,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00019278505799066124,
+      "loss": 1.7155,
+      "step": 529
+    },
+    {
+      "epoch": 0.15905169179983494,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019276999548124718,
+      "loss": 1.5752,
+      "step": 530
+    },
+    {
+      "epoch": 0.15935178933153274,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00019275493297183312,
+      "loss": 1.5679,
+      "step": 531
+    },
+    {
+      "epoch": 0.15965188686323056,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00019273987046241906,
+      "loss": 1.6421,
+      "step": 532
+    },
+    {
+      "epoch": 0.15995198439492836,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019272480795300498,
+      "loss": 1.6256,
+      "step": 533
+    },
+    {
+      "epoch": 0.16025208192662616,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00019270974544359092,
+      "loss": 1.5849,
+      "step": 534
+    },
+    {
+      "epoch": 0.16055217945832395,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00019269468293417686,
+      "loss": 1.7302,
+      "step": 535
+    },
+    {
+      "epoch": 0.16085227699002175,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00019267962042476277,
+      "loss": 1.7273,
+      "step": 536
+    },
+    {
+      "epoch": 0.16115237452171957,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001926645579153487,
+      "loss": 1.484,
+      "step": 537
+    },
+    {
+      "epoch": 0.16145247205341737,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019264949540593465,
+      "loss": 1.858,
+      "step": 538
+    },
+    {
+      "epoch": 0.16175256958511516,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019263443289652056,
+      "loss": 1.8602,
+      "step": 539
+    },
+    {
+      "epoch": 0.16205266711681296,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001926193703871065,
+      "loss": 1.775,
+      "step": 540
+    },
+    {
+      "epoch": 0.16235276464851076,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00019260430787769244,
+      "loss": 1.7084,
+      "step": 541
+    },
+    {
+      "epoch": 0.16265286218020858,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00019258924536827836,
+      "loss": 1.8766,
+      "step": 542
+    },
+    {
+      "epoch": 0.16295295971190638,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001925741828588643,
+      "loss": 1.836,
+      "step": 543
+    },
+    {
+      "epoch": 0.16325305724360417,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00019255912034945024,
+      "loss": 1.5194,
+      "step": 544
+    },
+    {
+      "epoch": 0.16355315477530197,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019254405784003615,
+      "loss": 1.6375,
+      "step": 545
+    },
+    {
+      "epoch": 0.16385325230699976,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.0001925289953306221,
+      "loss": 1.9624,
+      "step": 546
+    },
+    {
+      "epoch": 0.1641533498386976,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019251393282120803,
+      "loss": 1.6165,
+      "step": 547
+    },
+    {
+      "epoch": 0.16445344737039538,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019249887031179394,
+      "loss": 1.6888,
+      "step": 548
+    },
+    {
+      "epoch": 0.16475354490209318,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00019248380780237989,
+      "loss": 1.7981,
+      "step": 549
+    },
+    {
+      "epoch": 0.16505364243379098,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019246874529296583,
+      "loss": 1.4563,
+      "step": 550
+    },
+    {
+      "epoch": 0.16535373996548877,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00019245368278355177,
+      "loss": 1.6802,
+      "step": 551
+    },
+    {
+      "epoch": 0.1656538374971866,
+      "grad_norm": 2.375,
+      "learning_rate": 0.00019243862027413768,
+      "loss": 1.7944,
+      "step": 552
+    },
+    {
+      "epoch": 0.1659539350288844,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.0001924235577647236,
+      "loss": 1.5582,
+      "step": 553
+    },
+    {
+      "epoch": 0.1662540325605822,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019240849525530956,
+      "loss": 1.525,
+      "step": 554
+    },
+    {
+      "epoch": 0.16655413009227998,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019239343274589547,
+      "loss": 1.7083,
+      "step": 555
+    },
+    {
+      "epoch": 0.16685422762397778,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00019237837023648139,
+      "loss": 1.7821,
+      "step": 556
+    },
+    {
+      "epoch": 0.1671543251556756,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00019236330772706735,
+      "loss": 1.6192,
+      "step": 557
+    },
+    {
+      "epoch": 0.1674544226873734,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019234824521765327,
+      "loss": 1.615,
+      "step": 558
+    },
+    {
+      "epoch": 0.1677545202190712,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.0001923331827082392,
+      "loss": 1.7328,
+      "step": 559
+    },
+    {
+      "epoch": 0.168054617750769,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00019231812019882515,
+      "loss": 1.8038,
+      "step": 560
+    },
+    {
+      "epoch": 0.1683547152824668,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019230305768941106,
+      "loss": 1.5963,
+      "step": 561
+    },
+    {
+      "epoch": 0.1686548128141646,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.000192287995179997,
+      "loss": 1.6813,
+      "step": 562
+    },
+    {
+      "epoch": 0.1689549103458624,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019227293267058294,
+      "loss": 1.698,
+      "step": 563
+    },
+    {
+      "epoch": 0.1692550078775602,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00019225787016116885,
+      "loss": 1.7285,
+      "step": 564
+    },
+    {
+      "epoch": 0.169555105409258,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001922428076517548,
+      "loss": 1.7494,
+      "step": 565
+    },
+    {
+      "epoch": 0.16985520294095582,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001922277451423407,
+      "loss": 1.8757,
+      "step": 566
+    },
+    {
+      "epoch": 0.17015530047265362,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00019221268263292665,
+      "loss": 1.4615,
+      "step": 567
+    },
+    {
+      "epoch": 0.17045539800435142,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001921976201235126,
+      "loss": 1.6237,
+      "step": 568
+    },
+    {
+      "epoch": 0.1707554955360492,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001921825576140985,
+      "loss": 1.7211,
+      "step": 569
+    },
+    {
+      "epoch": 0.171055593067747,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019216749510468447,
+      "loss": 1.6005,
+      "step": 570
+    },
+    {
+      "epoch": 0.17135569059944483,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019215243259527038,
+      "loss": 1.7864,
+      "step": 571
+    },
+    {
+      "epoch": 0.17165578813114263,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001921373700858563,
+      "loss": 1.7053,
+      "step": 572
+    },
+    {
+      "epoch": 0.17195588566284042,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019212230757644226,
+      "loss": 1.6067,
+      "step": 573
+    },
+    {
+      "epoch": 0.17225598319453822,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00019210724506702818,
+      "loss": 1.569,
+      "step": 574
+    },
+    {
+      "epoch": 0.17255608072623602,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001920921825576141,
+      "loss": 1.677,
+      "step": 575
+    },
+    {
+      "epoch": 0.17285617825793384,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019207712004820006,
+      "loss": 1.5929,
+      "step": 576
+    },
+    {
+      "epoch": 0.17315627578963164,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019206205753878597,
+      "loss": 1.7629,
+      "step": 577
+    },
+    {
+      "epoch": 0.17345637332132943,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001920469950293719,
+      "loss": 1.6935,
+      "step": 578
+    },
+    {
+      "epoch": 0.17375647085302723,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00019203193251995782,
+      "loss": 1.5948,
+      "step": 579
+    },
+    {
+      "epoch": 0.17405656838472502,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00019201687001054376,
+      "loss": 1.555,
+      "step": 580
+    },
+    {
+      "epoch": 0.17435666591642285,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001920018075011297,
+      "loss": 1.7226,
+      "step": 581
+    },
+    {
+      "epoch": 0.17465676344812064,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019198674499171562,
+      "loss": 1.7919,
+      "step": 582
+    },
+    {
+      "epoch": 0.17495686097981844,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019197168248230156,
+      "loss": 1.7068,
+      "step": 583
+    },
+    {
+      "epoch": 0.17525695851151624,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001919566199728875,
+      "loss": 1.5244,
+      "step": 584
+    },
+    {
+      "epoch": 0.17555705604321403,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.0001919415574634734,
+      "loss": 1.763,
+      "step": 585
+    },
+    {
+      "epoch": 0.17585715357491186,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019192649495405938,
+      "loss": 1.7039,
+      "step": 586
+    },
+    {
+      "epoch": 0.17615725110660965,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001919114324446453,
+      "loss": 1.7103,
+      "step": 587
+    },
+    {
+      "epoch": 0.17645734863830745,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001918963699352312,
+      "loss": 1.6664,
+      "step": 588
+    },
+    {
+      "epoch": 0.17675744617000524,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00019188130742581715,
+      "loss": 1.7018,
+      "step": 589
+    },
+    {
+      "epoch": 0.17705754370170304,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019186624491640309,
+      "loss": 1.6809,
+      "step": 590
+    },
+    {
+      "epoch": 0.17735764123340086,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.000191851182406989,
+      "loss": 1.7552,
+      "step": 591
+    },
+    {
+      "epoch": 0.17765773876509866,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019183611989757494,
+      "loss": 1.5148,
+      "step": 592
+    },
+    {
+      "epoch": 0.17795783629679646,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019182105738816088,
+      "loss": 1.7134,
+      "step": 593
+    },
+    {
+      "epoch": 0.17825793382849425,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001918059948787468,
+      "loss": 1.5393,
+      "step": 594
+    },
+    {
+      "epoch": 0.17855803136019205,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019179093236933273,
+      "loss": 1.5201,
+      "step": 595
+    },
+    {
+      "epoch": 0.17885812889188987,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00019177586985991867,
+      "loss": 1.41,
+      "step": 596
+    },
+    {
+      "epoch": 0.17915822642358767,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001917608073505046,
+      "loss": 1.5989,
+      "step": 597
+    },
+    {
+      "epoch": 0.17945832395528546,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019174574484109053,
+      "loss": 1.6824,
+      "step": 598
+    },
+    {
+      "epoch": 0.17975842148698326,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00019173068233167647,
+      "loss": 1.6174,
+      "step": 599
+    },
+    {
+      "epoch": 0.18005851901868108,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.0001917156198222624,
+      "loss": 1.6333,
+      "step": 600
+    },
+    {
+      "epoch": 0.18035861655037888,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019170055731284832,
+      "loss": 1.5945,
+      "step": 601
+    },
+    {
+      "epoch": 0.18065871408207668,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00019168549480343426,
+      "loss": 1.5724,
+      "step": 602
+    },
+    {
+      "epoch": 0.18095881161377447,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001916704322940202,
+      "loss": 1.7408,
+      "step": 603
+    },
+    {
+      "epoch": 0.18125890914547227,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00019165536978460611,
+      "loss": 1.7874,
+      "step": 604
+    },
+    {
+      "epoch": 0.1815590066771701,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019164030727519205,
+      "loss": 1.8732,
+      "step": 605
+    },
+    {
+      "epoch": 0.1818591042088679,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.000191625244765778,
+      "loss": 1.5112,
+      "step": 606
+    },
+    {
+      "epoch": 0.18215920174056568,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001916101822563639,
+      "loss": 1.5816,
+      "step": 607
+    },
+    {
+      "epoch": 0.18245929927226348,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00019159511974694985,
+      "loss": 1.6178,
+      "step": 608
+    },
+    {
+      "epoch": 0.18275939680396128,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001915800572375358,
+      "loss": 1.5747,
+      "step": 609
+    },
+    {
+      "epoch": 0.1830594943356591,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001915649947281217,
+      "loss": 1.6899,
+      "step": 610
+    },
+    {
+      "epoch": 0.1833595918673569,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00019154993221870764,
+      "loss": 1.6481,
+      "step": 611
+    },
+    {
+      "epoch": 0.1836596893990547,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019153486970929358,
+      "loss": 1.7611,
+      "step": 612
+    },
+    {
+      "epoch": 0.1839597869307525,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00019151980719987952,
+      "loss": 1.6108,
+      "step": 613
+    },
+    {
+      "epoch": 0.18425988446245029,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00019150474469046544,
+      "loss": 1.6823,
+      "step": 614
+    },
+    {
+      "epoch": 0.1845599819941481,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019148968218105138,
+      "loss": 1.5958,
+      "step": 615
+    },
+    {
+      "epoch": 0.1848600795258459,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019147461967163732,
+      "loss": 1.7122,
+      "step": 616
+    },
+    {
+      "epoch": 0.1851601770575437,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019145955716222323,
+      "loss": 1.7203,
+      "step": 617
+    },
+    {
+      "epoch": 0.1854602745892415,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00019144449465280917,
+      "loss": 1.6483,
+      "step": 618
+    },
+    {
+      "epoch": 0.1857603721209393,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001914294321433951,
+      "loss": 1.6077,
+      "step": 619
+    },
+    {
+      "epoch": 0.18606046965263712,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00019141436963398102,
+      "loss": 1.7108,
+      "step": 620
+    },
+    {
+      "epoch": 0.1863605671843349,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00019139930712456696,
+      "loss": 1.727,
+      "step": 621
+    },
+    {
+      "epoch": 0.1866606647160327,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001913842446151529,
+      "loss": 1.6468,
+      "step": 622
+    },
+    {
+      "epoch": 0.1869607622477305,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019136918210573882,
+      "loss": 1.513,
+      "step": 623
+    },
+    {
+      "epoch": 0.1872608597794283,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00019135411959632476,
+      "loss": 1.7358,
+      "step": 624
+    },
+    {
+      "epoch": 0.18756095731112613,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001913390570869107,
+      "loss": 1.6106,
+      "step": 625
+    },
+    {
+      "epoch": 0.18786105484282392,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.0001913239945774966,
+      "loss": 1.5019,
+      "step": 626
+    },
+    {
+      "epoch": 0.18816115237452172,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019130893206808255,
+      "loss": 1.6404,
+      "step": 627
+    },
+    {
+      "epoch": 0.1884612499062195,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001912938695586685,
+      "loss": 1.6606,
+      "step": 628
+    },
+    {
+      "epoch": 0.1887613474379173,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001912788070492544,
+      "loss": 1.8266,
+      "step": 629
+    },
+    {
+      "epoch": 0.18906144496961513,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00019126374453984035,
+      "loss": 1.7289,
+      "step": 630
+    },
+    {
+      "epoch": 0.18936154250131293,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019124868203042629,
+      "loss": 1.6096,
+      "step": 631
+    },
+    {
+      "epoch": 0.18966164003301073,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00019123361952101223,
+      "loss": 1.673,
+      "step": 632
+    },
+    {
+      "epoch": 0.18996173756470852,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019121855701159814,
+      "loss": 1.6054,
+      "step": 633
+    },
+    {
+      "epoch": 0.19026183509640635,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00019120349450218408,
+      "loss": 1.8019,
+      "step": 634
+    },
+    {
+      "epoch": 0.19056193262810414,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00019118843199277002,
+      "loss": 1.5414,
+      "step": 635
+    },
+    {
+      "epoch": 0.19086203015980194,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00019117336948335593,
+      "loss": 1.5755,
+      "step": 636
+    },
+    {
+      "epoch": 0.19116212769149973,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019115830697394185,
+      "loss": 1.9447,
+      "step": 637
+    },
+    {
+      "epoch": 0.19146222522319753,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.0001911432444645278,
+      "loss": 1.7597,
+      "step": 638
+    },
+    {
+      "epoch": 0.19176232275489535,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00019112818195511373,
+      "loss": 1.6838,
+      "step": 639
+    },
+    {
+      "epoch": 0.19206242028659315,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00019111311944569967,
+      "loss": 1.5105,
+      "step": 640
+    },
+    {
+      "epoch": 0.19236251781829095,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.0001910980569362856,
+      "loss": 1.6408,
+      "step": 641
+    },
+    {
+      "epoch": 0.19266261534998874,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00019108299442687152,
+      "loss": 1.7393,
+      "step": 642
+    },
+    {
+      "epoch": 0.19296271288168654,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00019106793191745746,
+      "loss": 1.582,
+      "step": 643
+    },
+    {
+      "epoch": 0.19326281041338436,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001910528694080434,
+      "loss": 1.4669,
+      "step": 644
+    },
+    {
+      "epoch": 0.19356290794508216,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019103780689862931,
+      "loss": 1.57,
+      "step": 645
+    },
+    {
+      "epoch": 0.19386300547677995,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019102274438921525,
+      "loss": 1.7287,
+      "step": 646
+    },
+    {
+      "epoch": 0.19416310300847775,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00019100768187980117,
+      "loss": 1.5792,
+      "step": 647
+    },
+    {
+      "epoch": 0.19446320054017555,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001909926193703871,
+      "loss": 1.7222,
+      "step": 648
+    },
+    {
+      "epoch": 0.19476329807187337,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019097755686097305,
+      "loss": 1.4899,
+      "step": 649
+    },
+    {
+      "epoch": 0.19506339560357117,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019096249435155896,
+      "loss": 1.7813,
+      "step": 650
+    },
+    {
+      "epoch": 0.19536349313526896,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00019094743184214493,
+      "loss": 1.4786,
+      "step": 651
+    },
+    {
+      "epoch": 0.19566359066696676,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00019093236933273084,
+      "loss": 1.6588,
+      "step": 652
+    },
+    {
+      "epoch": 0.19596368819866455,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00019091730682331676,
+      "loss": 1.912,
+      "step": 653
+    },
+    {
+      "epoch": 0.19626378573036238,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00019090224431390272,
+      "loss": 1.3395,
+      "step": 654
+    },
+    {
+      "epoch": 0.19656388326206017,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00019088718180448864,
+      "loss": 1.6949,
+      "step": 655
+    },
+    {
+      "epoch": 0.19686398079375797,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019087211929507455,
+      "loss": 1.9453,
+      "step": 656
+    },
+    {
+      "epoch": 0.19716407832545577,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019085705678566052,
+      "loss": 1.6662,
+      "step": 657
+    },
+    {
+      "epoch": 0.19746417585715356,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00019084199427624643,
+      "loss": 1.5202,
+      "step": 658
+    },
+    {
+      "epoch": 0.19776427338885139,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00019082693176683237,
+      "loss": 1.4931,
+      "step": 659
+    },
+    {
+      "epoch": 0.19806437092054918,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00019081186925741828,
+      "loss": 1.661,
+      "step": 660
+    },
+    {
+      "epoch": 0.19836446845224698,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019079680674800422,
+      "loss": 1.6424,
+      "step": 661
+    },
+    {
+      "epoch": 0.19866456598394477,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019078174423859016,
+      "loss": 1.5423,
+      "step": 662
+    },
+    {
+      "epoch": 0.19896466351564257,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019076668172917608,
+      "loss": 1.5711,
+      "step": 663
+    },
+    {
+      "epoch": 0.1992647610473404,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019075161921976202,
+      "loss": 1.7005,
+      "step": 664
+    },
+    {
+      "epoch": 0.1995648585790382,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00019073655671034796,
+      "loss": 1.6474,
+      "step": 665
+    },
+    {
+      "epoch": 0.199864956110736,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019072149420093387,
+      "loss": 1.5861,
+      "step": 666
+    },
+    {
+      "epoch": 0.20016505364243378,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00019070643169151984,
+      "loss": 1.7536,
+      "step": 667
+    },
+    {
+      "epoch": 0.2004651511741316,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00019069136918210575,
+      "loss": 1.5817,
+      "step": 668
+    },
+    {
+      "epoch": 0.2007652487058294,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00019067630667269166,
+      "loss": 1.7074,
+      "step": 669
+    },
+    {
+      "epoch": 0.2010653462375272,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019066124416327763,
+      "loss": 1.5273,
+      "step": 670
+    },
+    {
+      "epoch": 0.201365443769225,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00019064618165386355,
+      "loss": 1.6839,
+      "step": 671
+    },
+    {
+      "epoch": 0.2016655413009228,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019063111914444946,
+      "loss": 1.6038,
+      "step": 672
+    },
+    {
+      "epoch": 0.20196563883262061,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001906160566350354,
+      "loss": 1.7038,
+      "step": 673
+    },
+    {
+      "epoch": 0.2022657363643184,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00019060099412562134,
+      "loss": 1.7343,
+      "step": 674
+    },
+    {
+      "epoch": 0.2025658338960162,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00019058593161620725,
+      "loss": 1.5586,
+      "step": 675
+    },
+    {
+      "epoch": 0.202865931427714,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001905708691067932,
+      "loss": 1.6777,
+      "step": 676
+    },
+    {
+      "epoch": 0.2031660289594118,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00019055580659737913,
+      "loss": 1.7327,
+      "step": 677
+    },
+    {
+      "epoch": 0.20346612649110962,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00019054074408796507,
+      "loss": 1.4847,
+      "step": 678
+    },
+    {
+      "epoch": 0.20376622402280742,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.000190525681578551,
+      "loss": 1.5468,
+      "step": 679
+    },
+    {
+      "epoch": 0.20406632155450521,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00019051061906913693,
+      "loss": 1.5556,
+      "step": 680
+    },
+    {
+      "epoch": 0.204366419086203,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00019049555655972287,
+      "loss": 1.527,
+      "step": 681
+    },
+    {
+      "epoch": 0.2046665166179008,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019048049405030878,
+      "loss": 1.8336,
+      "step": 682
+    },
+    {
+      "epoch": 0.20496661414959863,
+      "grad_norm": 3.453125,
+      "learning_rate": 0.00019046543154089472,
+      "loss": 1.6739,
+      "step": 683
+    },
+    {
+      "epoch": 0.20526671168129643,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00019045036903148066,
+      "loss": 1.4906,
+      "step": 684
+    },
+    {
+      "epoch": 0.20556680921299422,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00019043530652206657,
+      "loss": 1.6114,
+      "step": 685
+    },
+    {
+      "epoch": 0.20586690674469202,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00019042024401265251,
+      "loss": 1.5725,
+      "step": 686
+    },
+    {
+      "epoch": 0.20616700427638981,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00019040518150323845,
+      "loss": 1.5277,
+      "step": 687
+    },
+    {
+      "epoch": 0.20646710180808764,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019039011899382437,
+      "loss": 1.5455,
+      "step": 688
+    },
+    {
+      "epoch": 0.20676719933978543,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001903750564844103,
+      "loss": 1.3985,
+      "step": 689
+    },
+    {
+      "epoch": 0.20706729687148323,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00019035999397499625,
+      "loss": 1.6458,
+      "step": 690
+    },
+    {
+      "epoch": 0.20736739440318103,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019034493146558216,
+      "loss": 1.553,
+      "step": 691
+    },
+    {
+      "epoch": 0.20766749193487882,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001903298689561681,
+      "loss": 1.6564,
+      "step": 692
+    },
+    {
+      "epoch": 0.20796758946657665,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00019031480644675404,
+      "loss": 1.7099,
+      "step": 693
+    },
+    {
+      "epoch": 0.20826768699827444,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00019029974393733998,
+      "loss": 1.6427,
+      "step": 694
+    },
+    {
+      "epoch": 0.20856778452997224,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001902846814279259,
+      "loss": 1.6304,
+      "step": 695
+    },
+    {
+      "epoch": 0.20886788206167003,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00019026961891851184,
+      "loss": 1.4267,
+      "step": 696
+    },
+    {
+      "epoch": 0.20916797959336783,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019025455640909778,
+      "loss": 1.4554,
+      "step": 697
+    },
+    {
+      "epoch": 0.20946807712506565,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001902394938996837,
+      "loss": 1.4931,
+      "step": 698
+    },
+    {
+      "epoch": 0.20976817465676345,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00019022443139026963,
+      "loss": 1.5458,
+      "step": 699
+    },
+    {
+      "epoch": 0.21006827218846125,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00019020936888085557,
+      "loss": 1.9542,
+      "step": 700
+    },
+    {
+      "epoch": 0.21036836972015904,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00019019430637144148,
+      "loss": 1.564,
+      "step": 701
+    },
+    {
+      "epoch": 0.21066846725185687,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00019017924386202742,
+      "loss": 1.7874,
+      "step": 702
+    },
+    {
+      "epoch": 0.21096856478355466,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00019016418135261336,
+      "loss": 1.5641,
+      "step": 703
+    },
+    {
+      "epoch": 0.21126866231525246,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019014911884319928,
+      "loss": 1.5493,
+      "step": 704
+    },
+    {
+      "epoch": 0.21156875984695026,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00019013405633378522,
+      "loss": 1.5492,
+      "step": 705
+    },
+    {
+      "epoch": 0.21186885737864805,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00019011899382437116,
+      "loss": 1.6935,
+      "step": 706
+    },
+    {
+      "epoch": 0.21216895491034587,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00019010393131495707,
+      "loss": 1.4549,
+      "step": 707
+    },
+    {
+      "epoch": 0.21246905244204367,
+      "grad_norm": 8.5,
+      "learning_rate": 0.000190088868805543,
+      "loss": 1.4584,
+      "step": 708
+    },
+    {
+      "epoch": 0.21276914997374147,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00019007380629612895,
+      "loss": 1.5422,
+      "step": 709
+    },
+    {
+      "epoch": 0.21306924750543926,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00019005874378671486,
+      "loss": 1.6308,
+      "step": 710
+    },
+    {
+      "epoch": 0.21336934503713706,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001900436812773008,
+      "loss": 1.583,
+      "step": 711
+    },
+    {
+      "epoch": 0.21366944256883488,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00019002861876788675,
+      "loss": 1.924,
+      "step": 712
+    },
+    {
+      "epoch": 0.21396954010053268,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019001355625847269,
+      "loss": 1.4461,
+      "step": 713
+    },
+    {
+      "epoch": 0.21426963763223048,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001899984937490586,
+      "loss": 1.5583,
+      "step": 714
+    },
+    {
+      "epoch": 0.21456973516392827,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00018998343123964454,
+      "loss": 1.8522,
+      "step": 715
+    },
+    {
+      "epoch": 0.21486983269562607,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00018996836873023048,
+      "loss": 1.6595,
+      "step": 716
+    },
+    {
+      "epoch": 0.2151699302273239,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001899533062208164,
+      "loss": 1.5361,
+      "step": 717
+    },
+    {
+      "epoch": 0.2154700277590217,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001899382437114023,
+      "loss": 1.5324,
+      "step": 718
+    },
+    {
+      "epoch": 0.21577012529071948,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00018992318120198827,
+      "loss": 1.4054,
+      "step": 719
+    },
+    {
+      "epoch": 0.21607022282241728,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001899081186925742,
+      "loss": 1.6949,
+      "step": 720
+    },
+    {
+      "epoch": 0.21637032035411508,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00018989305618316013,
+      "loss": 1.5098,
+      "step": 721
+    },
+    {
+      "epoch": 0.2166704178858129,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018987799367374607,
+      "loss": 1.3886,
+      "step": 722
+    },
+    {
+      "epoch": 0.2169705154175107,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018986293116433198,
+      "loss": 1.4431,
+      "step": 723
+    },
+    {
+      "epoch": 0.2172706129492085,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00018984786865491792,
+      "loss": 1.6994,
+      "step": 724
+    },
+    {
+      "epoch": 0.2175707104809063,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00018983280614550386,
+      "loss": 1.4272,
+      "step": 725
+    },
+    {
+      "epoch": 0.21787080801260408,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00018981774363608977,
+      "loss": 1.5263,
+      "step": 726
+    },
+    {
+      "epoch": 0.2181709055443019,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018980268112667571,
+      "loss": 1.6176,
+      "step": 727
+    },
+    {
+      "epoch": 0.2184710030759997,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00018978761861726165,
+      "loss": 1.8208,
+      "step": 728
+    },
+    {
+      "epoch": 0.2187711006076975,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00018977255610784757,
+      "loss": 1.7318,
+      "step": 729
+    },
+    {
+      "epoch": 0.2190711981393953,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001897574935984335,
+      "loss": 1.5174,
+      "step": 730
+    },
+    {
+      "epoch": 0.2193712956710931,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00018974243108901942,
+      "loss": 1.7697,
+      "step": 731
+    },
+    {
+      "epoch": 0.21967139320279092,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001897273685796054,
+      "loss": 1.4862,
+      "step": 732
+    },
+    {
+      "epoch": 0.2199714907344887,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001897123060701913,
+      "loss": 1.5477,
+      "step": 733
+    },
+    {
+      "epoch": 0.2202715882661865,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00018969724356077722,
+      "loss": 1.7161,
+      "step": 734
+    },
+    {
+      "epoch": 0.2205716857978843,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018968218105136318,
+      "loss": 1.3867,
+      "step": 735
+    },
+    {
+      "epoch": 0.22087178332958213,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001896671185419491,
+      "loss": 1.7228,
+      "step": 736
+    },
+    {
+      "epoch": 0.22117188086127992,
+      "grad_norm": 1.5,
+      "learning_rate": 0.000189652056032535,
+      "loss": 1.48,
+      "step": 737
+    },
+    {
+      "epoch": 0.22147197839297772,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018963699352312098,
+      "loss": 1.4978,
+      "step": 738
+    },
+    {
+      "epoch": 0.22177207592467552,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001896219310137069,
+      "loss": 1.8609,
+      "step": 739
+    },
+    {
+      "epoch": 0.2220721734563733,
+      "grad_norm": 2.390625,
+      "learning_rate": 0.00018960686850429283,
+      "loss": 1.5324,
+      "step": 740
+    },
+    {
+      "epoch": 0.22237227098807114,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00018959180599487877,
+      "loss": 1.5691,
+      "step": 741
+    },
+    {
+      "epoch": 0.22267236851976893,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00018957674348546468,
+      "loss": 1.7052,
+      "step": 742
+    },
+    {
+      "epoch": 0.22297246605146673,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00018956168097605062,
+      "loss": 1.5346,
+      "step": 743
+    },
+    {
+      "epoch": 0.22327256358316452,
+      "grad_norm": 9.9375,
+      "learning_rate": 0.00018954661846663654,
+      "loss": 1.6401,
+      "step": 744
+    },
+    {
+      "epoch": 0.22357266111486232,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018953155595722248,
+      "loss": 1.5302,
+      "step": 745
+    },
+    {
+      "epoch": 0.22387275864656014,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00018951649344780842,
+      "loss": 1.8174,
+      "step": 746
+    },
+    {
+      "epoch": 0.22417285617825794,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00018950143093839433,
+      "loss": 1.6151,
+      "step": 747
+    },
+    {
+      "epoch": 0.22447295370995574,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001894863684289803,
+      "loss": 1.4346,
+      "step": 748
+    },
+    {
+      "epoch": 0.22477305124165353,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001894713059195662,
+      "loss": 1.6974,
+      "step": 749
+    },
+    {
+      "epoch": 0.22507314877335133,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018945624341015212,
+      "loss": 1.6096,
+      "step": 750
+    },
+    {
+      "epoch": 0.22537324630504915,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001894411809007381,
+      "loss": 1.5651,
+      "step": 751
+    },
+    {
+      "epoch": 0.22567334383674695,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.000189426118391324,
+      "loss": 1.5615,
+      "step": 752
+    },
+    {
+      "epoch": 0.22597344136844474,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018941105588190992,
+      "loss": 1.6279,
+      "step": 753
+    },
+    {
+      "epoch": 0.22627353890014254,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018939599337249586,
+      "loss": 1.5484,
+      "step": 754
+    },
+    {
+      "epoch": 0.22657363643184034,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001893809308630818,
+      "loss": 1.576,
+      "step": 755
+    },
+    {
+      "epoch": 0.22687373396353816,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001893658683536677,
+      "loss": 1.506,
+      "step": 756
+    },
+    {
+      "epoch": 0.22717383149523596,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00018935080584425365,
+      "loss": 1.3614,
+      "step": 757
+    },
+    {
+      "epoch": 0.22747392902693375,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001893357433348396,
+      "loss": 1.4959,
+      "step": 758
+    },
+    {
+      "epoch": 0.22777402655863155,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018932068082542553,
+      "loss": 1.8017,
+      "step": 759
+    },
+    {
+      "epoch": 0.22807412409032934,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00018930561831601145,
+      "loss": 1.6914,
+      "step": 760
+    },
+    {
+      "epoch": 0.22837422162202717,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001892905558065974,
+      "loss": 1.5474,
+      "step": 761
+    },
+    {
+      "epoch": 0.22867431915372496,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018927549329718333,
+      "loss": 1.6794,
+      "step": 762
+    },
+    {
+      "epoch": 0.22897441668542276,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018926043078776924,
+      "loss": 1.6124,
+      "step": 763
+    },
+    {
+      "epoch": 0.22927451421712056,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018924536827835518,
+      "loss": 1.6386,
+      "step": 764
+    },
+    {
+      "epoch": 0.22957461174881835,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018923030576894112,
+      "loss": 1.5832,
+      "step": 765
+    },
+    {
+      "epoch": 0.22987470928051618,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00018921524325952703,
+      "loss": 1.4725,
+      "step": 766
+    },
+    {
+      "epoch": 0.23017480681221397,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00018920018075011297,
+      "loss": 1.7471,
+      "step": 767
+    },
+    {
+      "epoch": 0.23047490434391177,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018918511824069891,
+      "loss": 1.5943,
+      "step": 768
+    },
+    {
+      "epoch": 0.23077500187560956,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00018917005573128483,
+      "loss": 1.3792,
+      "step": 769
+    },
+    {
+      "epoch": 0.2310750994073074,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00018915499322187077,
+      "loss": 1.5829,
+      "step": 770
+    },
+    {
+      "epoch": 0.23137519693900518,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001891399307124567,
+      "loss": 1.6752,
+      "step": 771
+    },
+    {
+      "epoch": 0.23167529447070298,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018912486820304262,
+      "loss": 1.5761,
+      "step": 772
+    },
+    {
+      "epoch": 0.23197539200240078,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00018910980569362856,
+      "loss": 1.5289,
+      "step": 773
+    },
+    {
+      "epoch": 0.23227548953409857,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001890947431842145,
+      "loss": 1.5257,
+      "step": 774
+    },
+    {
+      "epoch": 0.2325755870657964,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00018907968067480044,
+      "loss": 1.5859,
+      "step": 775
+    },
+    {
+      "epoch": 0.2328756845974942,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00018906461816538636,
+      "loss": 1.6038,
+      "step": 776
+    },
+    {
+      "epoch": 0.233175782129192,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001890495556559723,
+      "loss": 1.8739,
+      "step": 777
+    },
+    {
+      "epoch": 0.23347587966088978,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00018903449314655824,
+      "loss": 1.8358,
+      "step": 778
+    },
+    {
+      "epoch": 0.23377597719258758,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018901943063714415,
+      "loss": 1.6768,
+      "step": 779
+    },
+    {
+      "epoch": 0.2340760747242854,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001890043681277301,
+      "loss": 1.5545,
+      "step": 780
+    },
+    {
+      "epoch": 0.2343761722559832,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00018898930561831603,
+      "loss": 1.6533,
+      "step": 781
+    },
+    {
+      "epoch": 0.234676269787681,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018897424310890194,
+      "loss": 1.492,
+      "step": 782
+    },
+    {
+      "epoch": 0.2349763673193788,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00018895918059948788,
+      "loss": 1.6258,
+      "step": 783
+    },
+    {
+      "epoch": 0.2352764648510766,
+      "grad_norm": 7.3125,
+      "learning_rate": 0.00018894411809007382,
+      "loss": 1.5295,
+      "step": 784
+    },
+    {
+      "epoch": 0.2355765623827744,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00018892905558065974,
+      "loss": 1.7118,
+      "step": 785
+    },
+    {
+      "epoch": 0.2358766599144722,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00018891399307124568,
+      "loss": 1.4167,
+      "step": 786
+    },
+    {
+      "epoch": 0.23617675744617,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00018889893056183162,
+      "loss": 1.5052,
+      "step": 787
+    },
+    {
+      "epoch": 0.2364768549778678,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00018888386805241753,
+      "loss": 1.5729,
+      "step": 788
+    },
+    {
+      "epoch": 0.2367769525095656,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00018886880554300347,
+      "loss": 1.5417,
+      "step": 789
+    },
+    {
+      "epoch": 0.23707705004126342,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001888537430335894,
+      "loss": 1.7645,
+      "step": 790
+    },
+    {
+      "epoch": 0.23737714757296122,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018883868052417532,
+      "loss": 1.36,
+      "step": 791
+    },
+    {
+      "epoch": 0.237677245104659,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018882361801476127,
+      "loss": 1.5504,
+      "step": 792
+    },
+    {
+      "epoch": 0.2379773426363568,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001888085555053472,
+      "loss": 1.4623,
+      "step": 793
+    },
+    {
+      "epoch": 0.2382774401680546,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018879349299593315,
+      "loss": 1.4978,
+      "step": 794
+    },
+    {
+      "epoch": 0.23857753769975243,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00018877843048651906,
+      "loss": 1.5888,
+      "step": 795
+    },
+    {
+      "epoch": 0.23887763523145022,
+      "grad_norm": 1.875,
+      "learning_rate": 0.000188763367977105,
+      "loss": 1.6102,
+      "step": 796
+    },
+    {
+      "epoch": 0.23917773276314802,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00018874830546769094,
+      "loss": 1.5791,
+      "step": 797
+    },
+    {
+      "epoch": 0.23947783029484582,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018873324295827685,
+      "loss": 1.3414,
+      "step": 798
+    },
+    {
+      "epoch": 0.2397779278265436,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001887181804488628,
+      "loss": 1.4814,
+      "step": 799
+    },
+    {
+      "epoch": 0.24007802535824144,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018870311793944873,
+      "loss": 1.6525,
+      "step": 800
+    },
+    {
+      "epoch": 0.24037812288993923,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00018868805543003465,
+      "loss": 1.63,
+      "step": 801
+    },
+    {
+      "epoch": 0.24067822042163703,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001886729929206206,
+      "loss": 1.5961,
+      "step": 802
+    },
+    {
+      "epoch": 0.24097831795333483,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018865793041120653,
+      "loss": 1.4084,
+      "step": 803
+    },
+    {
+      "epoch": 0.24127841548503265,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018864286790179244,
+      "loss": 1.5355,
+      "step": 804
+    },
+    {
+      "epoch": 0.24157851301673045,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018862780539237838,
+      "loss": 1.7496,
+      "step": 805
+    },
+    {
+      "epoch": 0.24187861054842824,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018861274288296432,
+      "loss": 1.5766,
+      "step": 806
+    },
+    {
+      "epoch": 0.24217870808012604,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018859768037355023,
+      "loss": 1.6386,
+      "step": 807
+    },
+    {
+      "epoch": 0.24247880561182383,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018858261786413617,
+      "loss": 1.5728,
+      "step": 808
+    },
+    {
+      "epoch": 0.24277890314352166,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00018856755535472211,
+      "loss": 1.6603,
+      "step": 809
+    },
+    {
+      "epoch": 0.24307900067521945,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00018855249284530803,
+      "loss": 1.7977,
+      "step": 810
+    },
+    {
+      "epoch": 0.24337909820691725,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00018853743033589397,
+      "loss": 1.7713,
+      "step": 811
+    },
+    {
+      "epoch": 0.24367919573861505,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001885223678264799,
+      "loss": 1.5415,
+      "step": 812
+    },
+    {
+      "epoch": 0.24397929327031284,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018850730531706585,
+      "loss": 1.511,
+      "step": 813
+    },
+    {
+      "epoch": 0.24427939080201067,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018849224280765176,
+      "loss": 1.6479,
+      "step": 814
+    },
+    {
+      "epoch": 0.24457948833370846,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018847718029823768,
+      "loss": 1.3702,
+      "step": 815
+    },
+    {
+      "epoch": 0.24487958586540626,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00018846211778882364,
+      "loss": 1.8684,
+      "step": 816
+    },
+    {
+      "epoch": 0.24517968339710405,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018844705527940956,
+      "loss": 1.4243,
+      "step": 817
+    },
+    {
+      "epoch": 0.24547978092880185,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018843199276999547,
+      "loss": 1.3965,
+      "step": 818
+    },
+    {
+      "epoch": 0.24577987846049967,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018841693026058144,
+      "loss": 1.5183,
+      "step": 819
+    },
+    {
+      "epoch": 0.24607997599219747,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018840186775116735,
+      "loss": 1.4727,
+      "step": 820
+    },
+    {
+      "epoch": 0.24638007352389527,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001883868052417533,
+      "loss": 1.576,
+      "step": 821
+    },
+    {
+      "epoch": 0.24668017105559306,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00018837174273233923,
+      "loss": 1.86,
+      "step": 822
+    },
+    {
+      "epoch": 0.24698026858729086,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018835668022292514,
+      "loss": 1.6826,
+      "step": 823
+    },
+    {
+      "epoch": 0.24728036611898868,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018834161771351108,
+      "loss": 1.6552,
+      "step": 824
+    },
+    {
+      "epoch": 0.24758046365068648,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.000188326555204097,
+      "loss": 1.436,
+      "step": 825
+    },
+    {
+      "epoch": 0.24788056118238427,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00018831149269468294,
+      "loss": 1.5401,
+      "step": 826
+    },
+    {
+      "epoch": 0.24818065871408207,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00018829643018526888,
+      "loss": 1.6269,
+      "step": 827
+    },
+    {
+      "epoch": 0.24848075624577987,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001882813676758548,
+      "loss": 1.4165,
+      "step": 828
+    },
+    {
+      "epoch": 0.2487808537774777,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018826630516644076,
+      "loss": 1.5385,
+      "step": 829
+    },
+    {
+      "epoch": 0.24908095130917549,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018825124265702667,
+      "loss": 1.4564,
+      "step": 830
+    },
+    {
+      "epoch": 0.24938104884087328,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018823618014761258,
+      "loss": 1.4745,
+      "step": 831
+    },
+    {
+      "epoch": 0.24968114637257108,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018822111763819855,
+      "loss": 1.4751,
+      "step": 832
+    },
+    {
+      "epoch": 0.24998124390426887,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00018820605512878447,
+      "loss": 1.4609,
+      "step": 833
+    },
+    {
+      "epoch": 0.2502813414359667,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018819099261937038,
+      "loss": 1.6295,
+      "step": 834
+    },
+    {
+      "epoch": 0.2505814389676645,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018817593010995635,
+      "loss": 1.6266,
+      "step": 835
+    },
+    {
+      "epoch": 0.2508815364993623,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018816086760054226,
+      "loss": 1.5164,
+      "step": 836
+    },
+    {
+      "epoch": 0.2511816340310601,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00018814580509112817,
+      "loss": 1.6065,
+      "step": 837
+    },
+    {
+      "epoch": 0.2514817315627579,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001881307425817141,
+      "loss": 1.469,
+      "step": 838
+    },
+    {
+      "epoch": 0.2517818290944557,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00018811568007230005,
+      "loss": 1.5094,
+      "step": 839
+    },
+    {
+      "epoch": 0.2520819266261535,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.000188100617562886,
+      "loss": 1.4787,
+      "step": 840
+    },
+    {
+      "epoch": 0.2523820241578513,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.0001880855550534719,
+      "loss": 1.6291,
+      "step": 841
+    },
+    {
+      "epoch": 0.2526821216895491,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00018807049254405785,
+      "loss": 1.5997,
+      "step": 842
+    },
+    {
+      "epoch": 0.2529822192212469,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001880554300346438,
+      "loss": 1.8283,
+      "step": 843
+    },
+    {
+      "epoch": 0.2532823167529447,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001880403675252297,
+      "loss": 1.495,
+      "step": 844
+    },
+    {
+      "epoch": 0.2535824142846425,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00018802530501581564,
+      "loss": 1.6444,
+      "step": 845
+    },
+    {
+      "epoch": 0.2538825118163403,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00018801024250640158,
+      "loss": 1.7503,
+      "step": 846
+    },
+    {
+      "epoch": 0.2541826093480381,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001879951799969875,
+      "loss": 1.5271,
+      "step": 847
+    },
+    {
+      "epoch": 0.2544827068797359,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018798011748757346,
+      "loss": 1.549,
+      "step": 848
+    },
+    {
+      "epoch": 0.2547828044114337,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.00018796505497815937,
+      "loss": 1.5427,
+      "step": 849
+    },
+    {
+      "epoch": 0.2550829019431315,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001879499924687453,
+      "loss": 1.7483,
+      "step": 850
+    },
+    {
+      "epoch": 0.25538299947482934,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018793492995933123,
+      "loss": 1.5898,
+      "step": 851
+    },
+    {
+      "epoch": 0.25568309700652714,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018791986744991717,
+      "loss": 1.5416,
+      "step": 852
+    },
+    {
+      "epoch": 0.25598319453822493,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018790480494050308,
+      "loss": 1.396,
+      "step": 853
+    },
+    {
+      "epoch": 0.25628329206992273,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018788974243108902,
+      "loss": 1.3253,
+      "step": 854
+    },
+    {
+      "epoch": 0.2565833896016205,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00018787467992167496,
+      "loss": 1.5604,
+      "step": 855
+    },
+    {
+      "epoch": 0.2568834871333183,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001878596174122609,
+      "loss": 1.6375,
+      "step": 856
+    },
+    {
+      "epoch": 0.2571835846650161,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00018784455490284682,
+      "loss": 1.5907,
+      "step": 857
+    },
+    {
+      "epoch": 0.2574836821967139,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018782949239343276,
+      "loss": 1.3674,
+      "step": 858
+    },
+    {
+      "epoch": 0.2577837797284117,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001878144298840187,
+      "loss": 1.1815,
+      "step": 859
+    },
+    {
+      "epoch": 0.25808387726010956,
+      "grad_norm": 2.0,
+      "learning_rate": 0.0001877993673746046,
+      "loss": 1.6408,
+      "step": 860
+    },
+    {
+      "epoch": 0.25838397479180736,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00018778430486519055,
+      "loss": 1.6759,
+      "step": 861
+    },
+    {
+      "epoch": 0.25868407232350515,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001877692423557765,
+      "loss": 1.3956,
+      "step": 862
+    },
+    {
+      "epoch": 0.25898416985520295,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001877541798463624,
+      "loss": 1.5198,
+      "step": 863
+    },
+    {
+      "epoch": 0.25928426738690075,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00018773911733694834,
+      "loss": 1.6724,
+      "step": 864
+    },
+    {
+      "epoch": 0.25958436491859854,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018772405482753428,
+      "loss": 1.6487,
+      "step": 865
+    },
+    {
+      "epoch": 0.25988446245029634,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001877089923181202,
+      "loss": 1.5761,
+      "step": 866
+    },
+    {
+      "epoch": 0.26018455998199413,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018769392980870614,
+      "loss": 1.5766,
+      "step": 867
+    },
+    {
+      "epoch": 0.26048465751369193,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018767886729929208,
+      "loss": 1.5054,
+      "step": 868
+    },
+    {
+      "epoch": 0.2607847550453897,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.000187663804789878,
+      "loss": 1.492,
+      "step": 869
+    },
+    {
+      "epoch": 0.2610848525770876,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018764874228046393,
+      "loss": 1.6325,
+      "step": 870
+    },
+    {
+      "epoch": 0.2613849501087854,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00018763367977104987,
+      "loss": 1.6097,
+      "step": 871
+    },
+    {
+      "epoch": 0.26168504764048317,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018761861726163578,
+      "loss": 1.517,
+      "step": 872
+    },
+    {
+      "epoch": 0.26198514517218097,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00018760355475222173,
+      "loss": 1.6121,
+      "step": 873
+    },
+    {
+      "epoch": 0.26228524270387876,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018758849224280767,
+      "loss": 1.5043,
+      "step": 874
+    },
+    {
+      "epoch": 0.26258534023557656,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001875734297333936,
+      "loss": 1.6334,
+      "step": 875
+    },
+    {
+      "epoch": 0.26288543776727435,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00018755836722397952,
+      "loss": 1.5185,
+      "step": 876
+    },
+    {
+      "epoch": 0.26318553529897215,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018754330471456546,
+      "loss": 1.5225,
+      "step": 877
+    },
+    {
+      "epoch": 0.26348563283066995,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001875282422051514,
+      "loss": 1.615,
+      "step": 878
+    },
+    {
+      "epoch": 0.26378573036236774,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.0001875131796957373,
+      "loss": 1.5653,
+      "step": 879
+    },
+    {
+      "epoch": 0.2640858278940656,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018749811718632325,
+      "loss": 1.5101,
+      "step": 880
+    },
+    {
+      "epoch": 0.2643859254257634,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001874830546769092,
+      "loss": 1.4432,
+      "step": 881
+    },
+    {
+      "epoch": 0.2646860229574612,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001874679921674951,
+      "loss": 1.3807,
+      "step": 882
+    },
+    {
+      "epoch": 0.264986120489159,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018745292965808105,
+      "loss": 1.5568,
+      "step": 883
+    },
+    {
+      "epoch": 0.2652862180208568,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.000187437867148667,
+      "loss": 1.5451,
+      "step": 884
+    },
+    {
+      "epoch": 0.2655863155525546,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001874228046392529,
+      "loss": 1.5673,
+      "step": 885
+    },
+    {
+      "epoch": 0.26588641308425237,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018740774212983884,
+      "loss": 1.596,
+      "step": 886
+    },
+    {
+      "epoch": 0.26618651061595017,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00018739267962042478,
+      "loss": 1.741,
+      "step": 887
+    },
+    {
+      "epoch": 0.26648660814764796,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.0001873776171110107,
+      "loss": 1.7496,
+      "step": 888
+    },
+    {
+      "epoch": 0.2667867056793458,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018736255460159663,
+      "loss": 1.5862,
+      "step": 889
+    },
+    {
+      "epoch": 0.2670868032110436,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018734749209218257,
+      "loss": 1.5307,
+      "step": 890
+    },
+    {
+      "epoch": 0.2673869007427414,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001873324295827685,
+      "loss": 1.6871,
+      "step": 891
+    },
+    {
+      "epoch": 0.2676869982744392,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018731736707335443,
+      "loss": 1.5818,
+      "step": 892
+    },
+    {
+      "epoch": 0.267987095806137,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00018730230456394037,
+      "loss": 1.6683,
+      "step": 893
+    },
+    {
+      "epoch": 0.2682871933378348,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001872872420545263,
+      "loss": 1.7646,
+      "step": 894
+    },
+    {
+      "epoch": 0.2685872908695326,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018727217954511222,
+      "loss": 1.6728,
+      "step": 895
+    },
+    {
+      "epoch": 0.2688873884012304,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018725711703569814,
+      "loss": 1.4545,
+      "step": 896
+    },
+    {
+      "epoch": 0.2691874859329282,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001872420545262841,
+      "loss": 1.3816,
+      "step": 897
+    },
+    {
+      "epoch": 0.269487583464626,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018722699201687002,
+      "loss": 1.685,
+      "step": 898
+    },
+    {
+      "epoch": 0.26978768099632383,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00018721192950745593,
+      "loss": 1.5815,
+      "step": 899
+    },
+    {
+      "epoch": 0.2700877785280216,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001871968669980419,
+      "loss": 1.6894,
+      "step": 900
+    },
+    {
+      "epoch": 0.2703878760597194,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.0001871818044886278,
+      "loss": 1.6515,
+      "step": 901
+    },
+    {
+      "epoch": 0.2706879735914172,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018716674197921375,
+      "loss": 1.4751,
+      "step": 902
+    },
+    {
+      "epoch": 0.270988071123115,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001871516794697997,
+      "loss": 1.5035,
+      "step": 903
+    },
+    {
+      "epoch": 0.2712881686548128,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001871366169603856,
+      "loss": 1.5951,
+      "step": 904
+    },
+    {
+      "epoch": 0.2715882661865106,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018712155445097154,
+      "loss": 1.3937,
+      "step": 905
+    },
+    {
+      "epoch": 0.2718883637182084,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018710649194155748,
+      "loss": 1.6583,
+      "step": 906
+    },
+    {
+      "epoch": 0.2721884612499062,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001870914294321434,
+      "loss": 1.472,
+      "step": 907
+    },
+    {
+      "epoch": 0.272488558781604,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018707636692272934,
+      "loss": 1.2891,
+      "step": 908
+    },
+    {
+      "epoch": 0.27278865631330185,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018706130441331525,
+      "loss": 1.5525,
+      "step": 909
+    },
+    {
+      "epoch": 0.27308875384499964,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018704624190390122,
+      "loss": 1.5346,
+      "step": 910
+    },
+    {
+      "epoch": 0.27338885137669744,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018703117939448713,
+      "loss": 1.4297,
+      "step": 911
+    },
+    {
+      "epoch": 0.27368894890839524,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00018701611688507304,
+      "loss": 1.5749,
+      "step": 912
+    },
+    {
+      "epoch": 0.27398904644009303,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.000187001054375659,
+      "loss": 1.6282,
+      "step": 913
+    },
+    {
+      "epoch": 0.2742891439717908,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018698599186624493,
+      "loss": 1.6072,
+      "step": 914
+    },
+    {
+      "epoch": 0.2745892415034886,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018697092935683084,
+      "loss": 1.6163,
+      "step": 915
+    },
+    {
+      "epoch": 0.2748893390351864,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.0001869558668474168,
+      "loss": 1.7244,
+      "step": 916
+    },
+    {
+      "epoch": 0.2751894365668842,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00018694080433800272,
+      "loss": 1.5564,
+      "step": 917
+    },
+    {
+      "epoch": 0.275489534098582,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018692574182858863,
+      "loss": 1.7548,
+      "step": 918
+    },
+    {
+      "epoch": 0.27578963163027986,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.0001869106793191746,
+      "loss": 1.5589,
+      "step": 919
+    },
+    {
+      "epoch": 0.27608972916197766,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001868956168097605,
+      "loss": 1.5608,
+      "step": 920
+    },
+    {
+      "epoch": 0.27638982669367546,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018688055430034645,
+      "loss": 1.7432,
+      "step": 921
+    },
+    {
+      "epoch": 0.27668992422537325,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018686549179093237,
+      "loss": 1.6445,
+      "step": 922
+    },
+    {
+      "epoch": 0.27699002175707105,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001868504292815183,
+      "loss": 1.3911,
+      "step": 923
+    },
+    {
+      "epoch": 0.27729011928876884,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00018683536677210425,
+      "loss": 1.3797,
+      "step": 924
+    },
+    {
+      "epoch": 0.27759021682046664,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018682030426269016,
+      "loss": 1.4835,
+      "step": 925
+    },
+    {
+      "epoch": 0.27789031435216444,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.0001868052417532761,
+      "loss": 1.6195,
+      "step": 926
+    },
+    {
+      "epoch": 0.27819041188386223,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00018679017924386204,
+      "loss": 1.4923,
+      "step": 927
+    },
+    {
+      "epoch": 0.2784905094155601,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018677511673444795,
+      "loss": 1.4913,
+      "step": 928
+    },
+    {
+      "epoch": 0.2787906069472579,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00018676005422503392,
+      "loss": 1.6255,
+      "step": 929
+    },
+    {
+      "epoch": 0.2790907044789557,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00018674499171561983,
+      "loss": 1.4437,
+      "step": 930
+    },
+    {
+      "epoch": 0.27939080201065347,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00018672992920620575,
+      "loss": 1.6764,
+      "step": 931
+    },
+    {
+      "epoch": 0.27969089954235127,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001867148666967917,
+      "loss": 1.3859,
+      "step": 932
+    },
+    {
+      "epoch": 0.27999099707404906,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018669980418737763,
+      "loss": 1.4084,
+      "step": 933
+    },
+    {
+      "epoch": 0.28029109460574686,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018668474167796354,
+      "loss": 1.5567,
+      "step": 934
+    },
+    {
+      "epoch": 0.28059119213744466,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018666967916854948,
+      "loss": 1.3425,
+      "step": 935
+    },
+    {
+      "epoch": 0.28089128966914245,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018665461665913542,
+      "loss": 1.6332,
+      "step": 936
+    },
+    {
+      "epoch": 0.28119138720084025,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00018663955414972136,
+      "loss": 1.3728,
+      "step": 937
+    },
+    {
+      "epoch": 0.2814914847325381,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00018662449164030728,
+      "loss": 1.6139,
+      "step": 938
+    },
+    {
+      "epoch": 0.2817915822642359,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018660942913089322,
+      "loss": 1.5678,
+      "step": 939
+    },
+    {
+      "epoch": 0.2820916797959337,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018659436662147916,
+      "loss": 1.6103,
+      "step": 940
+    },
+    {
+      "epoch": 0.2823917773276315,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018657930411206507,
+      "loss": 1.5688,
+      "step": 941
+    },
+    {
+      "epoch": 0.2826918748593293,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.000186564241602651,
+      "loss": 1.5517,
+      "step": 942
+    },
+    {
+      "epoch": 0.2829919723910271,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018654917909323695,
+      "loss": 1.4805,
+      "step": 943
+    },
+    {
+      "epoch": 0.2832920699227249,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018653411658382286,
+      "loss": 1.5486,
+      "step": 944
+    },
+    {
+      "epoch": 0.2835921674544227,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001865190540744088,
+      "loss": 1.7684,
+      "step": 945
+    },
+    {
+      "epoch": 0.28389226498612047,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018650399156499474,
+      "loss": 1.398,
+      "step": 946
+    },
+    {
+      "epoch": 0.28419236251781826,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018648892905558066,
+      "loss": 1.5348,
+      "step": 947
+    },
+    {
+      "epoch": 0.2844924600495161,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.0001864738665461666,
+      "loss": 1.7038,
+      "step": 948
+    },
+    {
+      "epoch": 0.2847925575812139,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018645880403675254,
+      "loss": 1.5491,
+      "step": 949
+    },
+    {
+      "epoch": 0.2850926551129117,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018644374152733845,
+      "loss": 1.6717,
+      "step": 950
+    },
+    {
+      "epoch": 0.2853927526446095,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001864286790179244,
+      "loss": 1.6765,
+      "step": 951
+    },
+    {
+      "epoch": 0.2856928501763073,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018641361650851033,
+      "loss": 1.5512,
+      "step": 952
+    },
+    {
+      "epoch": 0.2859929477080051,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018639855399909624,
+      "loss": 1.6231,
+      "step": 953
+    },
+    {
+      "epoch": 0.2862930452397029,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018638349148968219,
+      "loss": 1.5572,
+      "step": 954
+    },
+    {
+      "epoch": 0.2865931427714007,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00018636842898026813,
+      "loss": 1.6453,
+      "step": 955
+    },
+    {
+      "epoch": 0.2868932403030985,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018635336647085407,
+      "loss": 1.5303,
+      "step": 956
+    },
+    {
+      "epoch": 0.28719333783479634,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018633830396143998,
+      "loss": 1.6022,
+      "step": 957
+    },
+    {
+      "epoch": 0.28749343536649413,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018632324145202592,
+      "loss": 1.5709,
+      "step": 958
+    },
+    {
+      "epoch": 0.28779353289819193,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00018630817894261186,
+      "loss": 1.8368,
+      "step": 959
+    },
+    {
+      "epoch": 0.2880936304298897,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00018629311643319777,
+      "loss": 1.6817,
+      "step": 960
+    },
+    {
+      "epoch": 0.2883937279615875,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001862780539237837,
+      "loss": 1.4369,
+      "step": 961
+    },
+    {
+      "epoch": 0.2886938254932853,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018626299141436965,
+      "loss": 1.4956,
+      "step": 962
+    },
+    {
+      "epoch": 0.2889939230249831,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018624792890495557,
+      "loss": 1.477,
+      "step": 963
+    },
+    {
+      "epoch": 0.2892940205566809,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.0001862328663955415,
+      "loss": 1.5163,
+      "step": 964
+    },
+    {
+      "epoch": 0.2895941180883787,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018621780388612745,
+      "loss": 1.4776,
+      "step": 965
+    },
+    {
+      "epoch": 0.2898942156200765,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018620274137671336,
+      "loss": 1.6991,
+      "step": 966
+    },
+    {
+      "epoch": 0.29019431315177435,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001861876788672993,
+      "loss": 1.5699,
+      "step": 967
+    },
+    {
+      "epoch": 0.29049441068347215,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00018617261635788524,
+      "loss": 1.7586,
+      "step": 968
+    },
+    {
+      "epoch": 0.29079450821516994,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018615755384847115,
+      "loss": 1.5001,
+      "step": 969
+    },
+    {
+      "epoch": 0.29109460574686774,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001861424913390571,
+      "loss": 1.4838,
+      "step": 970
+    },
+    {
+      "epoch": 0.29139470327856554,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018612742882964303,
+      "loss": 1.5378,
+      "step": 971
+    },
+    {
+      "epoch": 0.29169480081026333,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018611236632022895,
+      "loss": 1.3811,
+      "step": 972
+    },
+    {
+      "epoch": 0.29199489834196113,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001860973038108149,
+      "loss": 1.4847,
+      "step": 973
+    },
+    {
+      "epoch": 0.2922949958736589,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00018608224130140083,
+      "loss": 1.3861,
+      "step": 974
+    },
+    {
+      "epoch": 0.2925950934053567,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018606717879198677,
+      "loss": 1.5546,
+      "step": 975
+    },
+    {
+      "epoch": 0.2928951909370545,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018605211628257268,
+      "loss": 1.4138,
+      "step": 976
+    },
+    {
+      "epoch": 0.29319528846875237,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00018603705377315862,
+      "loss": 1.484,
+      "step": 977
+    },
+    {
+      "epoch": 0.29349538600045016,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018602199126374456,
+      "loss": 1.4326,
+      "step": 978
+    },
+    {
+      "epoch": 0.29379548353214796,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00018600692875433048,
+      "loss": 1.5529,
+      "step": 979
+    },
+    {
+      "epoch": 0.29409558106384576,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001859918662449164,
+      "loss": 1.4406,
+      "step": 980
+    },
+    {
+      "epoch": 0.29439567859554355,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018597680373550236,
+      "loss": 1.5325,
+      "step": 981
+    },
+    {
+      "epoch": 0.29469577612724135,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018596174122608827,
+      "loss": 1.3712,
+      "step": 982
+    },
+    {
+      "epoch": 0.29499587365893915,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.0001859466787166742,
+      "loss": 1.4665,
+      "step": 983
+    },
+    {
+      "epoch": 0.29529597119063694,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00018593161620726015,
+      "loss": 1.727,
+      "step": 984
+    },
+    {
+      "epoch": 0.29559606872233474,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018591655369784606,
+      "loss": 1.5416,
+      "step": 985
+    },
+    {
+      "epoch": 0.29589616625403253,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.000185901491188432,
+      "loss": 1.4836,
+      "step": 986
+    },
+    {
+      "epoch": 0.2961962637857304,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018588642867901794,
+      "loss": 1.5397,
+      "step": 987
+    },
+    {
+      "epoch": 0.2964963613174282,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018587136616960386,
+      "loss": 1.4447,
+      "step": 988
+    },
+    {
+      "epoch": 0.296796458849126,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001858563036601898,
+      "loss": 1.4848,
+      "step": 989
+    },
+    {
+      "epoch": 0.2970965563808238,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.0001858412411507757,
+      "loss": 1.4899,
+      "step": 990
+    },
+    {
+      "epoch": 0.29739665391252157,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018582617864136168,
+      "loss": 1.4665,
+      "step": 991
+    },
+    {
+      "epoch": 0.29769675144421937,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001858111161319476,
+      "loss": 1.4594,
+      "step": 992
+    },
+    {
+      "epoch": 0.29799684897591716,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001857960536225335,
+      "loss": 1.4327,
+      "step": 993
+    },
+    {
+      "epoch": 0.29829694650761496,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00018578099111311947,
+      "loss": 1.6085,
+      "step": 994
+    },
+    {
+      "epoch": 0.29859704403931275,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00018576592860370539,
+      "loss": 1.5007,
+      "step": 995
+    },
+    {
+      "epoch": 0.2988971415710106,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001857508660942913,
+      "loss": 1.4262,
+      "step": 996
+    },
+    {
+      "epoch": 0.2991972391027084,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018573580358487727,
+      "loss": 1.6092,
+      "step": 997
+    },
+    {
+      "epoch": 0.2994973366344062,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00018572074107546318,
+      "loss": 1.7704,
+      "step": 998
+    },
+    {
+      "epoch": 0.299797434166104,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001857056785660491,
+      "loss": 1.4038,
+      "step": 999
+    },
+    {
+      "epoch": 0.3000975316978018,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00018569061605663506,
+      "loss": 1.7797,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3003976292294996,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00018567555354722097,
+      "loss": 1.5334,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3006977267611974,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001856604910378069,
+      "loss": 1.5897,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3009978242928952,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018564542852839283,
+      "loss": 1.4099,
+      "step": 1003
+    },
+    {
+      "epoch": 0.301297921824593,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00018563036601897877,
+      "loss": 1.5034,
+      "step": 1004
+    },
+    {
+      "epoch": 0.30159801935629077,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001856153035095647,
+      "loss": 1.6193,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3018981168879886,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018560024100015062,
+      "loss": 1.591,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3021982144196864,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00018558517849073656,
+      "loss": 1.6193,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3024983119513842,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001855701159813225,
+      "loss": 1.5885,
+      "step": 1008
+    },
+    {
+      "epoch": 0.302798409483082,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00018555505347190841,
+      "loss": 1.6505,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3030985070147798,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018553999096249438,
+      "loss": 1.5,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3033986045464776,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001855249284530803,
+      "loss": 1.5536,
+      "step": 1011
+    },
+    {
+      "epoch": 0.3036987020781754,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.0001855098659436662,
+      "loss": 1.6599,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3039987996098732,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00018549480343425218,
+      "loss": 1.5482,
+      "step": 1013
+    },
+    {
+      "epoch": 0.304298897141571,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001854797409248381,
+      "loss": 1.4333,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3045989946732688,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.000185464678415424,
+      "loss": 1.4825,
+      "step": 1015
+    },
+    {
+      "epoch": 0.30489909220496664,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00018544961590600994,
+      "loss": 1.5121,
+      "step": 1016
+    },
+    {
+      "epoch": 0.30519918973666443,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018543455339659588,
+      "loss": 1.7644,
+      "step": 1017
+    },
+    {
+      "epoch": 0.30549928726836223,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018541949088718182,
+      "loss": 1.4108,
+      "step": 1018
+    },
+    {
+      "epoch": 0.30579938480006,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018540442837776774,
+      "loss": 1.5991,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3060994823317578,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018538936586835368,
+      "loss": 1.5133,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3063995798634556,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018537430335893962,
+      "loss": 1.3556,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3066996773951534,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018535924084952553,
+      "loss": 1.4659,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3069997749268512,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00018534417834011147,
+      "loss": 1.3569,
+      "step": 1023
+    },
+    {
+      "epoch": 0.307299872458549,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001853291158306974,
+      "loss": 1.493,
+      "step": 1024
+    },
+    {
+      "epoch": 0.30759996999024686,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018531405332128332,
+      "loss": 1.4151,
+      "step": 1025
+    },
+    {
+      "epoch": 0.30790006752194465,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018529899081186926,
+      "loss": 1.3797,
+      "step": 1026
+    },
+    {
+      "epoch": 0.30820016505364245,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001852839283024552,
+      "loss": 1.7316,
+      "step": 1027
+    },
+    {
+      "epoch": 0.30850026258534025,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018526886579304112,
+      "loss": 1.4219,
+      "step": 1028
+    },
+    {
+      "epoch": 0.30880036011703804,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018525380328362706,
+      "loss": 1.6769,
+      "step": 1029
+    },
+    {
+      "epoch": 0.30910045764873584,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.000185238740774213,
+      "loss": 1.4658,
+      "step": 1030
+    },
+    {
+      "epoch": 0.30940055518043363,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001852236782647989,
+      "loss": 1.5224,
+      "step": 1031
+    },
+    {
+      "epoch": 0.30970065271213143,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00018520861575538485,
+      "loss": 1.5557,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3100007502438292,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001851935532459708,
+      "loss": 1.5179,
+      "step": 1033
+    },
+    {
+      "epoch": 0.310300847775527,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001851784907365567,
+      "loss": 1.4555,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3106009453072249,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018516342822714265,
+      "loss": 1.2848,
+      "step": 1035
+    },
+    {
+      "epoch": 0.31090104283892267,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00018514836571772859,
+      "loss": 1.6182,
+      "step": 1036
+    },
+    {
+      "epoch": 0.31120114037062047,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018513330320831453,
+      "loss": 1.4735,
+      "step": 1037
+    },
+    {
+      "epoch": 0.31150123790231826,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00018511824069890044,
+      "loss": 1.3831,
+      "step": 1038
+    },
+    {
+      "epoch": 0.31180133543401606,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018510317818948638,
+      "loss": 1.4181,
+      "step": 1039
+    },
+    {
+      "epoch": 0.31210143296571385,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018508811568007232,
+      "loss": 1.6536,
+      "step": 1040
+    },
+    {
+      "epoch": 0.31240153049741165,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018507305317065823,
+      "loss": 1.5512,
+      "step": 1041
+    },
+    {
+      "epoch": 0.31270162802910945,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018505799066124417,
+      "loss": 1.444,
+      "step": 1042
+    },
+    {
+      "epoch": 0.31300172556080724,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018504292815183011,
+      "loss": 1.5243,
+      "step": 1043
+    },
+    {
+      "epoch": 0.31330182309250504,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018502786564241603,
+      "loss": 1.4794,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3136019206242029,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00018501280313300197,
+      "loss": 1.5701,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3139020181559007,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001849977406235879,
+      "loss": 1.7243,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3142021156875985,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018498267811417382,
+      "loss": 1.5668,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3145022132192963,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018496761560475976,
+      "loss": 1.5988,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3148023107509941,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001849525530953457,
+      "loss": 1.5717,
+      "step": 1049
+    },
+    {
+      "epoch": 0.31510240828269187,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018493749058593161,
+      "loss": 1.5149,
+      "step": 1050
+    },
+    {
+      "epoch": 0.31540250581438967,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00018492242807651755,
+      "loss": 1.566,
+      "step": 1051
+    },
+    {
+      "epoch": 0.31570260334608746,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001849073655671035,
+      "loss": 1.5474,
+      "step": 1052
+    },
+    {
+      "epoch": 0.31600270087778526,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001848923030576894,
+      "loss": 1.7225,
+      "step": 1053
+    },
+    {
+      "epoch": 0.31630279840948305,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018487724054827535,
+      "loss": 1.485,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3166028959411809,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001848621780388613,
+      "loss": 1.6119,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3169029934728787,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018484711552944723,
+      "loss": 1.3722,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3172030910045765,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018483205302003314,
+      "loss": 1.3462,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3175031885362743,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018481699051061908,
+      "loss": 1.4333,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3178032860679721,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018480192800120502,
+      "loss": 1.3674,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3181033835996699,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018478686549179094,
+      "loss": 1.4724,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3184034811313677,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00018477180298237685,
+      "loss": 1.6675,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3187035786630655,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00018475674047296282,
+      "loss": 1.489,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3190036761947633,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018474167796354873,
+      "loss": 1.5348,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3193037737264611,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018472661545413467,
+      "loss": 1.4522,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3196038712581589,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001847115529447206,
+      "loss": 1.4841,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3199039687898567,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018469649043530652,
+      "loss": 1.6342,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3202040663215545,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018468142792589246,
+      "loss": 1.4871,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3205041638532523,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001846663654164784,
+      "loss": 1.3574,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3208042613849501,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018465130290706432,
+      "loss": 1.3815,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3211043589166479,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00018463624039765026,
+      "loss": 1.6563,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3214044564483457,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001846211778882362,
+      "loss": 1.5642,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3217045539800435,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018460611537882214,
+      "loss": 1.389,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3220046515117413,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018459105286940805,
+      "loss": 1.4774,
+      "step": 1073
+    },
+    {
+      "epoch": 0.32230474904343914,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018457599035999396,
+      "loss": 1.6185,
+      "step": 1074
+    },
+    {
+      "epoch": 0.32260484657513694,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00018456092785057993,
+      "loss": 1.5243,
+      "step": 1075
+    },
+    {
+      "epoch": 0.32290494410683473,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00018454586534116585,
+      "loss": 1.4958,
+      "step": 1076
+    },
+    {
+      "epoch": 0.32320504163853253,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018453080283175176,
+      "loss": 1.487,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3235051391702303,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018451574032233773,
+      "loss": 1.6165,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3238052367019281,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018450067781292364,
+      "loss": 1.5625,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3241053342336259,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00018448561530350955,
+      "loss": 1.5259,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3244054317653237,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018447055279409552,
+      "loss": 1.4858,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3247055292970215,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018445549028468143,
+      "loss": 1.4443,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3250056268287193,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018444042777526737,
+      "loss": 1.4689,
+      "step": 1083
+    },
+    {
+      "epoch": 0.32530572436041716,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018442536526585331,
+      "loss": 1.5442,
+      "step": 1084
+    },
+    {
+      "epoch": 0.32560582189211495,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018441030275643923,
+      "loss": 1.5635,
+      "step": 1085
+    },
+    {
+      "epoch": 0.32590591942381275,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018439524024702517,
+      "loss": 1.4781,
+      "step": 1086
+    },
+    {
+      "epoch": 0.32620601695551055,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018438017773761108,
+      "loss": 1.4127,
+      "step": 1087
+    },
+    {
+      "epoch": 0.32650611448720834,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018436511522819702,
+      "loss": 1.5467,
+      "step": 1088
+    },
+    {
+      "epoch": 0.32680621201890614,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018435005271878296,
+      "loss": 1.6259,
+      "step": 1089
+    },
+    {
+      "epoch": 0.32710630955060394,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018433499020936887,
+      "loss": 1.4853,
+      "step": 1090
+    },
+    {
+      "epoch": 0.32740640708230173,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018431992769995484,
+      "loss": 1.6549,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3277065046139995,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018430486519054075,
+      "loss": 1.3987,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3280066021456974,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018428980268112667,
+      "loss": 1.429,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3283066996773952,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018427474017171264,
+      "loss": 1.6325,
+      "step": 1094
+    },
+    {
+      "epoch": 0.32860679720909297,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018425967766229855,
+      "loss": 1.3559,
+      "step": 1095
+    },
+    {
+      "epoch": 0.32890689474079077,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018424461515288446,
+      "loss": 1.462,
+      "step": 1096
+    },
+    {
+      "epoch": 0.32920699227248856,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001842295526434704,
+      "loss": 1.464,
+      "step": 1097
+    },
+    {
+      "epoch": 0.32950708980418636,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018421449013405634,
+      "loss": 1.4963,
+      "step": 1098
+    },
+    {
+      "epoch": 0.32980718733588416,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018419942762464228,
+      "loss": 1.6644,
+      "step": 1099
+    },
+    {
+      "epoch": 0.33010728486758195,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001841843651152282,
+      "loss": 1.6634,
+      "step": 1100
+    },
+    {
+      "epoch": 0.33040738239927975,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018416930260581414,
+      "loss": 1.661,
+      "step": 1101
+    },
+    {
+      "epoch": 0.33070747993097754,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018415424009640008,
+      "loss": 1.6341,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3310075774626754,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.000184139177586986,
+      "loss": 1.396,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3313076749943732,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018412411507757193,
+      "loss": 1.4635,
+      "step": 1104
+    },
+    {
+      "epoch": 0.331607772526071,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018410905256815787,
+      "loss": 1.6444,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3319078700577688,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018409399005874378,
+      "loss": 1.2616,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3322079675894666,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00018407892754932972,
+      "loss": 1.5884,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3325080651211644,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018406386503991566,
+      "loss": 1.4907,
+      "step": 1108
+    },
+    {
+      "epoch": 0.33280816265286217,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018404880253050158,
+      "loss": 1.6792,
+      "step": 1109
+    },
+    {
+      "epoch": 0.33310826018455997,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018403374002108752,
+      "loss": 1.7447,
+      "step": 1110
+    },
+    {
+      "epoch": 0.33340835771625776,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018401867751167346,
+      "loss": 1.5654,
+      "step": 1111
+    },
+    {
+      "epoch": 0.33370845524795556,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00018400361500225937,
+      "loss": 1.421,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3340085527796534,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001839885524928453,
+      "loss": 1.6987,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3343086503113512,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00018397348998343125,
+      "loss": 1.6056,
+      "step": 1114
+    },
+    {
+      "epoch": 0.334608747843049,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018395842747401717,
+      "loss": 1.4732,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3349088453747468,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001839433649646031,
+      "loss": 1.4702,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3352089429064446,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00018392830245518905,
+      "loss": 1.6773,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3355090404381424,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018391323994577499,
+      "loss": 1.5165,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3358091379698402,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001838981774363609,
+      "loss": 1.5727,
+      "step": 1119
+    },
+    {
+      "epoch": 0.336109235501538,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00018388311492694684,
+      "loss": 1.3256,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3364093330332358,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018386805241753278,
+      "loss": 1.4564,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3367094305649336,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001838529899081187,
+      "loss": 1.6336,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3370095280966314,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018383792739870463,
+      "loss": 1.48,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3373096256283292,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018382286488929057,
+      "loss": 1.5353,
+      "step": 1124
+    },
+    {
+      "epoch": 0.337609723160027,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001838078023798765,
+      "loss": 1.5878,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3379098206917248,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018379273987046243,
+      "loss": 1.4537,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3382099182234226,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018377767736104837,
+      "loss": 1.6593,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3385100157551204,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018376261485163428,
+      "loss": 1.7235,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3388101132868182,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00018374755234222022,
+      "loss": 1.5893,
+      "step": 1129
+    },
+    {
+      "epoch": 0.339110210818516,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018373248983280616,
+      "loss": 1.5723,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3394103083502138,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018371742732339207,
+      "loss": 1.3538,
+      "step": 1131
+    },
+    {
+      "epoch": 0.33971040588191165,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018370236481397801,
+      "loss": 1.5282,
+      "step": 1132
+    },
+    {
+      "epoch": 0.34001050341360944,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00018368730230456396,
+      "loss": 1.7757,
+      "step": 1133
+    },
+    {
+      "epoch": 0.34031060094530724,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018367223979514987,
+      "loss": 1.4502,
+      "step": 1134
+    },
+    {
+      "epoch": 0.34061069847700504,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001836571772857358,
+      "loss": 1.3864,
+      "step": 1135
+    },
+    {
+      "epoch": 0.34091079600870283,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018364211477632175,
+      "loss": 1.3355,
+      "step": 1136
+    },
+    {
+      "epoch": 0.34121089354040063,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001836270522669077,
+      "loss": 1.3584,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3415109910720984,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001836119897574936,
+      "loss": 1.4669,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3418110886037962,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018359692724807954,
+      "loss": 1.5578,
+      "step": 1139
+    },
+    {
+      "epoch": 0.342111186135494,
+      "grad_norm": 1.875,
+      "learning_rate": 0.00018358186473866548,
+      "loss": 1.677,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3424112836671918,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001835668022292514,
+      "loss": 1.4877,
+      "step": 1141
+    },
+    {
+      "epoch": 0.34271138119888966,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00018355173971983734,
+      "loss": 1.7458,
+      "step": 1142
+    },
+    {
+      "epoch": 0.34301147873058746,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018353667721042328,
+      "loss": 1.6546,
+      "step": 1143
+    },
+    {
+      "epoch": 0.34331157626228526,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001835216147010092,
+      "loss": 1.6732,
+      "step": 1144
+    },
+    {
+      "epoch": 0.34361167379398305,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018350655219159513,
+      "loss": 1.4605,
+      "step": 1145
+    },
+    {
+      "epoch": 0.34391177132568085,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018349148968218107,
+      "loss": 1.4104,
+      "step": 1146
+    },
+    {
+      "epoch": 0.34421186885737864,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00018347642717276698,
+      "loss": 1.5629,
+      "step": 1147
+    },
+    {
+      "epoch": 0.34451196638907644,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018346136466335292,
+      "loss": 1.4986,
+      "step": 1148
+    },
+    {
+      "epoch": 0.34481206392077424,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018344630215393886,
+      "loss": 1.4269,
+      "step": 1149
+    },
+    {
+      "epoch": 0.34511216145247203,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018343123964452478,
+      "loss": 1.5288,
+      "step": 1150
+    },
+    {
+      "epoch": 0.34541225898416983,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018341617713511072,
+      "loss": 1.5612,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3457123565158677,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018340111462569666,
+      "loss": 1.7006,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3460124540475655,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001833860521162826,
+      "loss": 1.4951,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3463125515792633,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001833709896068685,
+      "loss": 1.5741,
+      "step": 1154
+    },
+    {
+      "epoch": 0.34661264911096107,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018335592709745445,
+      "loss": 1.5187,
+      "step": 1155
+    },
+    {
+      "epoch": 0.34691274664265886,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001833408645880404,
+      "loss": 1.5228,
+      "step": 1156
+    },
+    {
+      "epoch": 0.34721284417435666,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001833258020786263,
+      "loss": 1.4322,
+      "step": 1157
+    },
+    {
+      "epoch": 0.34751294170605446,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018331073956921222,
+      "loss": 1.465,
+      "step": 1158
+    },
+    {
+      "epoch": 0.34781303923775225,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018329567705979819,
+      "loss": 1.4681,
+      "step": 1159
+    },
+    {
+      "epoch": 0.34811313676945005,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001832806145503841,
+      "loss": 1.5021,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3484132343011479,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018326555204097004,
+      "loss": 1.5936,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3487133318328457,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018325048953155598,
+      "loss": 1.521,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3490134293645435,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001832354270221419,
+      "loss": 1.5242,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3493135268962413,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018322036451272783,
+      "loss": 1.4083,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3496136244279391,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00018320530200331377,
+      "loss": 1.4233,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3499137219596369,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001831902394938997,
+      "loss": 1.6193,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3502138194913347,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018317517698448563,
+      "loss": 1.3828,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3505139170230325,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00018316011447507154,
+      "loss": 1.5177,
+      "step": 1168
+    },
+    {
+      "epoch": 0.35081401455473027,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018314505196565748,
+      "loss": 1.5117,
+      "step": 1169
+    },
+    {
+      "epoch": 0.35111411208642807,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00018312998945624342,
+      "loss": 1.5507,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3514142096181259,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018311492694682933,
+      "loss": 1.4672,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3517143071498237,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001830998644374153,
+      "loss": 1.5675,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3520144046815215,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018308480192800121,
+      "loss": 1.3824,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3523145022132193,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018306973941858713,
+      "loss": 1.3632,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3526145997449171,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001830546769091731,
+      "loss": 1.4864,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3529146972766149,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.000183039614399759,
+      "loss": 1.3885,
+      "step": 1176
+    },
+    {
+      "epoch": 0.3532147948083127,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018302455189034492,
+      "loss": 1.6913,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3535148923400105,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001830094893809309,
+      "loss": 1.5454,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3538149898717083,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001829944268715168,
+      "loss": 1.5729,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3541150874034061,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018297936436210274,
+      "loss": 1.7296,
+      "step": 1180
+    },
+    {
+      "epoch": 0.35441518493510393,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018296430185268866,
+      "loss": 1.3344,
+      "step": 1181
+    },
+    {
+      "epoch": 0.35471528246680173,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001829492393432746,
+      "loss": 1.4049,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3550153799984995,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00018293417683386054,
+      "loss": 1.4347,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3553154775301973,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018291911432444645,
+      "loss": 1.5985,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3556155750618951,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001829040518150324,
+      "loss": 1.6664,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3559156725935929,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00018288898930561833,
+      "loss": 1.6796,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3562157701252907,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018287392679620424,
+      "loss": 1.4223,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3565158676569885,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001828588642867902,
+      "loss": 1.416,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3568159651886863,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018284380177737612,
+      "loss": 1.4607,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3571160627203841,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018282873926796204,
+      "loss": 1.6294,
+      "step": 1190
+    },
+    {
+      "epoch": 0.35741616025208195,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.000182813676758548,
+      "loss": 1.6671,
+      "step": 1191
+    },
+    {
+      "epoch": 0.35771625778377975,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018279861424913392,
+      "loss": 1.4268,
+      "step": 1192
+    },
+    {
+      "epoch": 0.35801635531547754,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00018278355173971983,
+      "loss": 1.5886,
+      "step": 1193
+    },
+    {
+      "epoch": 0.35831645284717534,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00018276848923030577,
+      "loss": 1.5541,
+      "step": 1194
+    },
+    {
+      "epoch": 0.35861655037887313,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001827534267208917,
+      "loss": 1.7319,
+      "step": 1195
+    },
+    {
+      "epoch": 0.35891664791057093,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018273836421147763,
+      "loss": 1.4454,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3592167454422687,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018272330170206357,
+      "loss": 1.5426,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3595168429739665,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001827082391926495,
+      "loss": 1.4023,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3598169405056643,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018269317668323545,
+      "loss": 1.5891,
+      "step": 1199
+    },
+    {
+      "epoch": 0.36011703803736217,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018267811417382136,
+      "loss": 1.6734,
+      "step": 1200
+    },
+    {
+      "epoch": 0.36041713556905997,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001826630516644073,
+      "loss": 1.502,
+      "step": 1201
+    },
+    {
+      "epoch": 0.36071723310075776,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018264798915499324,
+      "loss": 1.3725,
+      "step": 1202
+    },
+    {
+      "epoch": 0.36101733063245556,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018263292664557915,
+      "loss": 1.5563,
+      "step": 1203
+    },
+    {
+      "epoch": 0.36131742816415335,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001826178641361651,
+      "loss": 1.4485,
+      "step": 1204
+    },
+    {
+      "epoch": 0.36161752569585115,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018260280162675103,
+      "loss": 1.3636,
+      "step": 1205
+    },
+    {
+      "epoch": 0.36191762322754895,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018258773911733695,
+      "loss": 1.4474,
+      "step": 1206
+    },
+    {
+      "epoch": 0.36221772075924674,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001825726766079229,
+      "loss": 1.2702,
+      "step": 1207
+    },
+    {
+      "epoch": 0.36251781829094454,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018255761409850883,
+      "loss": 1.3956,
+      "step": 1208
+    },
+    {
+      "epoch": 0.36281791582264233,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018254255158909474,
+      "loss": 1.6294,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3631180133543402,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018252748907968068,
+      "loss": 1.4714,
+      "step": 1210
+    },
+    {
+      "epoch": 0.363418110886038,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018251242657026662,
+      "loss": 1.5119,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3637182084177358,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018249736406085253,
+      "loss": 1.4259,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3640183059494336,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00018248230155143847,
+      "loss": 1.5371,
+      "step": 1213
+    },
+    {
+      "epoch": 0.36431840348113137,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018246723904202442,
+      "loss": 1.606,
+      "step": 1214
+    },
+    {
+      "epoch": 0.36461850101282917,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018245217653261036,
+      "loss": 1.4921,
+      "step": 1215
+    },
+    {
+      "epoch": 0.36491859854452696,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018243711402319627,
+      "loss": 1.3798,
+      "step": 1216
+    },
+    {
+      "epoch": 0.36521869607622476,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001824220515137822,
+      "loss": 1.4758,
+      "step": 1217
+    },
+    {
+      "epoch": 0.36551879360792255,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018240698900436815,
+      "loss": 1.2529,
+      "step": 1218
+    },
+    {
+      "epoch": 0.36581889113962035,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018239192649495406,
+      "loss": 1.2925,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3661189886713182,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018237686398554,
+      "loss": 1.5067,
+      "step": 1220
+    },
+    {
+      "epoch": 0.366419086203016,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00018236180147612594,
+      "loss": 1.4247,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3667191837347138,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018234673896671186,
+      "loss": 1.4453,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3670192812664116,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001823316764572978,
+      "loss": 1.4473,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3673193787981094,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018231661394788374,
+      "loss": 1.5146,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3676194763298072,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018230155143846965,
+      "loss": 1.4678,
+      "step": 1225
+    },
+    {
+      "epoch": 0.367919573861505,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001822864889290556,
+      "loss": 1.5658,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3682196713932028,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018227142641964153,
+      "loss": 1.301,
+      "step": 1227
+    },
+    {
+      "epoch": 0.36851976892490057,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018225636391022744,
+      "loss": 1.6967,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3688198664565984,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018224130140081338,
+      "loss": 1.5656,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3691199639882962,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018222623889139932,
+      "loss": 1.4321,
+      "step": 1230
+    },
+    {
+      "epoch": 0.369420061519994,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00018221117638198524,
+      "loss": 1.5329,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3697201590516918,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018219611387257118,
+      "loss": 1.4806,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3700202565833896,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018218105136315712,
+      "loss": 1.4863,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3703203541150874,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00018216598885374306,
+      "loss": 1.4558,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3706204516467852,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018215092634432897,
+      "loss": 1.6157,
+      "step": 1235
+    },
+    {
+      "epoch": 0.370920549178483,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001821358638349149,
+      "loss": 1.4377,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3712206467101808,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018212080132550085,
+      "loss": 1.3446,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3715207442418786,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018210573881608677,
+      "loss": 1.2719,
+      "step": 1238
+    },
+    {
+      "epoch": 0.37182084177357644,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018209067630667268,
+      "loss": 1.495,
+      "step": 1239
+    },
+    {
+      "epoch": 0.37212093930527423,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00018207561379725865,
+      "loss": 1.6259,
+      "step": 1240
+    },
+    {
+      "epoch": 0.37242103683697203,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00018206055128784456,
+      "loss": 1.5863,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3727211343686698,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001820454887784305,
+      "loss": 1.3801,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3730212319003676,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018203042626901644,
+      "loss": 1.4744,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3733213294320654,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018201536375960235,
+      "loss": 1.4595,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3736214269637632,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001820003012501883,
+      "loss": 1.5042,
+      "step": 1245
+    },
+    {
+      "epoch": 0.373921524495461,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00018198523874077423,
+      "loss": 1.5239,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3742216220271588,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018197017623136015,
+      "loss": 1.4747,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3745217195588566,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.0001819551137219461,
+      "loss": 1.3579,
+      "step": 1248
+    },
+    {
+      "epoch": 0.37482181709055445,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018194005121253203,
+      "loss": 1.5114,
+      "step": 1249
+    },
+    {
+      "epoch": 0.37512191462225225,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018192498870311794,
+      "loss": 1.3312,
+      "step": 1250
+    },
+    {
+      "epoch": 0.37542201215395005,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00018190992619370388,
+      "loss": 1.9791,
+      "step": 1251
+    },
+    {
+      "epoch": 0.37572210968564784,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001818948636842898,
+      "loss": 1.7578,
+      "step": 1252
+    },
+    {
+      "epoch": 0.37602220721734564,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00018187980117487576,
+      "loss": 1.5338,
+      "step": 1253
+    },
+    {
+      "epoch": 0.37632230474904343,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00018186473866546167,
+      "loss": 1.4274,
+      "step": 1254
+    },
+    {
+      "epoch": 0.37662240228074123,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001818496761560476,
+      "loss": 1.5909,
+      "step": 1255
+    },
+    {
+      "epoch": 0.376922499812439,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018183461364663356,
+      "loss": 1.4188,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3772225973441368,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018181955113721947,
+      "loss": 1.6454,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3775226948758346,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018180448862780538,
+      "loss": 1.6341,
+      "step": 1258
+    },
+    {
+      "epoch": 0.37782279240753247,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018178942611839135,
+      "loss": 1.5203,
+      "step": 1259
+    },
+    {
+      "epoch": 0.37812288993923027,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018177436360897726,
+      "loss": 1.3347,
+      "step": 1260
+    },
+    {
+      "epoch": 0.37842298747092806,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001817593010995632,
+      "loss": 1.6238,
+      "step": 1261
+    },
+    {
+      "epoch": 0.37872308500262586,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00018174423859014914,
+      "loss": 1.5044,
+      "step": 1262
+    },
+    {
+      "epoch": 0.37902318253432365,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018172917608073506,
+      "loss": 1.5297,
+      "step": 1263
+    },
+    {
+      "epoch": 0.37932328006602145,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.000181714113571321,
+      "loss": 1.5091,
+      "step": 1264
+    },
+    {
+      "epoch": 0.37962337759771925,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001816990510619069,
+      "loss": 1.4267,
+      "step": 1265
+    },
+    {
+      "epoch": 0.37992347512941704,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00018168398855249285,
+      "loss": 1.5707,
+      "step": 1266
+    },
+    {
+      "epoch": 0.38022357266111484,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001816689260430788,
+      "loss": 1.5961,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3805236701928127,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001816538635336647,
+      "loss": 1.4994,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3808237677245105,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018163880102425067,
+      "loss": 1.4226,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3811238652562083,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018162373851483658,
+      "loss": 1.5234,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3814239627879061,
+      "grad_norm": 1.875,
+      "learning_rate": 0.0001816086760054225,
+      "loss": 1.5703,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3817240603196039,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018159361349600846,
+      "loss": 1.6474,
+      "step": 1272
+    },
+    {
+      "epoch": 0.38202415785130167,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018157855098659438,
+      "loss": 1.5379,
+      "step": 1273
+    },
+    {
+      "epoch": 0.38232425538299947,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001815634884771803,
+      "loss": 1.5261,
+      "step": 1274
+    },
+    {
+      "epoch": 0.38262435291469726,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00018154842596776623,
+      "loss": 1.6148,
+      "step": 1275
+    },
+    {
+      "epoch": 0.38292445044639506,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018153336345835217,
+      "loss": 1.2961,
+      "step": 1276
+    },
+    {
+      "epoch": 0.38322454797809286,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018151830094893809,
+      "loss": 1.4594,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3835246455097907,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018150323843952403,
+      "loss": 1.4943,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3838247430414885,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018148817593010997,
+      "loss": 1.337,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3841248405731863,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001814731134206959,
+      "loss": 1.545,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3844249381048841,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00018145805091128182,
+      "loss": 1.5701,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3847250356365819,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018144298840186776,
+      "loss": 1.5943,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3850251331682797,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001814279258924537,
+      "loss": 1.4053,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3853252306999775,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001814128633830396,
+      "loss": 1.5405,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3856253282316753,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00018139780087362555,
+      "loss": 1.3923,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3859254257633731,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001813827383642115,
+      "loss": 1.6602,
+      "step": 1286
+    },
+    {
+      "epoch": 0.38622552329507087,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001813676758547974,
+      "loss": 1.4851,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3865256208267687,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018135261334538335,
+      "loss": 1.3762,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3868257183584665,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001813375508359693,
+      "loss": 1.2325,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3871258158901643,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001813224883265552,
+      "loss": 1.2877,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3874259134218621,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00018130742581714114,
+      "loss": 1.3455,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3877260109535599,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00018129236330772708,
+      "loss": 1.7059,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3880261084852577,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.000181277300798313,
+      "loss": 1.4718,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3883262060169555,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018126223828889893,
+      "loss": 1.3146,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3886263035486533,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018124717577948488,
+      "loss": 1.4029,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3889264010803511,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00018123211327007082,
+      "loss": 1.4421,
+      "step": 1296
+    },
+    {
+      "epoch": 0.38922649861204894,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018121705076065673,
+      "loss": 1.4713,
+      "step": 1297
+    },
+    {
+      "epoch": 0.38952659614374674,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018120198825124267,
+      "loss": 1.4229,
+      "step": 1298
+    },
+    {
+      "epoch": 0.38982669367544454,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001811869257418286,
+      "loss": 1.5183,
+      "step": 1299
+    },
+    {
+      "epoch": 0.39012679120714233,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018117186323241452,
+      "loss": 1.5479,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3904268887388401,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018115680072300046,
+      "loss": 1.486,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3907269862705379,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001811417382135864,
+      "loss": 1.4879,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3910270838022357,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00018112667570417232,
+      "loss": 1.4292,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3913271813339335,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018111161319475826,
+      "loss": 1.5659,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3916272788656313,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001810965506853442,
+      "loss": 1.4714,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3919273763973291,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001810814881759301,
+      "loss": 1.6376,
+      "step": 1306
+    },
+    {
+      "epoch": 0.39222747392902696,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018106642566651605,
+      "loss": 1.4629,
+      "step": 1307
+    },
+    {
+      "epoch": 0.39252757146072476,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.000181051363157102,
+      "loss": 1.705,
+      "step": 1308
+    },
+    {
+      "epoch": 0.39282766899242255,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001810363006476879,
+      "loss": 1.4012,
+      "step": 1309
+    },
+    {
+      "epoch": 0.39312776652412035,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00018102123813827384,
+      "loss": 1.6108,
+      "step": 1310
+    },
+    {
+      "epoch": 0.39342786405581814,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018100617562885978,
+      "loss": 1.4896,
+      "step": 1311
+    },
+    {
+      "epoch": 0.39372796158751594,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001809911131194457,
+      "loss": 1.4154,
+      "step": 1312
+    },
+    {
+      "epoch": 0.39402805911921374,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018097605061003164,
+      "loss": 1.5588,
+      "step": 1313
+    },
+    {
+      "epoch": 0.39432815665091153,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018096098810061758,
+      "loss": 1.4915,
+      "step": 1314
+    },
+    {
+      "epoch": 0.39462825418260933,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00018094592559120352,
+      "loss": 1.5737,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3949283517143071,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018093086308178943,
+      "loss": 1.4076,
+      "step": 1316
+    },
+    {
+      "epoch": 0.395228449246005,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00018091580057237537,
+      "loss": 1.6264,
+      "step": 1317
+    },
+    {
+      "epoch": 0.39552854677770277,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001809007380629613,
+      "loss": 1.5613,
+      "step": 1318
+    },
+    {
+      "epoch": 0.39582864430940057,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018088567555354723,
+      "loss": 1.561,
+      "step": 1319
+    },
+    {
+      "epoch": 0.39612874184109836,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00018087061304413317,
+      "loss": 1.3679,
+      "step": 1320
+    },
+    {
+      "epoch": 0.39642883937279616,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001808555505347191,
+      "loss": 1.5045,
+      "step": 1321
+    },
+    {
+      "epoch": 0.39672893690449396,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018084048802530502,
+      "loss": 1.4962,
+      "step": 1322
+    },
+    {
+      "epoch": 0.39702903443619175,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00018082542551589096,
+      "loss": 1.5476,
+      "step": 1323
+    },
+    {
+      "epoch": 0.39732913196788955,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001808103630064769,
+      "loss": 1.4432,
+      "step": 1324
+    },
+    {
+      "epoch": 0.39762922949958734,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001807953004970628,
+      "loss": 1.5359,
+      "step": 1325
+    },
+    {
+      "epoch": 0.39792932703128514,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00018078023798764875,
+      "loss": 1.6741,
+      "step": 1326
+    },
+    {
+      "epoch": 0.398229424562983,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001807651754782347,
+      "loss": 1.3267,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3985295220946808,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001807501129688206,
+      "loss": 1.7069,
+      "step": 1328
+    },
+    {
+      "epoch": 0.3988296196263786,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018073505045940655,
+      "loss": 1.3988,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3991297171580764,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.0001807199879499925,
+      "loss": 1.633,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3994298146897742,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001807049254405784,
+      "loss": 1.626,
+      "step": 1331
+    },
+    {
+      "epoch": 0.399729912221472,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00018068986293116434,
+      "loss": 1.4526,
+      "step": 1332
+    },
+    {
+      "epoch": 0.40003000975316977,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018067480042175025,
+      "loss": 1.3597,
+      "step": 1333
+    },
+    {
+      "epoch": 0.40033010728486756,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018065973791233622,
+      "loss": 1.5184,
+      "step": 1334
+    },
+    {
+      "epoch": 0.40063020481656536,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018064467540292213,
+      "loss": 1.4315,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4009303023482632,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00018062961289350805,
+      "loss": 1.5643,
+      "step": 1336
+    },
+    {
+      "epoch": 0.401230399879961,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00018061455038409402,
+      "loss": 1.6331,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4015304974116588,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018059948787467993,
+      "loss": 1.5273,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4018305949433566,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018058442536526584,
+      "loss": 1.3533,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4021306924750544,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001805693628558518,
+      "loss": 1.3344,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4024307900067522,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018055430034643772,
+      "loss": 1.6337,
+      "step": 1341
+    },
+    {
+      "epoch": 0.40273088753845,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00018053923783702366,
+      "loss": 1.6274,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4030309850701478,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001805241753276096,
+      "loss": 1.4707,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4033310826018456,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018050911281819552,
+      "loss": 1.3989,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4036311801335434,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018049405030878146,
+      "loss": 1.5383,
+      "step": 1345
+    },
+    {
+      "epoch": 0.40393127766524123,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018047898779936737,
+      "loss": 1.4,
+      "step": 1346
+    },
+    {
+      "epoch": 0.404231375196939,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001804639252899533,
+      "loss": 1.37,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4045314727286368,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018044886278053925,
+      "loss": 1.2958,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4048315702603346,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00018043380027112516,
+      "loss": 1.451,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4051316677920324,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00018041873776171113,
+      "loss": 1.5109,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4054317653237302,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00018040367525229704,
+      "loss": 1.3292,
+      "step": 1351
+    },
+    {
+      "epoch": 0.405731862855428,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018038861274288296,
+      "loss": 1.365,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4060319603871258,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018037355023346892,
+      "loss": 1.475,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4063320579188236,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00018035848772405484,
+      "loss": 1.573,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4066321554505214,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018034342521464075,
+      "loss": 1.2978,
+      "step": 1355
+    },
+    {
+      "epoch": 0.40693225298221924,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00018032836270522672,
+      "loss": 1.2807,
+      "step": 1356
+    },
+    {
+      "epoch": 0.40723235051391704,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018031330019581263,
+      "loss": 1.5733,
+      "step": 1357
+    },
+    {
+      "epoch": 0.40753244804561484,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00018029823768639855,
+      "loss": 1.3851,
+      "step": 1358
+    },
+    {
+      "epoch": 0.40783254557731263,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018028317517698449,
+      "loss": 1.5227,
+      "step": 1359
+    },
+    {
+      "epoch": 0.40813264310901043,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018026811266757043,
+      "loss": 1.4466,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4084327406407082,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018025305015815637,
+      "loss": 1.5674,
+      "step": 1361
+    },
+    {
+      "epoch": 0.408732838172406,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00018023798764874228,
+      "loss": 1.584,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4090329357041038,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00018022292513932822,
+      "loss": 1.3032,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4093330332358016,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00018020786262991416,
+      "loss": 1.341,
+      "step": 1364
+    },
+    {
+      "epoch": 0.40963313076749946,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00018019280012050007,
+      "loss": 1.4231,
+      "step": 1365
+    },
+    {
+      "epoch": 0.40993322829919726,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.000180177737611086,
+      "loss": 1.4941,
+      "step": 1366
+    },
+    {
+      "epoch": 0.41023332583089506,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018016267510167195,
+      "loss": 1.4035,
+      "step": 1367
+    },
+    {
+      "epoch": 0.41053342336259285,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00018014761259225787,
+      "loss": 1.4461,
+      "step": 1368
+    },
+    {
+      "epoch": 0.41083352089429065,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.0001801325500828438,
+      "loss": 1.7714,
+      "step": 1369
+    },
+    {
+      "epoch": 0.41113361842598845,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00018011748757342975,
+      "loss": 1.432,
+      "step": 1370
+    },
+    {
+      "epoch": 0.41143371595768624,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00018010242506401566,
+      "loss": 1.601,
+      "step": 1371
+    },
+    {
+      "epoch": 0.41173381348938404,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001800873625546016,
+      "loss": 1.4111,
+      "step": 1372
+    },
+    {
+      "epoch": 0.41203391102108183,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018007230004518754,
+      "loss": 1.4485,
+      "step": 1373
+    },
+    {
+      "epoch": 0.41233400855277963,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00018005723753577345,
+      "loss": 1.3593,
+      "step": 1374
+    },
+    {
+      "epoch": 0.4126341060844775,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001800421750263594,
+      "loss": 1.4183,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4129342036161753,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00018002711251694534,
+      "loss": 1.386,
+      "step": 1376
+    },
+    {
+      "epoch": 0.4132343011478731,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00018001205000753128,
+      "loss": 1.4322,
+      "step": 1377
+    },
+    {
+      "epoch": 0.41353439867957087,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001799969874981172,
+      "loss": 1.452,
+      "step": 1378
+    },
+    {
+      "epoch": 0.41383449621126867,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017998192498870313,
+      "loss": 1.4755,
+      "step": 1379
+    },
+    {
+      "epoch": 0.41413459374296646,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017996686247928907,
+      "loss": 1.4199,
+      "step": 1380
+    },
+    {
+      "epoch": 0.41443469127466426,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00017995179996987498,
+      "loss": 1.5432,
+      "step": 1381
+    },
+    {
+      "epoch": 0.41473478880636205,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017993673746046092,
+      "loss": 1.3065,
+      "step": 1382
+    },
+    {
+      "epoch": 0.41503488633805985,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017992167495104686,
+      "loss": 1.4612,
+      "step": 1383
+    },
+    {
+      "epoch": 0.41533498386975765,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017990661244163278,
+      "loss": 1.4473,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4156350814014555,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017989154993221872,
+      "loss": 1.5235,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4159351789331533,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017987648742280466,
+      "loss": 1.4225,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4162352764648511,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017986142491339057,
+      "loss": 1.5347,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4165353739965489,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001798463624039765,
+      "loss": 1.3585,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4168354715282467,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017983129989456245,
+      "loss": 1.4722,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4171355690599445,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017981623738514836,
+      "loss": 1.5159,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4174356665916423,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001798011748757343,
+      "loss": 1.4462,
+      "step": 1391
+    },
+    {
+      "epoch": 0.41773576412334007,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017978611236632024,
+      "loss": 1.5315,
+      "step": 1392
+    },
+    {
+      "epoch": 0.41803586165503787,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00017977104985690616,
+      "loss": 1.4445,
+      "step": 1393
+    },
+    {
+      "epoch": 0.41833595918673566,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001797559873474921,
+      "loss": 1.3101,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4186360567184335,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017974092483807804,
+      "loss": 1.5016,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4189361542501313,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017972586232866398,
+      "loss": 1.6294,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4192362517818291,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.0001797107998192499,
+      "loss": 1.4812,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4195363493135269,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017969573730983583,
+      "loss": 1.4232,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4198364468452247,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017968067480042177,
+      "loss": 1.4952,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4201365443769225,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017966561229100769,
+      "loss": 1.4685,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4204366419086203,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00017965054978159363,
+      "loss": 1.4839,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4207367394403181,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00017963548727217957,
+      "loss": 1.3333,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4210368369720159,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017962042476276548,
+      "loss": 1.4283,
+      "step": 1403
+    },
+    {
+      "epoch": 0.42133693450371373,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017960536225335142,
+      "loss": 1.3913,
+      "step": 1404
+    },
+    {
+      "epoch": 0.42163703203541153,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00017959029974393736,
+      "loss": 1.5024,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4219371295671093,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017957523723452327,
+      "loss": 1.4014,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4222372270988071,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001795601747251092,
+      "loss": 1.5321,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4225373246305049,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017954511221569515,
+      "loss": 1.394,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4228374221622027,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017953004970628107,
+      "loss": 1.6826,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4231375196939005,
+      "grad_norm": 1.375,
+      "learning_rate": 0.000179514987196867,
+      "loss": 1.4817,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4234376172255983,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017949992468745295,
+      "loss": 1.2468,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4237377147572961,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017948486217803886,
+      "loss": 1.4821,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4240378122889939,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001794697996686248,
+      "loss": 1.4521,
+      "step": 1413
+    },
+    {
+      "epoch": 0.42433790982069175,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017945473715921074,
+      "loss": 1.5434,
+      "step": 1414
+    },
+    {
+      "epoch": 0.42463800735238955,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00017943967464979668,
+      "loss": 1.4248,
+      "step": 1415
+    },
+    {
+      "epoch": 0.42493810488408734,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001794246121403826,
+      "loss": 1.3033,
+      "step": 1416
+    },
+    {
+      "epoch": 0.42523820241578514,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001794095496309685,
+      "loss": 1.4244,
+      "step": 1417
+    },
+    {
+      "epoch": 0.42553829994748293,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017939448712155448,
+      "loss": 1.4755,
+      "step": 1418
+    },
+    {
+      "epoch": 0.42583839747918073,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001793794246121404,
+      "loss": 1.5145,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4261384950108785,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001793643621027263,
+      "loss": 1.3148,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4264385925425763,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017934929959331227,
+      "loss": 1.7471,
+      "step": 1421
+    },
+    {
+      "epoch": 0.4267386900742741,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017933423708389818,
+      "loss": 1.4099,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4270387876059719,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00017931917457448412,
+      "loss": 1.4809,
+      "step": 1423
+    },
+    {
+      "epoch": 0.42733888513766977,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017930411206507006,
+      "loss": 1.3443,
+      "step": 1424
+    },
+    {
+      "epoch": 0.42763898266936756,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017928904955565598,
+      "loss": 1.3324,
+      "step": 1425
+    },
+    {
+      "epoch": 0.42793908020106536,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017927398704624192,
+      "loss": 1.5537,
+      "step": 1426
+    },
+    {
+      "epoch": 0.42823917773276315,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017925892453682786,
+      "loss": 1.5573,
+      "step": 1427
+    },
+    {
+      "epoch": 0.42853927526446095,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00017924386202741377,
+      "loss": 1.6209,
+      "step": 1428
+    },
+    {
+      "epoch": 0.42883937279615875,
+      "grad_norm": 1.75,
+      "learning_rate": 0.0001792287995179997,
+      "loss": 1.5307,
+      "step": 1429
+    },
+    {
+      "epoch": 0.42913947032785654,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017921373700858562,
+      "loss": 1.491,
+      "step": 1430
+    },
+    {
+      "epoch": 0.42943956785955434,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001791986744991716,
+      "loss": 1.4682,
+      "step": 1431
+    },
+    {
+      "epoch": 0.42973966539125213,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001791836119897575,
+      "loss": 1.3672,
+      "step": 1432
+    },
+    {
+      "epoch": 0.43003976292295,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00017916854948034342,
+      "loss": 1.5423,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4303398604546478,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017915348697092938,
+      "loss": 1.3178,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4306399579863456,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001791384244615153,
+      "loss": 1.5193,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4309400555180434,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001791233619521012,
+      "loss": 1.3752,
+      "step": 1436
+    },
+    {
+      "epoch": 0.43124015304974117,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017910829944268718,
+      "loss": 1.2543,
+      "step": 1437
+    },
+    {
+      "epoch": 0.43154025058143897,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001790932369332731,
+      "loss": 1.4975,
+      "step": 1438
+    },
+    {
+      "epoch": 0.43184034811313676,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.000179078174423859,
+      "loss": 1.5729,
+      "step": 1439
+    },
+    {
+      "epoch": 0.43214044564483456,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017906311191444495,
+      "loss": 1.5097,
+      "step": 1440
+    },
+    {
+      "epoch": 0.43244054317653235,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017904804940503089,
+      "loss": 1.5565,
+      "step": 1441
+    },
+    {
+      "epoch": 0.43274064070823015,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017903298689561683,
+      "loss": 1.4658,
+      "step": 1442
+    },
+    {
+      "epoch": 0.433040738239928,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017901792438620274,
+      "loss": 1.3659,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4333408357716258,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00017900286187678868,
+      "loss": 1.435,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4336409333033236,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017898779936737462,
+      "loss": 1.4398,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4339410308350214,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00017897273685796053,
+      "loss": 1.5693,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4342411283667192,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017895767434854647,
+      "loss": 1.4636,
+      "step": 1447
+    },
+    {
+      "epoch": 0.434541225898417,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017894261183913241,
+      "loss": 1.3669,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4348413234301148,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00017892754932971833,
+      "loss": 1.4625,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4351414209618126,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001789124868203043,
+      "loss": 1.6731,
+      "step": 1450
+    },
+    {
+      "epoch": 0.43544151849351037,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001788974243108902,
+      "loss": 1.3704,
+      "step": 1451
+    },
+    {
+      "epoch": 0.43574161602520817,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00017888236180147612,
+      "loss": 1.273,
+      "step": 1452
+    },
+    {
+      "epoch": 0.436041713556906,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017886729929206206,
+      "loss": 1.4466,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4363418110886038,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.000178852236782648,
+      "loss": 1.4342,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4366419086203016,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017883717427323391,
+      "loss": 1.3789,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4369420061519994,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017882211176381985,
+      "loss": 1.3388,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4372421036836972,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001788070492544058,
+      "loss": 1.7755,
+      "step": 1457
+    },
+    {
+      "epoch": 0.437542201215395,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00017879198674499174,
+      "loss": 1.3748,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4378422987470928,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017877692423557765,
+      "loss": 1.1952,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4381423962787906,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001787618617261636,
+      "loss": 1.4297,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4384424938104884,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017874679921674953,
+      "loss": 1.4504,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4387425913421862,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017873173670733544,
+      "loss": 1.45,
+      "step": 1462
+    },
+    {
+      "epoch": 0.43904268887388403,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017871667419792138,
+      "loss": 1.4691,
+      "step": 1463
+    },
+    {
+      "epoch": 0.43934278640558183,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00017870161168850732,
+      "loss": 1.4145,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4396428839372796,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017868654917909324,
+      "loss": 1.4319,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4399429814689774,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017867148666967918,
+      "loss": 1.6455,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4402430790006752,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00017865642416026512,
+      "loss": 1.4594,
+      "step": 1467
+    },
+    {
+      "epoch": 0.440543176532373,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017864136165085103,
+      "loss": 1.545,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4408432740640708,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017862629914143697,
+      "loss": 1.3448,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4411433715957686,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001786112366320229,
+      "loss": 1.4586,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4414434691274664,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017859617412260882,
+      "loss": 1.8423,
+      "step": 1471
+    },
+    {
+      "epoch": 0.44174356665916426,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017858111161319476,
+      "loss": 1.4754,
+      "step": 1472
+    },
+    {
+      "epoch": 0.44204366419086205,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001785660491037807,
+      "loss": 1.5889,
+      "step": 1473
+    },
+    {
+      "epoch": 0.44234376172255985,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00017855098659436662,
+      "loss": 1.3541,
+      "step": 1474
+    },
+    {
+      "epoch": 0.44264385925425764,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00017853592408495256,
+      "loss": 1.3647,
+      "step": 1475
+    },
+    {
+      "epoch": 0.44294395678595544,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001785208615755385,
+      "loss": 1.4987,
+      "step": 1476
+    },
+    {
+      "epoch": 0.44324405431765324,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017850579906612444,
+      "loss": 1.5614,
+      "step": 1477
+    },
+    {
+      "epoch": 0.44354415184935103,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017849073655671035,
+      "loss": 1.4074,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4438442493810488,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001784756740472963,
+      "loss": 1.5085,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4441443469127466,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017846061153788223,
+      "loss": 1.7233,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00017844554902846815,
+      "loss": 1.4665,
+      "step": 1481
+    },
+    {
+      "epoch": 0.44474454197614227,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017843048651905409,
+      "loss": 1.5516,
+      "step": 1482
+    },
+    {
+      "epoch": 0.44504463950784007,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017841542400964003,
+      "loss": 1.4761,
+      "step": 1483
+    },
+    {
+      "epoch": 0.44534473703953786,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017840036150022594,
+      "loss": 1.5358,
+      "step": 1484
+    },
+    {
+      "epoch": 0.44564483457123566,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017838529899081188,
+      "loss": 1.495,
+      "step": 1485
+    },
+    {
+      "epoch": 0.44594493210293346,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017837023648139782,
+      "loss": 1.5607,
+      "step": 1486
+    },
+    {
+      "epoch": 0.44624502963463125,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017835517397198373,
+      "loss": 1.3847,
+      "step": 1487
+    },
+    {
+      "epoch": 0.44654512716632905,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00017834011146256967,
+      "loss": 1.4687,
+      "step": 1488
+    },
+    {
+      "epoch": 0.44684522469802684,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00017832504895315561,
+      "loss": 1.3506,
+      "step": 1489
+    },
+    {
+      "epoch": 0.44714532222972464,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017830998644374153,
+      "loss": 1.4505,
+      "step": 1490
+    },
+    {
+      "epoch": 0.44744541976142244,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017829492393432747,
+      "loss": 1.4814,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4477455172931203,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001782798614249134,
+      "loss": 1.6193,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4480456148248181,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017826479891549932,
+      "loss": 1.3177,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4483457123565159,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017824973640608526,
+      "loss": 1.4615,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4486458098882137,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001782346738966712,
+      "loss": 1.4205,
+      "step": 1495
+    },
+    {
+      "epoch": 0.44894590741991147,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017821961138725714,
+      "loss": 1.5125,
+      "step": 1496
+    },
+    {
+      "epoch": 0.44924600495160927,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00017820454887784305,
+      "loss": 1.6038,
+      "step": 1497
+    },
+    {
+      "epoch": 0.44954610248330706,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.000178189486368429,
+      "loss": 1.556,
+      "step": 1498
+    },
+    {
+      "epoch": 0.44984620001500486,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00017817442385901494,
+      "loss": 1.4614,
+      "step": 1499
+    },
+    {
+      "epoch": 0.45014629754670266,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017815936134960085,
+      "loss": 1.2991,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4504463950784005,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017814429884018676,
+      "loss": 1.4926,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4507464926100983,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017812923633077273,
+      "loss": 1.5056,
+      "step": 1502
+    },
+    {
+      "epoch": 0.4510465901417961,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00017811417382135864,
+      "loss": 1.4131,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4513466876734939,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017809911131194458,
+      "loss": 1.4855,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4516467852051917,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00017808404880253052,
+      "loss": 1.6035,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4519468827368895,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017806898629311644,
+      "loss": 1.5364,
+      "step": 1506
+    },
+    {
+      "epoch": 0.4522469802685873,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017805392378370238,
+      "loss": 1.3304,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4525470778002851,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017803886127428832,
+      "loss": 1.3782,
+      "step": 1508
+    },
+    {
+      "epoch": 0.4528471753319829,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017802379876487423,
+      "loss": 1.4598,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4531472728636807,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017800873625546017,
+      "loss": 1.3191,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4534473703953785,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017799367374604608,
+      "loss": 1.252,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4537474679270763,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017797861123663205,
+      "loss": 1.3543,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4540475654587741,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017796354872721796,
+      "loss": 1.4726,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4543476629904719,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017794848621780388,
+      "loss": 1.4823,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4546477605221697,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017793342370838984,
+      "loss": 1.6547,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4549478580538675,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017791836119897576,
+      "loss": 1.2958,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4552479555855653,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017790329868956167,
+      "loss": 1.5808,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4555480531172631,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017788823618014764,
+      "loss": 1.6256,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4558481506489609,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017787317367073355,
+      "loss": 1.5886,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4561482481806587,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017785811116131947,
+      "loss": 1.5499,
+      "step": 1520
+    },
+    {
+      "epoch": 0.45644834571235654,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017784304865190543,
+      "loss": 1.4487,
+      "step": 1521
+    },
+    {
+      "epoch": 0.45674844324405434,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017782798614249135,
+      "loss": 1.4654,
+      "step": 1522
+    },
+    {
+      "epoch": 0.45704854077575213,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00017781292363307729,
+      "loss": 1.6,
+      "step": 1523
+    },
+    {
+      "epoch": 0.45734863830744993,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001777978611236632,
+      "loss": 1.5656,
+      "step": 1524
+    },
+    {
+      "epoch": 0.4576487358391477,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017778279861424914,
+      "loss": 1.3812,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4579488333708455,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00017776773610483508,
+      "loss": 1.4992,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4582489309025433,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.000177752673595421,
+      "loss": 1.6243,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4585490284342411,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017773761108600693,
+      "loss": 1.4253,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4588491259659389,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017772254857659287,
+      "loss": 1.6903,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4591492234976367,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001777074860671788,
+      "loss": 1.3985,
+      "step": 1530
+    },
+    {
+      "epoch": 0.45944932102933456,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017769242355776475,
+      "loss": 1.4711,
+      "step": 1531
+    },
+    {
+      "epoch": 0.45974941856103235,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017767736104835067,
+      "loss": 1.294,
+      "step": 1532
+    },
+    {
+      "epoch": 0.46004951609273015,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017766229853893658,
+      "loss": 1.3346,
+      "step": 1533
+    },
+    {
+      "epoch": 0.46034961362442794,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017764723602952255,
+      "loss": 1.2489,
+      "step": 1534
+    },
+    {
+      "epoch": 0.46064971115612574,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017763217352010846,
+      "loss": 1.3234,
+      "step": 1535
+    },
+    {
+      "epoch": 0.46094980868782354,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017761711101069437,
+      "loss": 1.2538,
+      "step": 1536
+    },
+    {
+      "epoch": 0.46124990621952133,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017760204850128031,
+      "loss": 1.3649,
+      "step": 1537
+    },
+    {
+      "epoch": 0.46155000375121913,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017758698599186626,
+      "loss": 1.3912,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4618501012829169,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.0001775719234824522,
+      "loss": 1.6035,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4621501988146148,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.0001775568609730381,
+      "loss": 1.6694,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4624502963463126,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017754179846362405,
+      "loss": 1.5529,
+      "step": 1541
+    },
+    {
+      "epoch": 0.46275039387801037,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017752673595421,
+      "loss": 1.6387,
+      "step": 1542
+    },
+    {
+      "epoch": 0.46305049140970816,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.0001775116734447959,
+      "loss": 1.6164,
+      "step": 1543
+    },
+    {
+      "epoch": 0.46335058894140596,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017749661093538184,
+      "loss": 1.5704,
+      "step": 1544
+    },
+    {
+      "epoch": 0.46365068647310376,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017748154842596778,
+      "loss": 1.4121,
+      "step": 1545
+    },
+    {
+      "epoch": 0.46395078400480155,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001774664859165537,
+      "loss": 1.4327,
+      "step": 1546
+    },
+    {
+      "epoch": 0.46425088153649935,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017745142340713964,
+      "loss": 1.5116,
+      "step": 1547
+    },
+    {
+      "epoch": 0.46455097906819715,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017743636089772558,
+      "loss": 1.4245,
+      "step": 1548
+    },
+    {
+      "epoch": 0.46485107659989494,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001774212983883115,
+      "loss": 1.258,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4651511741315928,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017740623587889743,
+      "loss": 1.5888,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4654512716632906,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017739117336948337,
+      "loss": 1.286,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4657513691949884,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017737611086006928,
+      "loss": 1.4132,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4660514667266862,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017736104835065522,
+      "loss": 1.4601,
+      "step": 1553
+    },
+    {
+      "epoch": 0.466351564258384,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017734598584124116,
+      "loss": 1.4937,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4666516617900818,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017733092333182708,
+      "loss": 1.5062,
+      "step": 1555
+    },
+    {
+      "epoch": 0.46695175932177957,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017731586082241302,
+      "loss": 1.5911,
+      "step": 1556
+    },
+    {
+      "epoch": 0.46725185685347737,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017730079831299896,
+      "loss": 1.3095,
+      "step": 1557
+    },
+    {
+      "epoch": 0.46755195438517516,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001772857358035849,
+      "loss": 1.3644,
+      "step": 1558
+    },
+    {
+      "epoch": 0.46785205191687296,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001772706732941708,
+      "loss": 1.4329,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4681521494485708,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017725561078475675,
+      "loss": 1.4017,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4684522469802686,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001772405482753427,
+      "loss": 1.3681,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4687523445119664,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001772254857659286,
+      "loss": 1.7392,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4690524420436642,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017721042325651455,
+      "loss": 1.5023,
+      "step": 1563
+    },
+    {
+      "epoch": 0.469352539575362,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017719536074710049,
+      "loss": 1.5186,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4696526371070598,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001771802982376864,
+      "loss": 1.3802,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4699527346387576,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017716523572827234,
+      "loss": 1.252,
+      "step": 1566
+    },
+    {
+      "epoch": 0.4702528321704554,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017715017321885828,
+      "loss": 1.4893,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4705529297021532,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001771351107094442,
+      "loss": 1.4714,
+      "step": 1568
+    },
+    {
+      "epoch": 0.47085302723385103,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017712004820003013,
+      "loss": 1.3368,
+      "step": 1569
+    },
+    {
+      "epoch": 0.4711531247655488,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017710498569061607,
+      "loss": 1.3873,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4714532222972466,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.000177089923181202,
+      "loss": 1.6891,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4717533198289444,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017707486067178793,
+      "loss": 1.2792,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4720534173606422,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017705979816237387,
+      "loss": 1.4133,
+      "step": 1573
+    },
+    {
+      "epoch": 0.47235351489234,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017704473565295978,
+      "loss": 1.551,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4726536124240378,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017702967314354572,
+      "loss": 1.3599,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4729537099557356,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017701461063413166,
+      "loss": 1.3987,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4732538074874334,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001769995481247176,
+      "loss": 1.579,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4735539050191312,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017698448561530351,
+      "loss": 1.4576,
+      "step": 1578
+    },
+    {
+      "epoch": 0.47385400255082905,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017696942310588946,
+      "loss": 1.4424,
+      "step": 1579
+    },
+    {
+      "epoch": 0.47415410008252684,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001769543605964754,
+      "loss": 1.385,
+      "step": 1580
+    },
+    {
+      "epoch": 0.47445419761422464,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001769392980870613,
+      "loss": 1.3461,
+      "step": 1581
+    },
+    {
+      "epoch": 0.47475429514592243,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017692423557764722,
+      "loss": 1.494,
+      "step": 1582
+    },
+    {
+      "epoch": 0.47505439267762023,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001769091730682332,
+      "loss": 1.3562,
+      "step": 1583
+    },
+    {
+      "epoch": 0.475354490209318,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001768941105588191,
+      "loss": 1.3217,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4756545877410158,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017687904804940504,
+      "loss": 1.4101,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4759546852727136,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017686398553999098,
+      "loss": 1.4218,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4762547828044114,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001768489230305769,
+      "loss": 1.4415,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4765548803361092,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017683386052116284,
+      "loss": 1.2545,
+      "step": 1588
+    },
+    {
+      "epoch": 0.47685497786780706,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017681879801174878,
+      "loss": 1.4279,
+      "step": 1589
+    },
+    {
+      "epoch": 0.47715507539950486,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001768037355023347,
+      "loss": 1.2719,
+      "step": 1590
+    },
+    {
+      "epoch": 0.47745517293120265,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017678867299292063,
+      "loss": 1.3807,
+      "step": 1591
+    },
+    {
+      "epoch": 0.47775527046290045,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017677361048350657,
+      "loss": 1.6153,
+      "step": 1592
+    },
+    {
+      "epoch": 0.47805536799459825,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001767585479740925,
+      "loss": 1.3972,
+      "step": 1593
+    },
+    {
+      "epoch": 0.47835546552629604,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017674348546467842,
+      "loss": 1.3494,
+      "step": 1594
+    },
+    {
+      "epoch": 0.47865556305799384,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017672842295526434,
+      "loss": 1.1675,
+      "step": 1595
+    },
+    {
+      "epoch": 0.47895566058969163,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001767133604458503,
+      "loss": 1.5152,
+      "step": 1596
+    },
+    {
+      "epoch": 0.47925575812138943,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00017669829793643622,
+      "loss": 1.4653,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4795558556530872,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017668323542702213,
+      "loss": 1.5217,
+      "step": 1598
+    },
+    {
+      "epoch": 0.4798559531847851,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001766681729176081,
+      "loss": 1.4362,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4801560507164829,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.000176653110408194,
+      "loss": 1.4197,
+      "step": 1600
+    },
+    {
+      "epoch": 0.48045614824818067,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017663804789877993,
+      "loss": 1.3259,
+      "step": 1601
+    },
+    {
+      "epoch": 0.48075624577987847,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001766229853893659,
+      "loss": 1.366,
+      "step": 1602
+    },
+    {
+      "epoch": 0.48105634331157626,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001766079228799518,
+      "loss": 1.3749,
+      "step": 1603
+    },
+    {
+      "epoch": 0.48135644084327406,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017659286037053775,
+      "loss": 1.5157,
+      "step": 1604
+    },
+    {
+      "epoch": 0.48165653837497185,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017657779786112366,
+      "loss": 1.4252,
+      "step": 1605
+    },
+    {
+      "epoch": 0.48195663590666965,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001765627353517096,
+      "loss": 1.6927,
+      "step": 1606
+    },
+    {
+      "epoch": 0.48225673343836745,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017654767284229554,
+      "loss": 1.4589,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4825568309700653,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017653261033288145,
+      "loss": 1.4061,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4828569285017631,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001765175478234674,
+      "loss": 1.2426,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4831570260334609,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017650248531405333,
+      "loss": 1.5331,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4834571235651587,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017648742280463925,
+      "loss": 1.2654,
+      "step": 1611
+    },
+    {
+      "epoch": 0.4837572210968565,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017647236029522521,
+      "loss": 1.5413,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4840573186285543,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017645729778581113,
+      "loss": 1.4424,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4843574161602521,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017644223527639704,
+      "loss": 1.2362,
+      "step": 1614
+    },
+    {
+      "epoch": 0.48465751369194987,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.000176427172766983,
+      "loss": 1.5662,
+      "step": 1615
+    },
+    {
+      "epoch": 0.48495761122364767,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017641211025756892,
+      "loss": 1.3734,
+      "step": 1616
+    },
+    {
+      "epoch": 0.48525770875534546,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017639704774815483,
+      "loss": 1.3632,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4855578062870433,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017638198523874077,
+      "loss": 1.375,
+      "step": 1618
+    },
+    {
+      "epoch": 0.4858579038187411,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017636692272932672,
+      "loss": 1.5785,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4861580013504389,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017635186021991266,
+      "loss": 1.4296,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4864580988821367,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017633679771049857,
+      "loss": 1.5416,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4867581964138345,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001763217352010845,
+      "loss": 1.4066,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4870582939455323,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017630667269167045,
+      "loss": 1.4252,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4873583914772301,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017629161018225636,
+      "loss": 1.2378,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4876584890089279,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001762765476728423,
+      "loss": 1.3684,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4879585865406257,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00017626148516342824,
+      "loss": 1.4344,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4882586840723235,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017624642265401416,
+      "loss": 1.2898,
+      "step": 1627
+    },
+    {
+      "epoch": 0.48855878160402133,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001762313601446001,
+      "loss": 1.3503,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4888588791357191,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017621629763518604,
+      "loss": 1.484,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4891589766674169,
+      "grad_norm": 3.984375,
+      "learning_rate": 0.00017620123512577195,
+      "loss": 1.453,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4894590741991147,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001761861726163579,
+      "loss": 1.3146,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4897591717308125,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017617111010694383,
+      "loss": 1.438,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4900592692625103,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017615604759752974,
+      "loss": 1.4568,
+      "step": 1633
+    },
+    {
+      "epoch": 0.4903593667942081,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017614098508811568,
+      "loss": 1.4505,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4906594643259059,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017612592257870162,
+      "loss": 1.4656,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4909595618576037,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017611086006928754,
+      "loss": 1.337,
+      "step": 1636
+    },
+    {
+      "epoch": 0.49125965938930155,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017609579755987348,
+      "loss": 1.3603,
+      "step": 1637
+    },
+    {
+      "epoch": 0.49155975692099935,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017608073505045942,
+      "loss": 1.4979,
+      "step": 1638
+    },
+    {
+      "epoch": 0.49185985445269714,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017606567254104536,
+      "loss": 1.4537,
+      "step": 1639
+    },
+    {
+      "epoch": 0.49215995198439494,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017605061003163127,
+      "loss": 1.4514,
+      "step": 1640
+    },
+    {
+      "epoch": 0.49246004951609273,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001760355475222172,
+      "loss": 1.4254,
+      "step": 1641
+    },
+    {
+      "epoch": 0.49276014704779053,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017602048501280315,
+      "loss": 1.6551,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4930602445794883,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017600542250338907,
+      "loss": 1.463,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4933603421111861,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.000175990359993975,
+      "loss": 1.5136,
+      "step": 1644
+    },
+    {
+      "epoch": 0.4936604396428839,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00017597529748456095,
+      "loss": 1.6423,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4939605371745817,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017596023497514686,
+      "loss": 1.3826,
+      "step": 1646
+    },
+    {
+      "epoch": 0.49426063470627957,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001759451724657328,
+      "loss": 1.4349,
+      "step": 1647
+    },
+    {
+      "epoch": 0.49456073223797736,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017593010995631874,
+      "loss": 1.255,
+      "step": 1648
+    },
+    {
+      "epoch": 0.49486082976967516,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017591504744690465,
+      "loss": 1.5038,
+      "step": 1649
+    },
+    {
+      "epoch": 0.49516092730137296,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001758999849374906,
+      "loss": 1.415,
+      "step": 1650
+    },
+    {
+      "epoch": 0.49546102483307075,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017588492242807653,
+      "loss": 1.3623,
+      "step": 1651
+    },
+    {
+      "epoch": 0.49576112236476855,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017586985991866245,
+      "loss": 1.4155,
+      "step": 1652
+    },
+    {
+      "epoch": 0.49606121989646634,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001758547974092484,
+      "loss": 1.6317,
+      "step": 1653
+    },
+    {
+      "epoch": 0.49636131742816414,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017583973489983433,
+      "loss": 1.55,
+      "step": 1654
+    },
+    {
+      "epoch": 0.49666141495986194,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017582467239042024,
+      "loss": 1.5341,
+      "step": 1655
+    },
+    {
+      "epoch": 0.49696151249155973,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017580960988100618,
+      "loss": 1.3299,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4972616100232576,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017579454737159212,
+      "loss": 1.4279,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4975617075549554,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017577948486217806,
+      "loss": 1.3859,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4978618050866532,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017576442235276397,
+      "loss": 1.3164,
+      "step": 1659
+    },
+    {
+      "epoch": 0.49816190261835097,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017574935984334992,
+      "loss": 1.4908,
+      "step": 1660
+    },
+    {
+      "epoch": 0.49846200015004877,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017573429733393586,
+      "loss": 1.5873,
+      "step": 1661
+    },
+    {
+      "epoch": 0.49876209768174656,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017571923482452177,
+      "loss": 1.5183,
+      "step": 1662
+    },
+    {
+      "epoch": 0.49906219521344436,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001757041723151077,
+      "loss": 1.3758,
+      "step": 1663
+    },
+    {
+      "epoch": 0.49936229274514216,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017568910980569365,
+      "loss": 1.4422,
+      "step": 1664
+    },
+    {
+      "epoch": 0.49966239027683995,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017567404729627956,
+      "loss": 1.4643,
+      "step": 1665
+    },
+    {
+      "epoch": 0.49996248780853775,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.0001756589847868655,
+      "loss": 1.4455,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5002625853402356,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017564392227745144,
+      "loss": 1.4651,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5005626828719334,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017562885976803736,
+      "loss": 1.3627,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5008627804036312,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001756137972586233,
+      "loss": 1.3963,
+      "step": 1669
+    },
+    {
+      "epoch": 0.501162877935329,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017559873474920924,
+      "loss": 1.4744,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5014629754670268,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.00017558367223979515,
+      "loss": 1.8055,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5017630729987246,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001755686097303811,
+      "loss": 1.4459,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5020631705304224,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017555354722096703,
+      "loss": 1.6567,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5023632680621202,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017553848471155297,
+      "loss": 1.5646,
+      "step": 1674
+    },
+    {
+      "epoch": 0.502663365593818,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017552342220213888,
+      "loss": 1.5302,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5029634631255158,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001755083596927248,
+      "loss": 1.2843,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5032635606572136,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017549329718331076,
+      "loss": 1.6471,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5035636581889114,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017547823467389668,
+      "loss": 1.3244,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5038637557206092,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001754631721644826,
+      "loss": 1.4017,
+      "step": 1679
+    },
+    {
+      "epoch": 0.504163853252307,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017544810965506856,
+      "loss": 1.4422,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5044639507840049,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00017543304714565447,
+      "loss": 1.5523,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5047640483157027,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00017541798463624039,
+      "loss": 1.4661,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5050641458474004,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00017540292212682635,
+      "loss": 1.6074,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5053642433790982,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00017538785961741227,
+      "loss": 1.6245,
+      "step": 1684
+    },
+    {
+      "epoch": 0.505664340910796,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001753727971079982,
+      "loss": 1.4143,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5059644384424938,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017535773459858415,
+      "loss": 1.2795,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5062645359741916,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017534267208917006,
+      "loss": 1.3119,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5065646335058894,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.000175327609579756,
+      "loss": 1.4655,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5068647310375872,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001753125470703419,
+      "loss": 1.3434,
+      "step": 1689
+    },
+    {
+      "epoch": 0.507164828569285,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017529748456092785,
+      "loss": 1.5543,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5074649261009828,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001752824220515138,
+      "loss": 1.4226,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5077650236326806,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.0001752673595420997,
+      "loss": 1.498,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5080651211643784,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00017525229703268567,
+      "loss": 1.6815,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5083652186960762,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.0001752372345232716,
+      "loss": 1.3535,
+      "step": 1694
+    },
+    {
+      "epoch": 0.508665316227774,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.0001752221720138575,
+      "loss": 1.5217,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5089654137594718,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017520710950444347,
+      "loss": 1.3794,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5092655112911696,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017519204699502938,
+      "loss": 1.3182,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5095656088228674,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001751769844856153,
+      "loss": 1.5142,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5098657063545652,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017516192197620126,
+      "loss": 1.4175,
+      "step": 1699
+    },
+    {
+      "epoch": 0.510165803886263,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017514685946678718,
+      "loss": 1.3963,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5104659014179609,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017513179695737312,
+      "loss": 1.403,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5107659989496587,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017511673444795903,
+      "loss": 1.3018,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5110660964813565,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017510167193854497,
+      "loss": 1.3394,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5113661940130543,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001750866094291309,
+      "loss": 1.5439,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5116662915447521,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017507154691971682,
+      "loss": 1.4285,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5119663890764499,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017505648441030276,
+      "loss": 1.3022,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5122664866081477,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001750414219008887,
+      "loss": 1.4812,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5125665841398455,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017502635939147462,
+      "loss": 1.5397,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5128666816715433,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017501129688206056,
+      "loss": 1.3874,
+      "step": 1709
+    },
+    {
+      "epoch": 0.513166779203241,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001749962343726465,
+      "loss": 1.4127,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5134668767349388,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001749811718632324,
+      "loss": 1.4313,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5137669742666366,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017496610935381835,
+      "loss": 1.2639,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5140670717983344,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001749510468444043,
+      "loss": 1.4308,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5143671693300322,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001749359843349902,
+      "loss": 1.4168,
+      "step": 1714
+    },
+    {
+      "epoch": 0.51466726686173,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017492092182557614,
+      "loss": 1.4266,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5149673643934278,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017490585931616208,
+      "loss": 1.4323,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5152674619251256,
+      "grad_norm": 1.375,
+      "learning_rate": 0.000174890796806748,
+      "loss": 1.5957,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5155675594568234,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017487573429733394,
+      "loss": 1.3062,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5158676569885212,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017486067178791988,
+      "loss": 1.4062,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5161677545202191,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017484560927850582,
+      "loss": 1.4026,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5164678520519169,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017483054676909173,
+      "loss": 1.4653,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5167679495836147,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00017481548425967767,
+      "loss": 1.4186,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5170680471153125,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001748004217502636,
+      "loss": 1.5322,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5173681446470103,
+      "grad_norm": 3.28125,
+      "learning_rate": 0.00017478535924084953,
+      "loss": 1.4276,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5176682421787081,
+      "grad_norm": 1.6953125,
+      "learning_rate": 0.00017477029673143547,
+      "loss": 1.5628,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5179683397104059,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001747552342220214,
+      "loss": 1.4406,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5182684372421037,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017474017171260732,
+      "loss": 1.3759,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5185685347738015,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017472510920319326,
+      "loss": 1.5824,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5188686323054993,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001747100466937792,
+      "loss": 1.517,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5191687298371971,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001746949841843651,
+      "loss": 1.3128,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5194688273688949,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017467992167495105,
+      "loss": 1.519,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5197689249005927,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.000174664859165537,
+      "loss": 1.4052,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5200690224322905,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001746497966561229,
+      "loss": 1.5532,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5203691199639883,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017463473414670885,
+      "loss": 1.4392,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5206692174956861,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001746196716372948,
+      "loss": 1.4521,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5209693150273839,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001746046091278807,
+      "loss": 1.3003,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5212694125590817,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017458954661846664,
+      "loss": 1.422,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5215695100907795,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017457448410905258,
+      "loss": 1.4514,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5218696076224774,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017455942159963852,
+      "loss": 1.5289,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5221697051541752,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00017454435909022443,
+      "loss": 1.5201,
+      "step": 1740
+    },
+    {
+      "epoch": 0.522469802685873,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017452929658081038,
+      "loss": 1.4065,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5227699002175707,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00017451423407139632,
+      "loss": 1.416,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5230699977492685,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017449917156198223,
+      "loss": 1.322,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5233700952809663,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017448410905256817,
+      "loss": 1.2824,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5236701928126641,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001744690465431541,
+      "loss": 1.6403,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5239702903443619,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017445398403374002,
+      "loss": 1.2771,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5242703878760597,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017443892152432596,
+      "loss": 1.6098,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5245704854077575,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001744238590149119,
+      "loss": 1.3361,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5248705829394553,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017440879650549782,
+      "loss": 1.3638,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5251706804711531,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017439373399608376,
+      "loss": 1.185,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5254707780028509,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001743786714866697,
+      "loss": 1.5022,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5257708755345487,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001743636089772556,
+      "loss": 1.4268,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5260709730662465,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00017434854646784155,
+      "loss": 1.5076,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5263710705979443,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001743334839584275,
+      "loss": 1.3569,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5266711681296421,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017431842144901343,
+      "loss": 1.3874,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5269712656613399,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017430335893959934,
+      "loss": 1.4339,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5272713631930377,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017428829643018528,
+      "loss": 1.3524,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5275714607247355,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00017427323392077122,
+      "loss": 1.5504,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5278715582564334,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00017425817141135714,
+      "loss": 1.4165,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5281716557881312,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017424310890194305,
+      "loss": 1.4148,
+      "step": 1760
+    },
+    {
+      "epoch": 0.528471753319829,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017422804639252902,
+      "loss": 1.4283,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5287718508515268,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017421298388311493,
+      "loss": 1.3939,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5290719483832246,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017419792137370085,
+      "loss": 1.4829,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5293720459149224,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001741828588642868,
+      "loss": 1.2617,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5296721434466202,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017416779635487273,
+      "loss": 1.4137,
+      "step": 1765
+    },
+    {
+      "epoch": 0.529972240978318,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017415273384545867,
+      "loss": 1.3688,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5302723385100158,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001741376713360446,
+      "loss": 1.5503,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5305724360417136,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017412260882663052,
+      "loss": 1.5974,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5308725335734114,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00017410754631721646,
+      "loss": 1.4101,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5311726311051091,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.0001740924838078024,
+      "loss": 1.4584,
+      "step": 1770
+    },
+    {
+      "epoch": 0.531472728636807,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.0001740774212983883,
+      "loss": 1.339,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5317728261685047,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017406235878897425,
+      "loss": 1.3225,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5320729237002025,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017404729627956017,
+      "loss": 1.4809,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5323730212319003,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017403223377014613,
+      "loss": 1.426,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5326731187635981,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017401717126073205,
+      "loss": 1.5084,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5329732162952959,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017400210875131796,
+      "loss": 1.3427,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5332733138269937,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017398704624190393,
+      "loss": 1.4403,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5335734113586916,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017397198373248984,
+      "loss": 1.3289,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5338735088903894,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017395692122307575,
+      "loss": 1.3675,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5341736064220872,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00017394185871366172,
+      "loss": 1.3883,
+      "step": 1780
+    },
+    {
+      "epoch": 0.534473703953785,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017392679620424764,
+      "loss": 1.2456,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5347738014854828,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017391173369483358,
+      "loss": 1.4854,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5350738990171806,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001738966711854195,
+      "loss": 1.4006,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5353739965488784,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00017388160867600543,
+      "loss": 1.3797,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5356740940805762,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017386654616659137,
+      "loss": 1.6834,
+      "step": 1785
+    },
+    {
+      "epoch": 0.535974191612274,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017385148365717728,
+      "loss": 1.3221,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5362742891439718,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017383642114776322,
+      "loss": 1.2381,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5365743866756696,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017382135863834916,
+      "loss": 1.3963,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5368744842073674,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017380629612893508,
+      "loss": 1.4378,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5371745817390652,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017379123361952102,
+      "loss": 1.3576,
+      "step": 1790
+    },
+    {
+      "epoch": 0.537474679270763,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017377617111010696,
+      "loss": 1.4937,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5377747768024608,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00017376110860069287,
+      "loss": 1.5982,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5380748743341586,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00017374604609127884,
+      "loss": 1.5881,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5383749718658564,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00017373098358186475,
+      "loss": 1.3502,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5386750693975542,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017371592107245066,
+      "loss": 1.3344,
+      "step": 1795
+    },
+    {
+      "epoch": 0.538975166929252,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001737008585630366,
+      "loss": 1.4586,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5392752644609498,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00017368579605362254,
+      "loss": 1.4361,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5395753619926477,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017367073354420846,
+      "loss": 1.4,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5398754595243455,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001736556710347944,
+      "loss": 1.3734,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5401755570560433,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017364060852538034,
+      "loss": 1.4759,
+      "step": 1800
+    },
+    {
+      "epoch": 0.540475654587741,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00017362554601596628,
+      "loss": 1.4379,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5407757521194388,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001736104835065522,
+      "loss": 1.6134,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5410758496511366,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017359542099713813,
+      "loss": 1.469,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5413759471828344,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00017358035848772407,
+      "loss": 1.5022,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5416760447145322,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00017356529597830999,
+      "loss": 1.3711,
+      "step": 1805
+    },
+    {
+      "epoch": 0.54197614224623,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017355023346889593,
+      "loss": 1.5012,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5422762397779278,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017353517095948187,
+      "loss": 1.3672,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5425763373096256,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017352010845006778,
+      "loss": 1.4671,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5428764348413234,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017350504594065372,
+      "loss": 1.379,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5431765323730212,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017348998343123966,
+      "loss": 1.3674,
+      "step": 1810
+    },
+    {
+      "epoch": 0.543476629904719,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017347492092182557,
+      "loss": 1.4918,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5437767274364168,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001734598584124115,
+      "loss": 1.4317,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5440768249681146,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017344479590299745,
+      "loss": 1.3553,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5443769224998124,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017342973339358337,
+      "loss": 1.4829,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5446770200315102,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001734146708841693,
+      "loss": 1.2717,
+      "step": 1815
+    },
+    {
+      "epoch": 0.544977117563208,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017339960837475525,
+      "loss": 1.4205,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5452772150949059,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017338454586534116,
+      "loss": 1.5072,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5455773126266037,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001733694833559271,
+      "loss": 1.4829,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5458774101583015,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017335442084651304,
+      "loss": 1.5063,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5461775076899993,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017333935833709898,
+      "loss": 1.5292,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5464776052216971,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001733242958276849,
+      "loss": 1.2802,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5467777027533949,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017330923331827084,
+      "loss": 1.3621,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5470778002850927,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017329417080885678,
+      "loss": 1.6059,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5473778978167905,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001732791082994427,
+      "loss": 1.5676,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5476779953484883,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017326404579002863,
+      "loss": 1.3093,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5479780928801861,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017324898328061457,
+      "loss": 1.4113,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5482781904118839,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017323392077120048,
+      "loss": 1.586,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5485782879435817,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017321885826178642,
+      "loss": 1.4331,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5488783854752795,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017320379575237236,
+      "loss": 1.33,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5491784830069772,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00017318873324295828,
+      "loss": 1.5194,
+      "step": 1830
+    },
+    {
+      "epoch": 0.549478580538675,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017317367073354422,
+      "loss": 1.5401,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5497786780703728,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017315860822413016,
+      "loss": 1.5781,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5500787756020706,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017314354571471607,
+      "loss": 1.3445,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5503788731337684,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.000173128483205302,
+      "loss": 1.4662,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5506789706654662,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017311342069588795,
+      "loss": 1.5719,
+      "step": 1835
+    },
+    {
+      "epoch": 0.550979068197164,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.0001730983581864739,
+      "loss": 1.4106,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5512791657288619,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001730832956770598,
+      "loss": 1.3589,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5515792632605597,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017306823316764574,
+      "loss": 1.3613,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5518793607922575,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017305317065823168,
+      "loss": 1.3914,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5521794583239553,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001730381081488176,
+      "loss": 1.4323,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5524795558556531,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001730230456394035,
+      "loss": 1.4085,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5527796533873509,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017300798312998948,
+      "loss": 1.3107,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5530797509190487,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001729929206205754,
+      "loss": 1.3548,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5533798484507465,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001729778581111613,
+      "loss": 1.2609,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5536799459824443,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017296279560174727,
+      "loss": 1.3064,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5539800435141421,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017294773309233319,
+      "loss": 1.3058,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5542801410458399,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017293267058291913,
+      "loss": 1.3936,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5545802385775377,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017291760807350507,
+      "loss": 1.4802,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5548803361092355,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017290254556409098,
+      "loss": 1.5111,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5551804336409333,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00017288748305467692,
+      "loss": 1.557,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5554805311726311,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017287242054526286,
+      "loss": 1.352,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5557806287043289,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017285735803584877,
+      "loss": 1.3129,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5560807262360267,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00017284229552643471,
+      "loss": 1.3773,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5563808237677245,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017282723301702063,
+      "loss": 1.5773,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5566809212994223,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.0001728121705076066,
+      "loss": 1.4381,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5569810188311202,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001727971079981925,
+      "loss": 1.2687,
+      "step": 1856
+    },
+    {
+      "epoch": 0.557281116362818,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017278204548877842,
+      "loss": 1.2856,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5575812138945158,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.0001727669829793644,
+      "loss": 1.5459,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5578813114262136,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001727519204699503,
+      "loss": 1.359,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5581814089579114,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00017273685796053621,
+      "loss": 1.2776,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5584815064896091,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017272179545112218,
+      "loss": 1.5734,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5587816040213069,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001727067329417081,
+      "loss": 1.3058,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5590817015530047,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017269167043229404,
+      "loss": 1.3721,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5593817990847025,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017267660792287998,
+      "loss": 1.5547,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5596818966164003,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001726615454134659,
+      "loss": 1.3633,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5599819941480981,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00017264648290405183,
+      "loss": 1.4625,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5602820916797959,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00017263142039463774,
+      "loss": 1.3944,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5605821892114937,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017261635788522368,
+      "loss": 1.5079,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5608822867431915,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017260129537580962,
+      "loss": 1.3981,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5611823842748893,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017258623286639554,
+      "loss": 1.3212,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5614824818065871,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017257117035698148,
+      "loss": 1.4912,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5617825793382849,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017255610784756742,
+      "loss": 1.6412,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5620826768699827,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017254104533815333,
+      "loss": 1.3439,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5623827744016805,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001725259828287393,
+      "loss": 1.4523,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5626828719333784,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001725109203193252,
+      "loss": 1.4729,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5629829694650762,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017249585780991112,
+      "loss": 1.2273,
+      "step": 1876
+    },
+    {
+      "epoch": 0.563283066996774,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001724807953004971,
+      "loss": 1.5048,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5635831645284718,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.000172465732791083,
+      "loss": 1.2854,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5638832620601696,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017245067028166892,
+      "loss": 1.5368,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5641833595918674,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017243560777225486,
+      "loss": 1.3044,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5644834571235652,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001724205452628408,
+      "loss": 1.4905,
+      "step": 1881
+    },
+    {
+      "epoch": 0.564783554655263,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017240548275342674,
+      "loss": 1.4784,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5650836521869608,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017239042024401265,
+      "loss": 1.1709,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5653837497186586,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001723753577345986,
+      "loss": 1.4377,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5656838472503564,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00017236029522518453,
+      "loss": 1.5231,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5659839447820542,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017234523271577045,
+      "loss": 1.4181,
+      "step": 1886
+    },
+    {
+      "epoch": 0.566284042313752,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017233017020635639,
+      "loss": 1.4176,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5665841398454498,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00017231510769694233,
+      "loss": 1.4735,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5668842373771475,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017230004518752824,
+      "loss": 1.5599,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5671843349088453,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00017228498267811418,
+      "loss": 1.5124,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5674844324405431,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017226992016870012,
+      "loss": 1.3677,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5677845299722409,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017225485765928603,
+      "loss": 1.3993,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5680846275039387,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017223979514987197,
+      "loss": 1.3791,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5683847250356365,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017222473264045791,
+      "loss": 1.4144,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5686848225673344,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017220967013104383,
+      "loss": 1.4976,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5689849200990322,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00017219460762162977,
+      "loss": 1.2837,
+      "step": 1896
+    },
+    {
+      "epoch": 0.56928501763073,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001721795451122157,
+      "loss": 1.3429,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5695851151624278,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017216448260280162,
+      "loss": 1.4205,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5698852126941256,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017214942009338756,
+      "loss": 1.3941,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5701853102258234,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001721343575839735,
+      "loss": 1.3525,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5704854077575212,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017211929507455944,
+      "loss": 1.3921,
+      "step": 1901
+    },
+    {
+      "epoch": 0.570785505289219,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017210423256514535,
+      "loss": 1.3309,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5710856028209168,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001720891700557313,
+      "loss": 1.3027,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5713857003526146,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00017207410754631724,
+      "loss": 1.3913,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5716857978843124,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017205904503690315,
+      "loss": 1.2913,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5719858954160102,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001720439825274891,
+      "loss": 1.2603,
+      "step": 1906
+    },
+    {
+      "epoch": 0.572285992947708,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017202892001807503,
+      "loss": 1.3933,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5725860904794058,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017201385750866094,
+      "loss": 1.3859,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5728861880111036,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00017199879499924688,
+      "loss": 1.3549,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5731862855428014,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017198373248983282,
+      "loss": 1.5665,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5734863830744992,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017196866998041874,
+      "loss": 1.5156,
+      "step": 1911
+    },
+    {
+      "epoch": 0.573786480606197,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017195360747100468,
+      "loss": 1.3828,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5740865781378948,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017193854496159062,
+      "loss": 1.4218,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5743866756695927,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017192348245217653,
+      "loss": 1.419,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5746867732012905,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017190841994276247,
+      "loss": 1.4382,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5749868707329883,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001718933574333484,
+      "loss": 1.5044,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5752869682646861,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017187829492393435,
+      "loss": 1.2938,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5755870657963839,
+      "grad_norm": 3.453125,
+      "learning_rate": 0.00017186323241452026,
+      "loss": 1.6619,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5758871633280817,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001718481699051062,
+      "loss": 1.6052,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5761872608597794,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00017183310739569214,
+      "loss": 1.4542,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5764873583914772,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017181804488627806,
+      "loss": 1.3482,
+      "step": 1921
+    },
+    {
+      "epoch": 0.576787455923175,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.000171802982376864,
+      "loss": 1.4744,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5770875534548728,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017178791986744994,
+      "loss": 1.4087,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5773876509865706,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017177285735803585,
+      "loss": 1.387,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5776877485182684,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017175779484862177,
+      "loss": 1.3048,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5779878460499662,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00017174273233920773,
+      "loss": 1.5437,
+      "step": 1926
+    },
+    {
+      "epoch": 0.578287943581664,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017172766982979365,
+      "loss": 1.2811,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5785880411133618,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00017171260732037959,
+      "loss": 1.6206,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5788881386450596,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017169754481096553,
+      "loss": 1.5305,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5791882361767574,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017168248230155144,
+      "loss": 1.4503,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5794883337084552,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017166741979213738,
+      "loss": 1.441,
+      "step": 1931
+    },
+    {
+      "epoch": 0.579788431240153,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017165235728272332,
+      "loss": 1.3836,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5800885287718508,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017163729477330923,
+      "loss": 1.4114,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5803886263035487,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00017162223226389517,
+      "loss": 1.6224,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5806887238352465,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017160716975448111,
+      "loss": 1.4272,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5809888213669443,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00017159210724506705,
+      "loss": 1.2714,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5812889188986421,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017157704473565297,
+      "loss": 1.3343,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5815890164303399,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00017156198222623888,
+      "loss": 1.4505,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5818891139620377,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017154691971682485,
+      "loss": 1.4835,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5821892114937355,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017153185720741076,
+      "loss": 1.3751,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5824893090254333,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017151679469799667,
+      "loss": 1.3615,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5827894065571311,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017150173218858264,
+      "loss": 1.3266,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5830895040888289,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017148666967916856,
+      "loss": 1.2912,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5833896016205267,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001714716071697545,
+      "loss": 1.3653,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5836896991522245,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017145654466034044,
+      "loss": 1.3381,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5839897966839223,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017144148215092635,
+      "loss": 1.4986,
+      "step": 1946
+    },
+    {
+      "epoch": 0.58428989421562,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001714264196415123,
+      "loss": 1.5132,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5845899917473178,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001714113571320982,
+      "loss": 1.3561,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5848900892790156,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017139629462268414,
+      "loss": 1.2613,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5851901868107134,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017138123211327008,
+      "loss": 1.415,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5854902843424112,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.000171366169603856,
+      "loss": 1.4213,
+      "step": 1951
+    },
+    {
+      "epoch": 0.585790381874109,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017135110709444194,
+      "loss": 1.3143,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5860904794058069,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00017133604458502788,
+      "loss": 1.2468,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5863905769375047,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001713209820756138,
+      "loss": 1.361,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5866906744692025,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017130591956619976,
+      "loss": 1.2898,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5869907720009003,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00017129085705678567,
+      "loss": 1.5257,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5872908695325981,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017127579454737158,
+      "loss": 1.465,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5875909670642959,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017126073203795755,
+      "loss": 1.6703,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5878910645959937,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00017124566952854346,
+      "loss": 1.493,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5881911621276915,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017123060701912938,
+      "loss": 1.3793,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5884912596593893,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017121554450971532,
+      "loss": 1.4132,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5887913571910871,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017120048200030126,
+      "loss": 1.3494,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5890914547227849,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001711854194908872,
+      "loss": 1.3878,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5893915522544827,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001711703569814731,
+      "loss": 1.4495,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5896916497861805,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017115529447205905,
+      "loss": 1.2517,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5899917473178783,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.000171140231962645,
+      "loss": 1.3871,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5902918448495761,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001711251694532309,
+      "loss": 1.4149,
+      "step": 1967
+    },
+    {
+      "epoch": 0.5905919423812739,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017111010694381685,
+      "loss": 1.406,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5908920399129717,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017109504443440279,
+      "loss": 1.3471,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5911921374446695,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001710799819249887,
+      "loss": 1.3088,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5914922349763673,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017106491941557467,
+      "loss": 1.3584,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5917923325080651,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017104985690616058,
+      "loss": 1.4751,
+      "step": 1972
+    },
+    {
+      "epoch": 0.592092430039763,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001710347943967465,
+      "loss": 1.1931,
+      "step": 1973
+    },
+    {
+      "epoch": 0.5923925275714608,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00017101973188733243,
+      "loss": 1.4128,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5926926251031586,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017100466937791837,
+      "loss": 1.3746,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5929927226348564,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001709896068685043,
+      "loss": 1.3378,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5932928201665542,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00017097454435909023,
+      "loss": 1.3517,
+      "step": 1977
+    },
+    {
+      "epoch": 0.593592917698252,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00017095948184967617,
+      "loss": 1.3099,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5938930152299498,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017094441934026208,
+      "loss": 1.3613,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5941931127616475,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017092935683084802,
+      "loss": 1.395,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5944932102933453,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00017091429432143396,
+      "loss": 1.1279,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5947933078250431,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001708992318120199,
+      "loss": 1.2798,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5950934053567409,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00017088416930260582,
+      "loss": 1.6579,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5953935028884387,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017086910679319176,
+      "loss": 1.4677,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5956936004201365,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001708540442837777,
+      "loss": 1.3478,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5959936979518343,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001708389817743636,
+      "loss": 1.3407,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5962937954835321,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00017082391926494955,
+      "loss": 1.6657,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5965938930152299,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001708088567555355,
+      "loss": 1.3768,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5968939905469277,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001707937942461214,
+      "loss": 1.3923,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5971940880786255,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017077873173670734,
+      "loss": 1.6085,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5974941856103233,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017076366922729328,
+      "loss": 1.4348,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5977942831420212,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001707486067178792,
+      "loss": 1.3991,
+      "step": 1992
+    },
+    {
+      "epoch": 0.598094380673719,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017073354420846514,
+      "loss": 1.4068,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5983944782054168,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017071848169905108,
+      "loss": 1.3129,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5986945757371146,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.000170703419189637,
+      "loss": 1.4585,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5989946732688124,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017068835668022293,
+      "loss": 1.5604,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5992947708005102,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017067329417080887,
+      "loss": 1.2656,
+      "step": 1997
+    },
+    {
+      "epoch": 0.599594868332208,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001706582316613948,
+      "loss": 1.2242,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5998949658639058,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00017064316915198072,
+      "loss": 1.4615,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6001950633956036,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017062810664256666,
+      "loss": 1.4079,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6004951609273014,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001706130441331526,
+      "loss": 1.401,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6007952584589992,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017059798162373852,
+      "loss": 1.2916,
+      "step": 2002
+    },
+    {
+      "epoch": 0.601095355990697,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00017058291911432446,
+      "loss": 1.4505,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6013954535223948,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001705678566049104,
+      "loss": 1.2513,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6016955510540926,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001705527940954963,
+      "loss": 1.4687,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6019956485857904,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017053773158608225,
+      "loss": 1.2582,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6022957461174882,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001705226690766682,
+      "loss": 1.3652,
+      "step": 2007
+    },
+    {
+      "epoch": 0.602595843649186,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001705076065672541,
+      "loss": 1.3325,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6028959411808837,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00017049254405784005,
+      "loss": 1.3567,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6031960387125815,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00017047748154842599,
+      "loss": 1.3234,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6034961362442793,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001704624190390119,
+      "loss": 1.493,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6037962337759772,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017044735652959784,
+      "loss": 1.5729,
+      "step": 2012
+    },
+    {
+      "epoch": 0.604096331307675,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017043229402018378,
+      "loss": 1.3537,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6043964288393728,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001704172315107697,
+      "loss": 1.2841,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6046965263710706,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00017040216900135563,
+      "loss": 1.6055,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6049966239027684,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00017038710649194157,
+      "loss": 1.5241,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6052967214344662,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017037204398252751,
+      "loss": 1.397,
+      "step": 2017
+    },
+    {
+      "epoch": 0.605596818966164,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00017035698147311343,
+      "loss": 1.3919,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6058969164978618,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00017034191896369934,
+      "loss": 1.3732,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6061970140295596,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001703268564542853,
+      "loss": 1.4192,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6064971115612574,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017031179394487122,
+      "loss": 1.4555,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6067972090929552,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00017029673143545713,
+      "loss": 1.4916,
+      "step": 2022
+    },
+    {
+      "epoch": 0.607097306624653,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001702816689260431,
+      "loss": 1.3684,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6073974041563508,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00017026660641662902,
+      "loss": 1.3847,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6076975016880486,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017025154390721496,
+      "loss": 1.2735,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6079975992197464,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001702364813978009,
+      "loss": 1.4429,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6082976967514442,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001702214188883868,
+      "loss": 1.3218,
+      "step": 2027
+    },
+    {
+      "epoch": 0.608597794283142,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017020635637897275,
+      "loss": 1.4759,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6088978918148398,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001701912938695587,
+      "loss": 1.2526,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6091979893465376,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001701762313601446,
+      "loss": 1.2939,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6094980868782355,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017016116885073054,
+      "loss": 1.315,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6097981844099333,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017014610634131646,
+      "loss": 1.3034,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6100982819416311,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001701310438319024,
+      "loss": 1.3837,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6103983794733289,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017011598132248834,
+      "loss": 1.2727,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6106984770050267,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00017010091881307425,
+      "loss": 1.1456,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6109985745367245,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017008585630366022,
+      "loss": 1.5611,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6112986720684223,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00017007079379424613,
+      "loss": 1.4972,
+      "step": 2037
+    },
+    {
+      "epoch": 0.61159876960012,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017005573128483204,
+      "loss": 1.3585,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6118988671318178,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.000170040668775418,
+      "loss": 1.3005,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6121989646635156,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00017002560626600392,
+      "loss": 1.4271,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6124990621952134,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017001054375658984,
+      "loss": 1.4276,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6127991597269112,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001699954812471758,
+      "loss": 1.1991,
+      "step": 2042
+    },
+    {
+      "epoch": 0.613099257258609,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016998041873776172,
+      "loss": 1.3941,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6133993547903068,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016996535622834766,
+      "loss": 1.3258,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6136994523220046,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016995029371893357,
+      "loss": 1.51,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6139995498537024,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001699352312095195,
+      "loss": 1.2953,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6142996473854002,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016992016870010545,
+      "loss": 1.3888,
+      "step": 2047
+    },
+    {
+      "epoch": 0.614599744917098,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016990510619069137,
+      "loss": 1.4715,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6148998424487958,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001698900436812773,
+      "loss": 1.352,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6151999399804937,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016987498117186325,
+      "loss": 1.3108,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6155000375121915,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016985991866244916,
+      "loss": 1.5483,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6158001350438893,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016984485615303513,
+      "loss": 1.3636,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6161002325755871,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016982979364362104,
+      "loss": 1.4585,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6164003301072849,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016981473113420695,
+      "loss": 1.3937,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6167004276389827,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001697996686247929,
+      "loss": 1.5205,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6170005251706805,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016978460611537883,
+      "loss": 1.5215,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6173006227023783,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016976954360596475,
+      "loss": 1.3466,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6176007202340761,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001697544810965507,
+      "loss": 1.3609,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6179008177657739,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016973941858713663,
+      "loss": 1.2542,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6182009152974717,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016972435607772254,
+      "loss": 1.3966,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6185010128291695,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016970929356830848,
+      "loss": 1.6225,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6188011103608673,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016969423105889442,
+      "loss": 1.2639,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6191012078925651,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016967916854948036,
+      "loss": 1.1531,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6194013054242629,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016966410604006628,
+      "loss": 1.426,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6197014029559607,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016964904353065222,
+      "loss": 1.3222,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6200015004876585,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00016963398102123816,
+      "loss": 1.5445,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6203015980193562,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016961891851182407,
+      "loss": 1.3555,
+      "step": 2067
+    },
+    {
+      "epoch": 0.620601695551054,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016960385600241,
+      "loss": 1.4551,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6209017930827518,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016958879349299595,
+      "loss": 1.3705,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6212018906144497,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00016957373098358186,
+      "loss": 1.4518,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6215019881461475,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001695586684741678,
+      "loss": 1.3668,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6218020856778453,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016954360596475374,
+      "loss": 1.3668,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6221021832095431,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016952854345533966,
+      "loss": 1.369,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6224022807412409,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001695134809459256,
+      "loss": 1.3567,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6227023782729387,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016949841843651154,
+      "loss": 1.2545,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6230024758046365,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016948335592709745,
+      "loss": 1.3604,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6233025733363343,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001694682934176834,
+      "loss": 1.4746,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6236026708680321,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016945323090826933,
+      "loss": 1.4639,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6239027683997299,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016943816839885527,
+      "loss": 1.3494,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6242028659314277,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016942310588944118,
+      "loss": 1.4447,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6245029634631255,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016940804338002712,
+      "loss": 1.6654,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6248030609948233,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016939298087061307,
+      "loss": 1.3087,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6251031585265211,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016937791836119898,
+      "loss": 1.382,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6254032560582189,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016936285585178492,
+      "loss": 1.2667,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6257033535899167,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016934779334237086,
+      "loss": 1.2419,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6260034511216145,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016933273083295677,
+      "loss": 1.3936,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6263035486533123,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001693176683235427,
+      "loss": 1.6209,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6266036461850101,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016930260581412865,
+      "loss": 1.2604,
+      "step": 2088
+    },
+    {
+      "epoch": 0.626903743716708,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016928754330471457,
+      "loss": 1.2737,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6272038412484058,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.0001692724807953005,
+      "loss": 1.326,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6275039387801036,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016925741828588645,
+      "loss": 1.2448,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6278040363118014,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00016924235577647236,
+      "loss": 1.6543,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6281041338434992,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001692272932670583,
+      "loss": 1.2636,
+      "step": 2093
+    },
+    {
+      "epoch": 0.628404231375197,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016921223075764424,
+      "loss": 1.2049,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6287043289068948,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00016919716824823015,
+      "loss": 1.4451,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6290044264385926,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001691821057388161,
+      "loss": 1.3878,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6293045239702904,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00016916704322940203,
+      "loss": 1.4188,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6296046215019881,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016915198071998797,
+      "loss": 1.3408,
+      "step": 2098
+    },
+    {
+      "epoch": 0.629904719033686,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001691369182105739,
+      "loss": 1.4261,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6302048165653837,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016912185570115983,
+      "loss": 1.246,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6305049140970815,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016910679319174577,
+      "loss": 1.4647,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6308050116287793,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016909173068233168,
+      "loss": 1.3262,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6311051091604771,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001690766681729176,
+      "loss": 1.2872,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6314052066921749,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016906160566350356,
+      "loss": 1.4019,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6317053042238727,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016904654315408948,
+      "loss": 1.3414,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6320054017555705,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016903148064467542,
+      "loss": 1.4664,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6323054992872683,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016901641813526136,
+      "loss": 1.3865,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6326055968189661,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016900135562584727,
+      "loss": 1.3692,
+      "step": 2108
+    },
+    {
+      "epoch": 0.632905694350664,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001689862931164332,
+      "loss": 1.2973,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6332057918823618,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016897123060701915,
+      "loss": 1.3764,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6335058894140596,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016895616809760506,
+      "loss": 1.3368,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6338059869457574,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.000168941105588191,
+      "loss": 1.5196,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6341060844774552,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016892604307877694,
+      "loss": 1.5256,
+      "step": 2113
+    },
+    {
+      "epoch": 0.634406182009153,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016891098056936286,
+      "loss": 1.2769,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6347062795408508,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001688959180599488,
+      "loss": 1.3793,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6350063770725486,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001688808555505347,
+      "loss": 1.3366,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6353064746042464,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016886579304112068,
+      "loss": 1.3324,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6356065721359442,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001688507305317066,
+      "loss": 1.424,
+      "step": 2118
+    },
+    {
+      "epoch": 0.635906669667642,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001688356680222925,
+      "loss": 1.4911,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6362067671993398,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016882060551287847,
+      "loss": 1.4226,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6365068647310376,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016880554300346438,
+      "loss": 1.3932,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6368069622627354,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001687904804940503,
+      "loss": 1.2991,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6371070597944332,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016877541798463627,
+      "loss": 1.2815,
+      "step": 2123
+    },
+    {
+      "epoch": 0.637407157326131,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016876035547522218,
+      "loss": 1.4308,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6377072548578288,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016874529296580812,
+      "loss": 1.4061,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6380073523895265,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00016873023045639403,
+      "loss": 1.3948,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6383074499212243,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016871516794697997,
+      "loss": 1.615,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6386075474529223,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001687001054375659,
+      "loss": 1.5501,
+      "step": 2128
+    },
+    {
+      "epoch": 0.63890764498462,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016868504292815183,
+      "loss": 1.4253,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6392077425163178,
+      "grad_norm": 2.703125,
+      "learning_rate": 0.00016866998041873777,
+      "loss": 1.3952,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6395078400480156,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001686549179093237,
+      "loss": 1.3179,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6398079375797134,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016863985539990962,
+      "loss": 1.3278,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6401080351114112,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001686247928904956,
+      "loss": 1.4715,
+      "step": 2133
+    },
+    {
+      "epoch": 0.640408132643109,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001686097303810815,
+      "loss": 1.4792,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6407082301748068,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001685946678716674,
+      "loss": 1.3861,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6410083277065046,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016857960536225338,
+      "loss": 1.3853,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6413084252382024,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001685645428528393,
+      "loss": 1.4075,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6416085227699002,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001685494803434252,
+      "loss": 1.281,
+      "step": 2138
+    },
+    {
+      "epoch": 0.641908620301598,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016853441783401115,
+      "loss": 1.1885,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6422087178332958,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001685193553245971,
+      "loss": 1.4923,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6425088153649936,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.000168504292815183,
+      "loss": 1.2623,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6428089128966914,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016848923030576894,
+      "loss": 1.3488,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6431090104283892,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016847416779635488,
+      "loss": 1.3217,
+      "step": 2143
+    },
+    {
+      "epoch": 0.643409107960087,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016845910528694082,
+      "loss": 1.4212,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6437092054917848,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016844404277752674,
+      "loss": 1.3235,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6440093030234826,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00016842898026811268,
+      "loss": 1.3334,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6443094005551804,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016841391775869862,
+      "loss": 1.4292,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6446094980868783,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016839885524928453,
+      "loss": 1.2635,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6449095956185761,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016838379273987047,
+      "loss": 1.221,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6452096931502739,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001683687302304564,
+      "loss": 1.3479,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6455097906819717,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016835366772104232,
+      "loss": 1.2157,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6458098882136695,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016833860521162826,
+      "loss": 1.5322,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6461099857453673,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001683235427022142,
+      "loss": 1.2209,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6464100832770651,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016830848019280012,
+      "loss": 1.2631,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6467101808087629,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016829341768338606,
+      "loss": 1.4527,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6470102783404607,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.000168278355173972,
+      "loss": 1.4074,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6473103758721584,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001682632926645579,
+      "loss": 1.4008,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6476104734038562,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016824823015514385,
+      "loss": 1.4183,
+      "step": 2158
+    },
+    {
+      "epoch": 0.647910570935554,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001682331676457298,
+      "loss": 1.5418,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6482106684672518,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016821810513631573,
+      "loss": 1.5038,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6485107659989496,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016820304262690164,
+      "loss": 1.4,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6488108635306474,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016818798011748758,
+      "loss": 1.3827,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6491109610623452,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016817291760807353,
+      "loss": 1.2823,
+      "step": 2163
+    },
+    {
+      "epoch": 0.649411058594043,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016815785509865944,
+      "loss": 1.3741,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6497111561257408,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016814279258924538,
+      "loss": 1.5198,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6500112536574386,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016812773007983132,
+      "loss": 1.3139,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6503113511891365,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016811266757041723,
+      "loss": 1.5163,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6506114487208343,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016809760506100317,
+      "loss": 1.467,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6509115462525321,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001680825425515891,
+      "loss": 1.3366,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6512116437842299,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016806748004217503,
+      "loss": 1.3525,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6515117413159277,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016805241753276097,
+      "loss": 1.4356,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6518118388476255,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001680373550233469,
+      "loss": 1.3246,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6521119363793233,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016802229251393282,
+      "loss": 1.3699,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6524120339110211,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016800723000451876,
+      "loss": 1.4648,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6527121314427189,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001679921674951047,
+      "loss": 1.511,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6530122289744167,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001679771049856906,
+      "loss": 1.2011,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6533123265061145,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016796204247627655,
+      "loss": 1.3102,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6536124240378123,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001679469799668625,
+      "loss": 1.2193,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6539125215695101,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016793191745744843,
+      "loss": 1.3843,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6542126191012079,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016791685494803435,
+      "loss": 1.4811,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6545127166329057,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001679017924386203,
+      "loss": 1.2737,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6548128141646035,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016788672992920623,
+      "loss": 1.2797,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6551129116963013,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016787166741979214,
+      "loss": 1.4352,
+      "step": 2183
+    },
+    {
+      "epoch": 0.655413009227999,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016785660491037805,
+      "loss": 1.3475,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6557131067596969,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016784154240096402,
+      "loss": 1.3562,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6560132042913948,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016782647989154994,
+      "loss": 1.4082,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6563133018230926,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016781141738213588,
+      "loss": 1.392,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6566133993547904,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016779635487272182,
+      "loss": 1.2578,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6569134968864881,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016778129236330773,
+      "loss": 1.3799,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6572135944181859,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016776622985389367,
+      "loss": 1.4125,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6575136919498837,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001677511673444796,
+      "loss": 1.3712,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6578137894815815,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016773610483506552,
+      "loss": 1.3932,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6581138870132793,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016772104232565146,
+      "loss": 1.3088,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6584139845449771,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001677059798162374,
+      "loss": 1.3538,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6587140820766749,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016769091730682332,
+      "loss": 1.4042,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6590141796083727,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016767585479740926,
+      "loss": 1.3308,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6593142771400705,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016766079228799517,
+      "loss": 1.4975,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6596143746717683,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00016764572977858114,
+      "loss": 1.5129,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6599144722034661,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016763066726916705,
+      "loss": 1.4197,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6602145697351639,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016761560475975296,
+      "loss": 1.3074,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6605146672668617,
+      "grad_norm": 1.859375,
+      "learning_rate": 0.00016760054225033893,
+      "loss": 1.5233,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6608147647985595,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016758547974092484,
+      "loss": 1.2591,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6611148623302573,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016757041723151076,
+      "loss": 1.353,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6614149598619551,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016755535472209673,
+      "loss": 1.4814,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6617150573936529,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016754029221268264,
+      "loss": 1.5387,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6620151549253508,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016752522970326858,
+      "loss": 1.2793,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6623152524570486,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016751016719385452,
+      "loss": 1.4917,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6626153499887464,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016749510468444043,
+      "loss": 1.314,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6629154475204442,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016748004217502637,
+      "loss": 1.591,
+      "step": 2209
+    },
+    {
+      "epoch": 0.663215545052142,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016746497966561229,
+      "loss": 1.3384,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6635156425838398,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016744991715619823,
+      "loss": 1.3663,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6638157401155376,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016743485464678417,
+      "loss": 1.4415,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6641158376472354,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016741979213737008,
+      "loss": 1.3108,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6644159351789332,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016740472962795605,
+      "loss": 1.4374,
+      "step": 2214
+    },
+    {
+      "epoch": 0.664716032710631,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016738966711854196,
+      "loss": 1.3652,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6650161302423288,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016737460460912787,
+      "loss": 1.4494,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6653162277740265,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016735954209971384,
+      "loss": 1.2973,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6656163253057243,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016734447959029975,
+      "loss": 1.3816,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6659164228374221,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016732941708088567,
+      "loss": 1.4064,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6662165203691199,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001673143545714716,
+      "loss": 1.2997,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6665166179008177,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016729929206205755,
+      "loss": 1.4165,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6668167154325155,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016728422955264346,
+      "loss": 1.3225,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6671168129642133,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001672691670432294,
+      "loss": 1.3503,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6674169104959111,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016725410453381534,
+      "loss": 1.5027,
+      "step": 2224
+    },
+    {
+      "epoch": 0.667717008027609,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016723904202440128,
+      "loss": 1.3812,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6680171055593068,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001672239795149872,
+      "loss": 1.332,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6683172030910046,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016720891700557314,
+      "loss": 1.4005,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6686173006227024,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016719385449615908,
+      "loss": 1.3724,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6689173981544002,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000167178791986745,
+      "loss": 1.3457,
+      "step": 2229
+    },
+    {
+      "epoch": 0.669217495686098,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016716372947733093,
+      "loss": 1.4396,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6695175932177958,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016714866696791687,
+      "loss": 1.4231,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6698176907494936,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016713360445850278,
+      "loss": 1.3275,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6701177882811914,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016711854194908872,
+      "loss": 1.4999,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6704178858128892,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016710347943967466,
+      "loss": 1.3031,
+      "step": 2234
+    },
+    {
+      "epoch": 0.670717983344587,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016708841693026058,
+      "loss": 1.3323,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6710180808762848,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016707335442084652,
+      "loss": 1.2973,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6713181784079826,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016705829191143246,
+      "loss": 1.3943,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6716182759396804,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016704322940201837,
+      "loss": 1.3391,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6719183734713782,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.0001670281668926043,
+      "loss": 1.5491,
+      "step": 2239
+    },
+    {
+      "epoch": 0.672218471003076,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016701310438319025,
+      "loss": 1.4836,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6725185685347738,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001669980418737762,
+      "loss": 1.2742,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6728186660664716,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001669829793643621,
+      "loss": 1.2701,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6731187635981694,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016696791685494804,
+      "loss": 1.3819,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6734188611298672,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016695285434553399,
+      "loss": 1.3507,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6737189586615651,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001669377918361199,
+      "loss": 1.3926,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6740190561932629,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016692272932670584,
+      "loss": 1.4646,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6743191537249607,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016690766681729178,
+      "loss": 1.3839,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6746192512566584,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001668926043078777,
+      "loss": 1.4323,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6749193487883562,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016687754179846363,
+      "loss": 1.552,
+      "step": 2249
+    },
+    {
+      "epoch": 0.675219446320054,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016686247928904957,
+      "loss": 1.4602,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6755195438517518,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016684741677963549,
+      "loss": 1.2683,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6758196413834496,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016683235427022143,
+      "loss": 1.4352,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6761197389151474,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00016681729176080737,
+      "loss": 1.5591,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6764198364468452,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016680222925139328,
+      "loss": 1.2423,
+      "step": 2254
+    },
+    {
+      "epoch": 0.676719933978543,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016678716674197922,
+      "loss": 1.1472,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6770200315102408,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016677210423256516,
+      "loss": 1.3632,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6773201290419386,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016675704172315107,
+      "loss": 1.3397,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6776202265736364,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016674197921373701,
+      "loss": 1.4747,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6779203241053342,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016672691670432295,
+      "loss": 1.5768,
+      "step": 2259
+    },
+    {
+      "epoch": 0.678220421637032,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001667118541949089,
+      "loss": 1.5026,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6785205191687298,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001666967916854948,
+      "loss": 1.3354,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6788206167004276,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00016668172917608075,
+      "loss": 1.5111,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6791207142321254,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.4864,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6794208117638233,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001666516041572526,
+      "loss": 1.4499,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6797209092955211,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016663654164783854,
+      "loss": 1.3754,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6800210068272189,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016662147913842448,
+      "loss": 1.2978,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6803211043589167,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001666064166290104,
+      "loss": 1.5103,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6806212018906145,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016659135411959634,
+      "loss": 1.4249,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6809212994223123,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016657629161018228,
+      "loss": 1.2607,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6812213969540101,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001665612291007682,
+      "loss": 1.5475,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6815214944857079,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016654616659135413,
+      "loss": 1.3454,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6818215920174057,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016653110408194007,
+      "loss": 1.3833,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6821216895491035,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00016651604157252598,
+      "loss": 1.4289,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6824217870808013,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016650097906311192,
+      "loss": 1.4188,
+      "step": 2274
+    },
+    {
+      "epoch": 0.682721884612499,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016648591655369786,
+      "loss": 1.5501,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6830219821441968,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016647085404428378,
+      "loss": 1.2889,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6833220796758946,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016645579153486972,
+      "loss": 1.4575,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6836221772075924,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016644072902545566,
+      "loss": 1.3279,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6839222747392902,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001664256665160416,
+      "loss": 1.3548,
+      "step": 2279
+    },
+    {
+      "epoch": 0.684222372270988,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001664106040066275,
+      "loss": 1.5138,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6845224698026858,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016639554149721342,
+      "loss": 1.3876,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6848225673343836,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001663804789877994,
+      "loss": 1.3444,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6851226648660814,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001663654164783853,
+      "loss": 1.4409,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6854227623977793,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016635035396897122,
+      "loss": 1.3196,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6857228599294771,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016633529145955719,
+      "loss": 1.3237,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6860229574611749,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001663202289501431,
+      "loss": 1.2925,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6863230549928727,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016630516644072904,
+      "loss": 1.5131,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6866231525245705,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016629010393131498,
+      "loss": 1.4874,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6869232500562683,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001662750414219009,
+      "loss": 1.5825,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6872233475879661,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016625997891248683,
+      "loss": 1.4458,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6875234451196639,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016624491640307275,
+      "loss": 1.274,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6878235426513617,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016622985389365869,
+      "loss": 1.4105,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6881236401830595,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016621479138424463,
+      "loss": 1.185,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6884237377147573,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016619972887483054,
+      "loss": 1.4877,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6887238352464551,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001661846663654165,
+      "loss": 1.2655,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6890239327781529,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016616960385600242,
+      "loss": 1.4218,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6893240303098507,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016615454134658833,
+      "loss": 1.3278,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6896241278415485,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001661394788371743,
+      "loss": 1.3682,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6899242253732463,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016612441632776021,
+      "loss": 1.2872,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6902243229049441,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016610935381834613,
+      "loss": 1.3642,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6905244204366419,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001660942913089321,
+      "loss": 1.2341,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6908245179683397,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.000166079228799518,
+      "loss": 1.1779,
+      "step": 2302
+    },
+    {
+      "epoch": 0.6911246155000376,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016606416629010392,
+      "loss": 1.4119,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6914247130317354,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00016604910378068986,
+      "loss": 1.3741,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6917248105634332,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001660340412712758,
+      "loss": 1.4136,
+      "step": 2305
+    },
+    {
+      "epoch": 0.692024908095131,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016601897876186174,
+      "loss": 1.2977,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6923250056268287,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016600391625244766,
+      "loss": 1.2931,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6926251031585265,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001659888537430336,
+      "loss": 1.5669,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6929252006902243,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00016597379123361954,
+      "loss": 1.5516,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6932252982219221,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016595872872420545,
+      "loss": 1.3889,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6935253957536199,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001659436662147914,
+      "loss": 1.4215,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6938254932853177,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016592860370537733,
+      "loss": 1.4341,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6941255908170155,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016591354119596324,
+      "loss": 1.5146,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6944256883487133,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001658984786865492,
+      "loss": 1.4048,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6947257858804111,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016588341617713512,
+      "loss": 1.2667,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6950258834121089,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016586835366772104,
+      "loss": 1.2473,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6953259809438067,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016585329115830698,
+      "loss": 1.2827,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6956260784755045,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016583822864889292,
+      "loss": 1.4406,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6959261760072023,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016582316613947883,
+      "loss": 1.2125,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6962262735389001,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016580810363006477,
+      "loss": 1.3531,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6965263710705979,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001657930411206507,
+      "loss": 1.3753,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6968264686022958,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016577797861123665,
+      "loss": 1.2808,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6971265661339936,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00016576291610182256,
+      "loss": 1.4129,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6974266636656914,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001657478535924085,
+      "loss": 1.3764,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6977267611973892,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00016573279108299445,
+      "loss": 1.3886,
+      "step": 2325
+    },
+    {
+      "epoch": 0.698026858729087,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016571772857358036,
+      "loss": 1.6038,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6983269562607848,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001657026660641663,
+      "loss": 1.5274,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6986270537924826,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016568760355475224,
+      "loss": 1.2512,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6989271513241804,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016567254104533815,
+      "loss": 1.3583,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6992272488558782,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001656574785359241,
+      "loss": 1.2966,
+      "step": 2330
+    },
+    {
+      "epoch": 0.699527346387576,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016564241602651003,
+      "loss": 1.3883,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6998274439192738,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016562735351709595,
+      "loss": 1.426,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7001275414509716,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016561229100768189,
+      "loss": 1.1721,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7004276389826694,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016559722849826783,
+      "loss": 1.4719,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7007277365143671,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016558216598885374,
+      "loss": 1.2412,
+      "step": 2335
+    },
+    {
+      "epoch": 0.701027834046065,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016556710347943968,
+      "loss": 1.4902,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7013279315777627,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016555204097002562,
+      "loss": 1.2423,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7016280291094605,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016553697846061153,
+      "loss": 1.2318,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7019281266411583,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016552191595119747,
+      "loss": 1.4884,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7022282241728561,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016550685344178341,
+      "loss": 1.2953,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7025283217045539,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016549179093236935,
+      "loss": 1.4376,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7028284192362518,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016547672842295527,
+      "loss": 1.2512,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7031285167679496,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001654616659135412,
+      "loss": 1.2608,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7034286142996474,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016544660340412715,
+      "loss": 1.3046,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7037287118313452,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016543154089471306,
+      "loss": 1.3317,
+      "step": 2345
+    },
+    {
+      "epoch": 0.704028809363043,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.000165416478385299,
+      "loss": 1.4813,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7043289068947408,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016540141587588494,
+      "loss": 1.3034,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7046290044264386,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016538635336647086,
+      "loss": 1.2364,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7049291019581364,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001653712908570568,
+      "loss": 1.4839,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7052291994898342,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016535622834764274,
+      "loss": 1.5186,
+      "step": 2350
+    },
+    {
+      "epoch": 0.705529297021532,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00016534116583822865,
+      "loss": 1.3338,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7058293945532298,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001653261033288146,
+      "loss": 1.344,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7061294920849276,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016531104081940053,
+      "loss": 1.1921,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7064295896166254,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016529597830998644,
+      "loss": 1.4924,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7067296871483232,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016528091580057238,
+      "loss": 1.2945,
+      "step": 2355
+    },
+    {
+      "epoch": 0.707029784680021,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016526585329115832,
+      "loss": 1.3608,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7073298822117188,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016525079078174424,
+      "loss": 1.281,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7076299797434166,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016523572827233018,
+      "loss": 1.2147,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7079300772751144,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00016522066576291612,
+      "loss": 1.4685,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7082301748068122,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016520560325350206,
+      "loss": 1.2432,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7085302723385101,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016519054074408797,
+      "loss": 1.4935,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7088303698702079,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016517547823467388,
+      "loss": 1.3243,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7091304674019057,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016516041572525985,
+      "loss": 1.4045,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7094305649336035,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016514535321584576,
+      "loss": 1.3593,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7097306624653013,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00016513029070643168,
+      "loss": 1.5028,
+      "step": 2365
+    },
+    {
+      "epoch": 0.710030759996999,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016511522819701765,
+      "loss": 1.2952,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7103308575286968,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016510016568760356,
+      "loss": 1.3847,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7106309550603946,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001650851031781895,
+      "loss": 1.3531,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7109310525920924,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016507004066877544,
+      "loss": 1.2346,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7112311501237902,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016505497815936135,
+      "loss": 1.3223,
+      "step": 2370
+    },
+    {
+      "epoch": 0.711531247655488,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001650399156499473,
+      "loss": 1.2628,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7118313451871858,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016502485314053323,
+      "loss": 1.3046,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7121314427188836,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016500979063111915,
+      "loss": 1.3139,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7124315402505814,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016499472812170509,
+      "loss": 1.3732,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7127316377822792,
+      "grad_norm": 1.375,
+      "learning_rate": 0.000164979665612291,
+      "loss": 1.3496,
+      "step": 2375
+    },
+    {
+      "epoch": 0.713031735313977,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016496460310287697,
+      "loss": 1.291,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7133318328456748,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016494954059346288,
+      "loss": 1.4153,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7136319303773726,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001649344780840488,
+      "loss": 1.1803,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7139320279090704,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016491941557463476,
+      "loss": 1.3748,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7142321254407682,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016490435306522067,
+      "loss": 1.5438,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7145322229724661,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001648892905558066,
+      "loss": 1.3813,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7148323205041639,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016487422804639255,
+      "loss": 1.2497,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7151324180358617,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016485916553697847,
+      "loss": 1.4777,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7154325155675595,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016484410302756438,
+      "loss": 1.1348,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7157326130992573,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016482904051815035,
+      "loss": 1.313,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7160327106309551,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016481397800873626,
+      "loss": 1.4981,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7163328081626529,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001647989154993222,
+      "loss": 1.2701,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7166329056943507,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016478385298990812,
+      "loss": 1.3422,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7169330032260485,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016476879048049406,
+      "loss": 1.3946,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7172331007577463,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016475372797108,
+      "loss": 1.5386,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7175331982894441,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001647386654616659,
+      "loss": 1.3938,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7178332958211419,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016472360295225185,
+      "loss": 1.3085,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7181333933528397,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001647085404428378,
+      "loss": 1.3475,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7184334908845375,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001646934779334237,
+      "loss": 1.3294,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7187335884162352,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00016467841542400967,
+      "loss": 1.2625,
+      "step": 2395
+    },
+    {
+      "epoch": 0.719033685947933,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016466335291459558,
+      "loss": 1.53,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7193337834796308,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001646482904051815,
+      "loss": 1.3099,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7196338810113286,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00016463322789576744,
+      "loss": 1.6263,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7199339785430264,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016461816538635338,
+      "loss": 1.3391,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7202340760747243,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001646031028769393,
+      "loss": 1.5896,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7205341736064221,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016458804036752523,
+      "loss": 1.3596,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7208342711381199,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016457297785811117,
+      "loss": 1.4132,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7211343686698177,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001645579153486971,
+      "loss": 1.3478,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7214344662015155,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016454285283928302,
+      "loss": 1.4263,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7217345637332133,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016452779032986896,
+      "loss": 1.354,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7220346612649111,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001645127278204549,
+      "loss": 1.4068,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7223347587966089,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016449766531104082,
+      "loss": 1.3618,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7226348563283067,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016448260280162676,
+      "loss": 1.2602,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7229349538600045,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001644675402922127,
+      "loss": 1.2946,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7232350513917023,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001644524777827986,
+      "loss": 1.2878,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7235351489234001,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016443741527338455,
+      "loss": 1.3785,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7238352464550979,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001644223527639705,
+      "loss": 1.5394,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7241353439867957,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001644072902545564,
+      "loss": 1.4331,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7244354415184935,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016439222774514235,
+      "loss": 1.317,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7247355390501913,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001643771652357283,
+      "loss": 1.5039,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7250356365818891,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001643621027263142,
+      "loss": 1.3948,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7253357341135869,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016434704021690014,
+      "loss": 1.312,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7256358316452847,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016433197770748608,
+      "loss": 1.2869,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7259359291769825,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000164316915198072,
+      "loss": 1.2166,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7262360267086804,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016430185268865793,
+      "loss": 1.3935,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7265361242403782,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016428679017924387,
+      "loss": 1.2943,
+      "step": 2421
+    },
+    {
+      "epoch": 0.726836221772076,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00016427172766982981,
+      "loss": 1.3658,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7271363193037738,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016425666516041573,
+      "loss": 1.193,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7274364168354716,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016424160265100167,
+      "loss": 1.1766,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7277365143671694,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001642265401415876,
+      "loss": 1.4953,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7280366118988671,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016421147763217352,
+      "loss": 1.4068,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7283367094305649,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016419641512275946,
+      "loss": 1.4113,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7286368069622627,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001641813526133454,
+      "loss": 1.3556,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7289369044939605,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016416629010393132,
+      "loss": 1.409,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7292370020256583,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00016415122759451726,
+      "loss": 1.2294,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7295370995573561,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001641361650851032,
+      "loss": 1.3016,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7298371970890539,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001641211025756891,
+      "loss": 1.446,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7301372946207517,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016410604006627505,
+      "loss": 1.4124,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7304373921524495,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.000164090977556861,
+      "loss": 1.3405,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7307374896841473,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001640759150474469,
+      "loss": 1.2619,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7310375872158451,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016406085253803284,
+      "loss": 1.5061,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7313376847475429,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016404579002861878,
+      "loss": 1.3273,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7316377822792407,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001640307275192047,
+      "loss": 1.2618,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7319378798109386,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016401566500979064,
+      "loss": 1.4892,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7322379773426364,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016400060250037658,
+      "loss": 1.3371,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7325380748743342,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016398553999096252,
+      "loss": 1.3791,
+      "step": 2441
+    },
+    {
+      "epoch": 0.732838172406032,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016397047748154843,
+      "loss": 1.4313,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7331382699377298,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016395541497213437,
+      "loss": 1.3041,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7334383674694276,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001639403524627203,
+      "loss": 1.4023,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7337384650011254,
+      "grad_norm": 1.6015625,
+      "learning_rate": 0.00016392528995330622,
+      "loss": 1.5263,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7340385625328232,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00016391022744389214,
+      "loss": 1.3521,
+      "step": 2446
+    },
+    {
+      "epoch": 0.734338660064521,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001638951649344781,
+      "loss": 1.3785,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7346387575962188,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016388010242506402,
+      "loss": 1.2985,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7349388551279166,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016386503991564996,
+      "loss": 1.2019,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7352389526596144,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001638499774062359,
+      "loss": 1.6236,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7355390501913122,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001638349148968218,
+      "loss": 1.2872,
+      "step": 2451
+    },
+    {
+      "epoch": 0.73583914772301,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016381985238740775,
+      "loss": 1.3471,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7361392452547078,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001638047898779937,
+      "loss": 1.2036,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7364393427864055,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001637897273685796,
+      "loss": 1.2311,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7367394403181033,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016377466485916555,
+      "loss": 1.2791,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7370395378498011,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016375960234975146,
+      "loss": 1.3627,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7373396353814989,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016374453984033743,
+      "loss": 1.1719,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7376397329131968,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016372947733092334,
+      "loss": 1.1474,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7379398304448946,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016371441482150925,
+      "loss": 1.3987,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7382399279765924,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016369935231209522,
+      "loss": 1.1942,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7385400255082902,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016368428980268113,
+      "loss": 1.3261,
+      "step": 2461
+    },
+    {
+      "epoch": 0.738840123039988,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016366922729326705,
+      "loss": 1.2698,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7391402205716858,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016365416478385301,
+      "loss": 1.321,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7394403181033836,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016363910227443893,
+      "loss": 1.4005,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7397404156350814,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016362403976502484,
+      "loss": 1.2717,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7400405131667792,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001636089772556108,
+      "loss": 1.1605,
+      "step": 2466
+    },
+    {
+      "epoch": 0.740340610698477,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016359391474619672,
+      "loss": 1.1858,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7406407082301748,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00016357885223678266,
+      "loss": 1.3643,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7409408057618726,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016356378972736858,
+      "loss": 1.4417,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7412409032935704,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00016354872721795452,
+      "loss": 1.4571,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7415410008252682,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016353366470854046,
+      "loss": 1.04,
+      "step": 2471
+    },
+    {
+      "epoch": 0.741841098356966,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016351860219912637,
+      "loss": 1.3014,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7421411958886638,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001635035396897123,
+      "loss": 1.2089,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7424412934203616,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016348847718029825,
+      "loss": 1.3155,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7427413909520594,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00016347341467088416,
+      "loss": 1.4545,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7430414884837572,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016345835216147013,
+      "loss": 1.4591,
+      "step": 2476
+    },
+    {
+      "epoch": 0.743341586015455,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016344328965205604,
+      "loss": 1.3655,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7436416835471529,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00016342822714264196,
+      "loss": 1.2625,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7439417810788507,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016341316463322792,
+      "loss": 1.4249,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7442418786105485,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016339810212381384,
+      "loss": 1.3508,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7445419761422463,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016338303961439975,
+      "loss": 1.3642,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7448420736739441,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001633679771049857,
+      "loss": 1.2477,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7451421712056419,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016335291459557163,
+      "loss": 1.457,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7454422687373397,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016333785208615757,
+      "loss": 1.5931,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7457423662690374,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016332278957674348,
+      "loss": 1.7071,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7460424638007352,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016330772706732942,
+      "loss": 1.2662,
+      "step": 2486
+    },
+    {
+      "epoch": 0.746342561332433,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016329266455791537,
+      "loss": 1.2835,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7466426588641308,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016327760204850128,
+      "loss": 1.3271,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7469427563958286,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016326253953908722,
+      "loss": 1.413,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7472428539275264,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016324747702967316,
+      "loss": 1.1623,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7475429514592242,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016323241452025907,
+      "loss": 1.3886,
+      "step": 2491
+    },
+    {
+      "epoch": 0.747843048990922,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000163217352010845,
+      "loss": 1.2108,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7481431465226198,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016320228950143095,
+      "loss": 1.2996,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7484432440543176,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016318722699201687,
+      "loss": 1.2665,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7487433415860154,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001631721644826028,
+      "loss": 1.3944,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7490434391177132,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016315710197318875,
+      "loss": 1.2135,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7493435366494111,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016314203946377466,
+      "loss": 1.2127,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7496436341811089,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001631269769543606,
+      "loss": 1.3336,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7499437317128067,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016311191444494654,
+      "loss": 1.4273,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7502438292445045,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016309685193553245,
+      "loss": 1.374,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7505439267762023,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001630817894261184,
+      "loss": 1.5211,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7508440243079001,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016306672691670433,
+      "loss": 1.4218,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7511441218395979,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00016305166440729027,
+      "loss": 1.5465,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7514442193712957,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001630366018978762,
+      "loss": 1.3066,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7517443169029935,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016302153938846213,
+      "loss": 1.418,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7520444144346913,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00016300647687904807,
+      "loss": 1.4893,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7523445119663891,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016299141436963398,
+      "loss": 1.3826,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7526446094980869,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016297635186021992,
+      "loss": 1.424,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7529447070297847,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016296128935080586,
+      "loss": 1.2858,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7532448045614825,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016294622684139178,
+      "loss": 1.4046,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7535449020931803,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00016293116433197772,
+      "loss": 1.3442,
+      "step": 2511
+    },
+    {
+      "epoch": 0.753844999624878,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016291610182256366,
+      "loss": 1.4919,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7541450971565758,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016290103931314957,
+      "loss": 1.2478,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7544451946882736,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001628859768037355,
+      "loss": 1.3096,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7547452922199714,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016287091429432145,
+      "loss": 1.5153,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7550453897516692,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016285585178490736,
+      "loss": 1.417,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7553454872833671,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001628407892754933,
+      "loss": 1.2427,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7556455848150649,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016282572676607924,
+      "loss": 1.342,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7559456823467627,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016281066425666516,
+      "loss": 1.1807,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7562457798784605,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001627956017472511,
+      "loss": 1.3075,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7565458774101583,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016278053923783704,
+      "loss": 1.3451,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7568459749418561,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016276547672842298,
+      "loss": 1.4251,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7571460724735539,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001627504142190089,
+      "loss": 1.3827,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7574461700052517,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016273535170959483,
+      "loss": 1.4294,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7577462675369495,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016272028920018077,
+      "loss": 1.389,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7580463650686473,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016270522669076668,
+      "loss": 1.454,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7583464626003451,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001626901641813526,
+      "loss": 1.2574,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7586465601320429,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016267510167193857,
+      "loss": 1.3663,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7589466576637407,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016266003916252448,
+      "loss": 1.4384,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7592467551954385,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016264497665311042,
+      "loss": 1.5697,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7595468527271363,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016262991414369636,
+      "loss": 1.2541,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7598469502588341,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016261485163428227,
+      "loss": 1.3737,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7601470477905319,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001625997891248682,
+      "loss": 1.2426,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7604471453222297,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016258472661545415,
+      "loss": 1.5617,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7607472428539275,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016256966410604007,
+      "loss": 1.1622,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7610473403856254,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.000162554601596626,
+      "loss": 1.3918,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7613474379173232,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016253953908721195,
+      "loss": 1.3319,
+      "step": 2537
+    },
+    {
+      "epoch": 0.761647535449021,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.0001625244765777979,
+      "loss": 1.3363,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7619476329807188,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001625094140683838,
+      "loss": 1.4448,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7622477305124166,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001624943515589697,
+      "loss": 1.1553,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7625478280441144,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016247928904955568,
+      "loss": 1.3294,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7628479255758122,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001624642265401416,
+      "loss": 1.6004,
+      "step": 2542
+    },
+    {
+      "epoch": 0.76314802310751,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001624491640307275,
+      "loss": 1.4915,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7634481206392078,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016243410152131347,
+      "loss": 1.4076,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7637482181709055,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001624190390118994,
+      "loss": 1.297,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7640483157026033,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001624039765024853,
+      "loss": 1.2142,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7643484132343011,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016238891399307127,
+      "loss": 1.396,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7646485107659989,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016237385148365718,
+      "loss": 1.4826,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7649486082976967,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016235878897424312,
+      "loss": 1.2971,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7652487058293945,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016234372646482906,
+      "loss": 1.4358,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7655488033610923,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016232866395541498,
+      "loss": 1.4168,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7658489008927901,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016231360144600092,
+      "loss": 1.3902,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7661489984244879,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016229853893658683,
+      "loss": 1.3763,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7664490959561857,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016228347642717277,
+      "loss": 1.2573,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7667491934878835,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001622684139177587,
+      "loss": 1.2746,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7670492910195814,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016225335140834462,
+      "loss": 1.3898,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7673493885512792,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001622382888989306,
+      "loss": 1.3284,
+      "step": 2557
+    },
+    {
+      "epoch": 0.767649486082977,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001622232263895165,
+      "loss": 1.3377,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7679495836146748,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016220816388010242,
+      "loss": 1.1953,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7682496811463726,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016219310137068838,
+      "loss": 1.273,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7685497786780704,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001621780388612743,
+      "loss": 1.2597,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7688498762097682,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001621629763518602,
+      "loss": 1.5422,
+      "step": 2562
+    },
+    {
+      "epoch": 0.769149973741466,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016214791384244615,
+      "loss": 1.2478,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7694500712731638,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001621328513330321,
+      "loss": 1.4945,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7697501688048616,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016211778882361803,
+      "loss": 1.2119,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7700502663365594,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016210272631420394,
+      "loss": 1.501,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7703503638682572,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016208766380478988,
+      "loss": 1.2269,
+      "step": 2567
+    },
+    {
+      "epoch": 0.770650461399955,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016207260129537583,
+      "loss": 1.4066,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7709505589316528,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016205753878596174,
+      "loss": 1.266,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7712506564633506,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016204247627654768,
+      "loss": 1.3929,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7715507539950484,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016202741376713362,
+      "loss": 1.3153,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7718508515267462,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016201235125771953,
+      "loss": 1.2703,
+      "step": 2572
+    },
+    {
+      "epoch": 0.772150949058444,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016199728874830547,
+      "loss": 1.5509,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7724510465901417,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001619822262388914,
+      "loss": 1.3476,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7727511441218397,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016196716372947733,
+      "loss": 1.1306,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7730512416535374,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016195210122006327,
+      "loss": 1.3271,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7733513391852352,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001619370387106492,
+      "loss": 1.336,
+      "step": 2577
+    },
+    {
+      "epoch": 0.773651436716933,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00016192197620123512,
+      "loss": 1.5909,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7739515342486308,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016190691369182106,
+      "loss": 1.2674,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7742516317803286,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.000161891851182407,
+      "loss": 1.1689,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7745517293120264,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001618767886729929,
+      "loss": 1.3817,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7748518268437242,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016186172616357885,
+      "loss": 1.3514,
+      "step": 2582
+    },
+    {
+      "epoch": 0.775151924375422,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001618466636541648,
+      "loss": 1.3885,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7754520219071198,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016183160114475073,
+      "loss": 1.4564,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7757521194388176,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016181653863533665,
+      "loss": 1.266,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7760522169705154,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001618014761259226,
+      "loss": 1.2088,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7763523145022132,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016178641361650853,
+      "loss": 1.2892,
+      "step": 2587
+    },
+    {
+      "epoch": 0.776652412033911,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016177135110709444,
+      "loss": 1.3157,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7769525095656088,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016175628859768038,
+      "loss": 1.2163,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7772526070973066,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016174122608826632,
+      "loss": 1.2627,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7775527046290044,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016172616357885224,
+      "loss": 1.1806,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7778528021607022,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016171110106943818,
+      "loss": 1.4765,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7781528996924,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016169603856002412,
+      "loss": 1.2464,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7784529972240979,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016168097605061003,
+      "loss": 1.271,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7787530947557957,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00016166591354119597,
+      "loss": 1.3756,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7790531922874935,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001616508510317819,
+      "loss": 1.4012,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7793532898191913,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016163578852236782,
+      "loss": 1.3908,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7796533873508891,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016162072601295376,
+      "loss": 1.3265,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7799534848825869,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001616056635035397,
+      "loss": 1.4816,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7802535824142847,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016159060099412562,
+      "loss": 1.2081,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7805536799459825,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016157553848471156,
+      "loss": 1.4784,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7808537774776803,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001615604759752975,
+      "loss": 1.4617,
+      "step": 2602
+    },
+    {
+      "epoch": 0.781153875009378,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016154541346588344,
+      "loss": 1.2588,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7814539725410758,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016153035095646935,
+      "loss": 1.2944,
+      "step": 2604
+    },
+    {
+      "epoch": 0.7817540700727736,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001615152884470553,
+      "loss": 1.243,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7820541676044714,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016150022593764123,
+      "loss": 1.2763,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7823542651361692,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016148516342822714,
+      "loss": 1.2895,
+      "step": 2607
+    },
+    {
+      "epoch": 0.782654362667867,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016147010091881308,
+      "loss": 1.2862,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7829544601995648,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016145503840939903,
+      "loss": 1.3878,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7832545577312626,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016143997589998494,
+      "loss": 1.2728,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7835546552629604,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00016142491339057088,
+      "loss": 1.2195,
+      "step": 2611
+    },
+    {
+      "epoch": 0.7838547527946582,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016140985088115682,
+      "loss": 1.4686,
+      "step": 2612
+    },
+    {
+      "epoch": 0.784154850326356,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016139478837174273,
+      "loss": 1.4764,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7844549478580539,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016137972586232867,
+      "loss": 1.3493,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7847550453897517,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.0001613646633529146,
+      "loss": 1.4249,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7850551429214495,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016134960084350053,
+      "loss": 1.3033,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7853552404531473,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016133453833408647,
+      "loss": 1.3315,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7856553379848451,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001613194758246724,
+      "loss": 1.3022,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7859554355165429,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016130441331525835,
+      "loss": 1.2474,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7862555330482407,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00016128935080584426,
+      "loss": 1.0205,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7865556305799385,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001612742882964302,
+      "loss": 1.2648,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7868557281116363,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016125922578701614,
+      "loss": 1.2351,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7871558256433341,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00016124416327760205,
+      "loss": 1.3173,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7874559231750319,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016122910076818797,
+      "loss": 1.3434,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7877560207067297,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00016121403825877393,
+      "loss": 1.3572,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7880561182384275,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00016119897574935985,
+      "loss": 1.4937,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7883562157701253,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016118391323994576,
+      "loss": 1.3907,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7886563133018231,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016116885073053173,
+      "loss": 1.2821,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7889564108335209,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016115378822111764,
+      "loss": 1.2965,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7892565083652187,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016113872571170358,
+      "loss": 1.3807,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7895566058969165,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00016112366320228952,
+      "loss": 1.4485,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7898567034286142,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016110860069287544,
+      "loss": 1.5083,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7901568009603122,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016109353818346138,
+      "loss": 1.476,
+      "step": 2633
+    },
+    {
+      "epoch": 0.79045689849201,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001610784756740473,
+      "loss": 1.2326,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7907569960237077,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016106341316463323,
+      "loss": 1.2415,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7910570935554055,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00016104835065521917,
+      "loss": 1.5179,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7913571910871033,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016103328814580508,
+      "loss": 1.3532,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7916572886188011,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016101822563639105,
+      "loss": 1.5036,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7919573861504989,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016100316312697696,
+      "loss": 1.1606,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7922574836821967,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016098810061756288,
+      "loss": 1.3987,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7925575812138945,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016097303810814884,
+      "loss": 1.3679,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7928576787455923,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016095797559873476,
+      "loss": 1.3997,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7931577762772901,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016094291308932067,
+      "loss": 1.1681,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7934578738089879,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016092785057990664,
+      "loss": 1.3417,
+      "step": 2644
+    },
+    {
+      "epoch": 0.7937579713406857,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016091278807049255,
+      "loss": 1.3156,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7940580688723835,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001608977255610785,
+      "loss": 1.3437,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7943581664040813,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001608826630516644,
+      "loss": 1.2768,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7946582639357791,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016086760054225034,
+      "loss": 1.3396,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7949583614674769,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016085253803283629,
+      "loss": 1.4577,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7952584589991747,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001608374755234222,
+      "loss": 1.3011,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7955585565308725,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016082241301400814,
+      "loss": 1.2038,
+      "step": 2651
+    },
+    {
+      "epoch": 0.7958586540625703,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016080735050459408,
+      "loss": 1.5311,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7961587515942682,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016079228799518,
+      "loss": 1.3579,
+      "step": 2653
+    },
+    {
+      "epoch": 0.796458849125966,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016077722548576593,
+      "loss": 1.2711,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7967589466576638,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016076216297635187,
+      "loss": 1.3668,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7970590441893616,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016074710046693779,
+      "loss": 1.226,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7973591417210594,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016073203795752375,
+      "loss": 1.5609,
+      "step": 2657
+    },
+    {
+      "epoch": 0.7976592392527572,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016071697544810967,
+      "loss": 1.3704,
+      "step": 2658
+    },
+    {
+      "epoch": 0.797959336784455,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00016070191293869558,
+      "loss": 1.4582,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7982594343161528,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016068685042928152,
+      "loss": 1.4081,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7985595318478506,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016067178791986746,
+      "loss": 1.3463,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7988596293795484,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016065672541045337,
+      "loss": 1.3146,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7991597269112461,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016064166290103931,
+      "loss": 1.3459,
+      "step": 2663
+    },
+    {
+      "epoch": 0.799459824442944,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016062660039162525,
+      "loss": 1.2289,
+      "step": 2664
+    },
+    {
+      "epoch": 0.7997599219746417,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001606115378822112,
+      "loss": 1.2372,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8000600195063395,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001605964753727971,
+      "loss": 1.1526,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8003601170380373,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016058141286338305,
+      "loss": 1.3241,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8006602145697351,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.000160566350353969,
+      "loss": 1.2164,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8009603121014329,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001605512878445549,
+      "loss": 1.45,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8012604096331307,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016053622533514084,
+      "loss": 1.3473,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8015605071648285,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00016052116282572678,
+      "loss": 1.4477,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8018606046965264,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001605061003163127,
+      "loss": 1.3464,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8021607022282242,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016049103780689864,
+      "loss": 1.5677,
+      "step": 2673
+    },
+    {
+      "epoch": 0.802460799759922,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016047597529748458,
+      "loss": 1.2047,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8027608972916198,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001604609127880705,
+      "loss": 1.3808,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8030609948233176,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016044585027865643,
+      "loss": 1.4955,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8033610923550154,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016043078776924237,
+      "loss": 1.1822,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8036611898867132,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00016041572525982828,
+      "loss": 1.5124,
+      "step": 2678
+    },
+    {
+      "epoch": 0.803961287418411,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00016040066275041422,
+      "loss": 1.519,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8042613849501088,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016038560024100016,
+      "loss": 1.2999,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8045614824818066,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00016037053773158608,
+      "loss": 1.4215,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8048615800135044,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016035547522217202,
+      "loss": 1.1833,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8051616775452022,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016034041271275796,
+      "loss": 1.4053,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8054617750769,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001603253502033439,
+      "loss": 1.2544,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8057618726085978,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001603102876939298,
+      "loss": 1.2291,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8060619701402956,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00016029522518451575,
+      "loss": 1.3917,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8063620676719934,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001602801626751017,
+      "loss": 1.2827,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8066621652036912,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001602651001656876,
+      "loss": 1.3315,
+      "step": 2688
+    },
+    {
+      "epoch": 0.806962262735389,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016025003765627354,
+      "loss": 1.3306,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8072623602670868,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016023497514685949,
+      "loss": 1.3486,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8075624577987845,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001602199126374454,
+      "loss": 1.3045,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8078625553304825,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00016020485012803134,
+      "loss": 1.3687,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8081626528621803,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00016018978761861728,
+      "loss": 1.4814,
+      "step": 2693
+    },
+    {
+      "epoch": 0.808462750393878,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001601747251092032,
+      "loss": 1.3221,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8087628479255758,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016015966259978913,
+      "loss": 1.3468,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8090629454572736,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016014460009037507,
+      "loss": 1.1956,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8093630429889714,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016012953758096099,
+      "loss": 1.6465,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8096631405206692,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00016011447507154693,
+      "loss": 1.2427,
+      "step": 2698
+    },
+    {
+      "epoch": 0.809963238052367,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016009941256213287,
+      "loss": 1.3986,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8102633355840648,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001600843500527188,
+      "loss": 1.3011,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8105634331157626,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00016006928754330472,
+      "loss": 1.2917,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8108635306474604,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00016005422503389066,
+      "loss": 1.1727,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8111636281791582,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001600391625244766,
+      "loss": 1.4837,
+      "step": 2703
+    },
+    {
+      "epoch": 0.811463725710856,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016002410001506251,
+      "loss": 1.2761,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8117638232425538,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016000903750564843,
+      "loss": 1.3012,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8120639207742516,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001599939749962344,
+      "loss": 1.3213,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8123640183059494,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001599789124868203,
+      "loss": 1.3919,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8126641158376472,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015996384997740622,
+      "loss": 1.2439,
+      "step": 2708
+    },
+    {
+      "epoch": 0.812964213369345,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001599487874679922,
+      "loss": 1.2689,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8132643109010428,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001599337249585781,
+      "loss": 1.345,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8135644084327407,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015991866244916404,
+      "loss": 1.2131,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8138645059644385,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00015990359993974998,
+      "loss": 1.5219,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8141646034961363,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001598885374303359,
+      "loss": 1.3164,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8144647010278341,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015987347492092184,
+      "loss": 1.3533,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8147647985595319,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015985841241150778,
+      "loss": 1.3267,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8150648960912297,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001598433499020937,
+      "loss": 1.3811,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8153649936229275,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015982828739267963,
+      "loss": 1.3447,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8156650911546253,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015981322488326554,
+      "loss": 1.4951,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8159651886863231,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001597981623738515,
+      "loss": 1.4477,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8162652862180209,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015978309986443742,
+      "loss": 1.3829,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8165653837497187,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015976803735502334,
+      "loss": 1.0778,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8168654812814165,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001597529748456093,
+      "loss": 1.4347,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8171655788131142,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015973791233619522,
+      "loss": 1.288,
+      "step": 2723
+    },
+    {
+      "epoch": 0.817465676344812,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015972284982678113,
+      "loss": 1.3615,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8177657738765098,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001597077873173671,
+      "loss": 1.4657,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8180658714082076,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000159692724807953,
+      "loss": 1.2755,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8183659689399054,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015967766229853895,
+      "loss": 1.3648,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8186660664716032,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001596625997891249,
+      "loss": 1.375,
+      "step": 2728
+    },
+    {
+      "epoch": 0.818966164003301,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001596475372797108,
+      "loss": 1.0844,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8192662615349989,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015963247477029675,
+      "loss": 1.4108,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8195663590666967,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00015961741226088266,
+      "loss": 1.2291,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8198664565983945,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001596023497514686,
+      "loss": 1.2477,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8201665541300923,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015958728724205454,
+      "loss": 1.3004,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8204666516617901,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015957222473264045,
+      "loss": 1.1984,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8207667491934879,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001595571622232264,
+      "loss": 1.3007,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8210668467251857,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015954209971381233,
+      "loss": 1.2744,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8213669442568835,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015952703720439825,
+      "loss": 1.4376,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8216670417885813,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001595119746949842,
+      "loss": 1.3306,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8219671393202791,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015949691218557013,
+      "loss": 1.2796,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8222672368519769,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015948184967615604,
+      "loss": 1.202,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8225673343836747,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015946678716674198,
+      "loss": 1.2287,
+      "step": 2741
+    },
+    {
+      "epoch": 0.8228674319153725,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015945172465732792,
+      "loss": 1.2571,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8231675294470703,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015943666214791383,
+      "loss": 1.2728,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8234676269787681,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015942159963849977,
+      "loss": 1.3113,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8237677245104659,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00015940653712908571,
+      "loss": 1.3675,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8240678220421637,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015939147461967165,
+      "loss": 1.2952,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8243679195738615,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015937641211025757,
+      "loss": 1.3293,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8246680171055593,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001593613496008435,
+      "loss": 1.3545,
+      "step": 2748
+    },
+    {
+      "epoch": 0.824968114637257,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015934628709142945,
+      "loss": 1.4297,
+      "step": 2749
+    },
+    {
+      "epoch": 0.825268212168955,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015933122458201536,
+      "loss": 1.4081,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8255683097006528,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001593161620726013,
+      "loss": 1.4732,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8258684072323506,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015930109956318724,
+      "loss": 1.2257,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8261685047640484,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015928603705377316,
+      "loss": 1.3439,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8264686022957461,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001592709745443591,
+      "loss": 1.2557,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8267686998274439,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015925591203494504,
+      "loss": 1.3251,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8270687973591417,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015924084952553095,
+      "loss": 1.2133,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8273688948908395,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001592257870161169,
+      "loss": 1.2834,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8276689924225373,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00015921072450670283,
+      "loss": 1.4182,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8279690899542351,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015919566199728874,
+      "loss": 1.3136,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8282691874859329,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015918059948787468,
+      "loss": 1.2912,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8285692850176307,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015916553697846062,
+      "loss": 1.3347,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8288693825493285,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015915047446904654,
+      "loss": 1.2097,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8291694800810263,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015913541195963248,
+      "loss": 1.2673,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8294695776127241,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015912034945021842,
+      "loss": 1.3591,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8297696751444219,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015910528694080436,
+      "loss": 1.2772,
+      "step": 2765
+    },
+    {
+      "epoch": 0.8300697726761197,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00015909022443139027,
+      "loss": 1.5034,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8303698702078175,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001590751619219762,
+      "loss": 1.413,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8306699677395153,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015906009941256215,
+      "loss": 1.2675,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8309700652712132,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015904503690314806,
+      "loss": 1.3063,
+      "step": 2769
+    },
+    {
+      "epoch": 0.831270162802911,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.000159029974393734,
+      "loss": 1.373,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8315702603346088,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00015901491188431995,
+      "loss": 1.2663,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8318703578663066,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015899984937490586,
+      "loss": 1.2567,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8321704553980044,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001589847868654918,
+      "loss": 1.3289,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8324705529297022,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015896972435607774,
+      "loss": 1.2979,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8327706504614,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015895466184666365,
+      "loss": 1.2304,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8330707479930978,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001589395993372496,
+      "loss": 1.2451,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8333708455247956,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015892453682783553,
+      "loss": 1.3133,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8336709430564934,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015890947431842145,
+      "loss": 1.4503,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8339710405881912,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015889441180900739,
+      "loss": 1.5183,
+      "step": 2779
+    },
+    {
+      "epoch": 0.834271138119889,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00015887934929959333,
+      "loss": 1.5375,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8345712356515868,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00015886428679017927,
+      "loss": 1.2736,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8348713331832845,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00015884922428076518,
+      "loss": 1.2476,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8351714307149823,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015883416177135112,
+      "loss": 1.4497,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8354715282466801,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00015881909926193706,
+      "loss": 1.3973,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8357716257783779,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015880403675252297,
+      "loss": 1.2988,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8360717233100757,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015878897424310891,
+      "loss": 1.2389,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8363718208417735,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015877391173369485,
+      "loss": 1.3063,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8366719183734713,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015875884922428077,
+      "loss": 1.4043,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8369720159051692,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015874378671486668,
+      "loss": 1.4215,
+      "step": 2789
+    },
+    {
+      "epoch": 0.837272113436867,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015872872420545265,
+      "loss": 1.3593,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8375722109685648,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00015871366169603856,
+      "loss": 1.0645,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8378723085002626,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001586985991866245,
+      "loss": 1.3344,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8381724060319604,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015868353667721044,
+      "loss": 1.4025,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8384725035636582,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015866847416779636,
+      "loss": 1.1842,
+      "step": 2794
+    },
+    {
+      "epoch": 0.838772601095356,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001586534116583823,
+      "loss": 1.3445,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8390726986270538,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015863834914896824,
+      "loss": 1.3102,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8393727961587516,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015862328663955415,
+      "loss": 1.4008,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8396728936904494,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001586082241301401,
+      "loss": 1.3777,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8399729912221472,
+      "grad_norm": 1.125,
+      "learning_rate": 0.000158593161620726,
+      "loss": 1.19,
+      "step": 2799
+    },
+    {
+      "epoch": 0.840273088753845,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015857809911131197,
+      "loss": 1.1533,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8405731862855428,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015856303660189788,
+      "loss": 1.2956,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8408732838172406,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001585479740924838,
+      "loss": 1.2912,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8411733813489384,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015853291158306976,
+      "loss": 1.1514,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8414734788806362,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00015851784907365568,
+      "loss": 1.3142,
+      "step": 2804
+    },
+    {
+      "epoch": 0.841773576412334,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001585027865642416,
+      "loss": 1.3295,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8420736739440318,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015848772405482756,
+      "loss": 1.4288,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8423737714757296,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00015847266154541347,
+      "loss": 1.4103,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8426738690074275,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001584575990359994,
+      "loss": 1.3713,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8429739665391253,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015844253652658535,
+      "loss": 1.1889,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8432740640708231,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015842747401717126,
+      "loss": 1.311,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8435741616025209,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001584124115077572,
+      "loss": 1.4402,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8438742591342187,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015839734899834312,
+      "loss": 1.252,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8441743566659164,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015838228648892906,
+      "loss": 1.315,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8444744541976142,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.000158367223979515,
+      "loss": 1.4532,
+      "step": 2814
+    },
+    {
+      "epoch": 0.844774551729312,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001583521614701009,
+      "loss": 1.2206,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8450746492610098,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015833709896068685,
+      "loss": 1.4517,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8453747467927076,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001583220364512728,
+      "loss": 1.3986,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8456748443244054,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001583069739418587,
+      "loss": 1.2857,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8459749418561032,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00015829191143244467,
+      "loss": 1.3342,
+      "step": 2819
+    },
+    {
+      "epoch": 0.846275039387801,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001582768489230306,
+      "loss": 1.3608,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8465751369194988,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001582617864136165,
+      "loss": 1.276,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8468752344511966,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015824672390420247,
+      "loss": 1.2275,
+      "step": 2822
+    },
+    {
+      "epoch": 0.8471753319828944,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015823166139478838,
+      "loss": 1.3776,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8474754295145922,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001582165988853743,
+      "loss": 1.443,
+      "step": 2824
+    },
+    {
+      "epoch": 0.84777552704629,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015820153637596023,
+      "loss": 1.12,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8480756245779878,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015818647386654617,
+      "loss": 1.363,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8483757221096856,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015817141135713211,
+      "loss": 1.1788,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8486758196413835,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015815634884771803,
+      "loss": 1.3076,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8489759171730813,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015814128633830397,
+      "loss": 1.3115,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8492760147047791,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.0001581262238288899,
+      "loss": 1.636,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8495761122364769,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015811116131947582,
+      "loss": 1.3405,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8498762097681747,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015809609881006176,
+      "loss": 1.3452,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8501763072998725,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001580810363006477,
+      "loss": 1.4867,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8504764048315703,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015806597379123362,
+      "loss": 1.3825,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8507765023632681,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00015805091128181958,
+      "loss": 1.2774,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8510765998949659,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001580358487724055,
+      "loss": 1.2353,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8513766974266637,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001580207862629914,
+      "loss": 1.3207,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8516767949583615,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015800572375357735,
+      "loss": 1.3922,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8519768924900593,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001579906612441633,
+      "loss": 1.4323,
+      "step": 2839
+    },
+    {
+      "epoch": 0.852276990021757,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001579755987347492,
+      "loss": 1.3938,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8525770875534548,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015796053622533514,
+      "loss": 1.2989,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8528771850851526,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00015794547371592108,
+      "loss": 1.4249,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8531772826168504,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000157930411206507,
+      "loss": 1.3279,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8534773801485482,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00015791534869709294,
+      "loss": 1.4211,
+      "step": 2844
+    },
+    {
+      "epoch": 0.853777477680246,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015790028618767888,
+      "loss": 1.12,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8540775752119438,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00015788522367826482,
+      "loss": 1.5178,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8543776727436417,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015787016116885073,
+      "loss": 1.3678,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8546777702753395,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015785509865943667,
+      "loss": 1.3119,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8549778678070373,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001578400361500226,
+      "loss": 1.2802,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8552779653387351,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015782497364060852,
+      "loss": 1.3966,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8555780628704329,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015780991113119446,
+      "loss": 1.4982,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8558781604021307,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001577948486217804,
+      "loss": 1.3462,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8561782579338285,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015777978611236632,
+      "loss": 1.2369,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8564783554655263,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015776472360295226,
+      "loss": 1.2343,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8567784529972241,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001577496610935382,
+      "loss": 1.2686,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8570785505289219,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001577345985841241,
+      "loss": 1.6219,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8573786480606197,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015771953607471005,
+      "loss": 1.4134,
+      "step": 2857
+    },
+    {
+      "epoch": 0.8576787455923175,
+      "grad_norm": 1.25,
+      "learning_rate": 0.000157704473565296,
+      "loss": 1.1966,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8579788431240153,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001576894110558819,
+      "loss": 1.4379,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8582789406557131,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015767434854646785,
+      "loss": 1.3265,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8585790381874109,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001576592860370538,
+      "loss": 1.4278,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8588791357191087,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015764422352763973,
+      "loss": 1.4928,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8591792332508065,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015762916101822564,
+      "loss": 1.0484,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8594793307825043,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00015761409850881158,
+      "loss": 1.3931,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8597794283142021,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015759903599939752,
+      "loss": 1.2861,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8600795258459,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00015758397348998343,
+      "loss": 1.2333,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8603796233775978,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015756891098056937,
+      "loss": 1.3269,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8606797209092956,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015755384847115531,
+      "loss": 1.2539,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8609798184409934,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00015753878596174123,
+      "loss": 1.3467,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8612799159726912,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00015752372345232714,
+      "loss": 1.2532,
+      "step": 2870
+    },
+    {
+      "epoch": 0.861580013504389,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001575086609429131,
+      "loss": 1.2299,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8618801110360867,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015749359843349902,
+      "loss": 1.2051,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8621802085677845,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00015747853592408496,
+      "loss": 1.3444,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8624803060994823,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001574634734146709,
+      "loss": 1.3147,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8627804036311801,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015744841090525682,
+      "loss": 1.3087,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8630805011628779,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015743334839584276,
+      "loss": 1.4173,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8633805986945757,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001574182858864287,
+      "loss": 1.3403,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8636806962262735,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001574032233770146,
+      "loss": 1.2317,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8639807937579713,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015738816086760055,
+      "loss": 1.408,
+      "step": 2879
+    },
+    {
+      "epoch": 0.8642808912896691,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001573730983581865,
+      "loss": 1.1982,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8645809888213669,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015735803584877243,
+      "loss": 1.1833,
+      "step": 2881
+    },
+    {
+      "epoch": 0.8648810863530647,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015734297333935834,
+      "loss": 1.4232,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8651811838847625,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015732791082994426,
+      "loss": 1.2107,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8654812814164603,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015731284832053022,
+      "loss": 1.2726,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8657813789481581,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015729778581111614,
+      "loss": 1.4095,
+      "step": 2885
+    },
+    {
+      "epoch": 0.866081476479856,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015728272330170205,
+      "loss": 1.3007,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8663815740115538,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015726766079228802,
+      "loss": 1.254,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8666816715432516,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015725259828287393,
+      "loss": 1.3154,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8669817690749494,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015723753577345987,
+      "loss": 1.1851,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8672818666066472,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001572224732640458,
+      "loss": 1.1795,
+      "step": 2890
+    },
+    {
+      "epoch": 0.867581964138345,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015720741075463172,
+      "loss": 1.2011,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8678820616700428,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015719234824521767,
+      "loss": 1.1736,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8681821592017406,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001571772857358036,
+      "loss": 1.2723,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8684822567334384,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015716222322638952,
+      "loss": 1.2732,
+      "step": 2894
+    },
+    {
+      "epoch": 0.8687823542651362,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015714716071697546,
+      "loss": 1.309,
+      "step": 2895
+    },
+    {
+      "epoch": 0.869082451796834,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00015713209820756137,
+      "loss": 1.4615,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8693825493285318,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001571170356981473,
+      "loss": 1.2697,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8696826468602296,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015710197318873325,
+      "loss": 1.3499,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8699827443919274,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015708691067931917,
+      "loss": 1.3952,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8702828419236252,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015707184816990513,
+      "loss": 1.2633,
+      "step": 2900
+    },
+    {
+      "epoch": 0.870582939455323,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015705678566049105,
+      "loss": 1.1553,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8708830369870207,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015704172315107696,
+      "loss": 1.2606,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8711831345187185,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015702666064166293,
+      "loss": 1.3165,
+      "step": 2903
+    },
+    {
+      "epoch": 0.8714832320504163,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00015701159813224884,
+      "loss": 1.2373,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8717833295821142,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015699653562283475,
+      "loss": 1.3337,
+      "step": 2905
+    },
+    {
+      "epoch": 0.872083427113812,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001569814731134207,
+      "loss": 1.4984,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8723835246455098,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015696641060400663,
+      "loss": 1.3207,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8726836221772076,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015695134809459257,
+      "loss": 1.2759,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8729837197089054,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001569362855851785,
+      "loss": 1.4313,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8732838172406032,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015692122307576443,
+      "loss": 1.2051,
+      "step": 2910
+    },
+    {
+      "epoch": 0.873583914772301,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015690616056635037,
+      "loss": 1.2296,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8738840123039988,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015689109805693628,
+      "loss": 1.2973,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8741841098356966,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015687603554752222,
+      "loss": 1.2607,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8744842073673944,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015686097303810816,
+      "loss": 1.2773,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8747843048990922,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00015684591052869408,
+      "loss": 1.5187,
+      "step": 2915
+    },
+    {
+      "epoch": 0.87508440243079,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015683084801928004,
+      "loss": 1.2379,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8753844999624878,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00015681578550986596,
+      "loss": 1.1631,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8756845974941856,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015680072300045187,
+      "loss": 1.2932,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8759846950258834,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001567856604910378,
+      "loss": 1.3215,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8762847925575812,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015677059798162375,
+      "loss": 1.3369,
+      "step": 2920
+    },
+    {
+      "epoch": 0.876584890089279,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015675553547220966,
+      "loss": 1.3909,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8768849876209768,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001567404729627956,
+      "loss": 1.2348,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8771850851526746,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015672541045338154,
+      "loss": 1.1187,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8774851826843724,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015671034794396746,
+      "loss": 1.3818,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8777852802160703,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001566952854345534,
+      "loss": 1.2405,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8780853777477681,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015668022292513934,
+      "loss": 1.2924,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8783854752794659,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015666516041572528,
+      "loss": 1.2966,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8786855728111637,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001566500979063112,
+      "loss": 1.219,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8789856703428615,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015663503539689713,
+      "loss": 1.4048,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8792857678745593,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015661997288748307,
+      "loss": 1.298,
+      "step": 2930
+    },
+    {
+      "epoch": 0.879585865406257,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00015660491037806898,
+      "loss": 1.5569,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8798859629379548,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015658984786865493,
+      "loss": 1.3185,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8801860604696526,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015657478535924087,
+      "loss": 1.3806,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8804861580013504,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015655972284982678,
+      "loss": 1.3427,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8807862555330482,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015654466034041272,
+      "loss": 1.3269,
+      "step": 2935
+    },
+    {
+      "epoch": 0.881086353064746,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015652959783099866,
+      "loss": 1.2887,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8813864505964438,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00015651453532158457,
+      "loss": 1.4228,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8816865481281416,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001564994728121705,
+      "loss": 1.3216,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8819866456598394,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015648441030275645,
+      "loss": 1.3772,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8822867431915372,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015646934779334237,
+      "loss": 1.3433,
+      "step": 2940
+    },
+    {
+      "epoch": 0.882586840723235,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001564542852839283,
+      "loss": 1.3045,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8828869382549328,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015643922277451425,
+      "loss": 1.378,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8831870357866306,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001564241602651002,
+      "loss": 1.2912,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8834871333183285,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001564090977556861,
+      "loss": 1.3545,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8837872308500263,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015639403524627204,
+      "loss": 1.3692,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8840873283817241,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015637897273685798,
+      "loss": 1.3644,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8843874259134219,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001563639102274439,
+      "loss": 1.3164,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8846875234451197,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015634884771802983,
+      "loss": 1.4904,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8849876209768175,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015633378520861577,
+      "loss": 1.4089,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8852877185085153,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001563187226992017,
+      "loss": 1.145,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8855878160402131,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015630366018978763,
+      "loss": 1.3659,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8858879135719109,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015628859768037357,
+      "loss": 1.3516,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8861880111036087,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00015627353517095948,
+      "loss": 1.3006,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8864881086353065,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00015625847266154542,
+      "loss": 1.4191,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8867882061670043,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015624341015213136,
+      "loss": 1.2818,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8870883036987021,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015622834764271728,
+      "loss": 1.371,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8873884012303999,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015621328513330322,
+      "loss": 1.297,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8876884987620977,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015619822262388916,
+      "loss": 1.184,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8879885962937955,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015618316011447507,
+      "loss": 1.3297,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8882886938254932,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.000156168097605061,
+      "loss": 1.3705,
+      "step": 2960
+    },
+    {
+      "epoch": 0.888588791357191,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015615303509564695,
+      "loss": 1.4826,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001561379725862329,
+      "loss": 1.2576,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8891889864205866,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001561229100768188,
+      "loss": 1.3966,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8894890839522845,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00015610784756740474,
+      "loss": 1.3998,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8897891814839823,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015609278505799068,
+      "loss": 1.2413,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8900892790156801,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001560777225485766,
+      "loss": 1.3101,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8903893765473779,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.0001560626600391625,
+      "loss": 1.3883,
+      "step": 2967
+    },
+    {
+      "epoch": 0.8906894740790757,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015604759752974848,
+      "loss": 1.4513,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8909895716107735,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001560325350203344,
+      "loss": 1.3054,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8912896691424713,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015601747251092033,
+      "loss": 1.3815,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8915897666741691,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015600241000150627,
+      "loss": 1.3399,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8918898642058669,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015598734749209218,
+      "loss": 1.2958,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8921899617375647,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015597228498267813,
+      "loss": 1.2362,
+      "step": 2973
+    },
+    {
+      "epoch": 0.8924900592692625,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00015595722247326407,
+      "loss": 1.3504,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8927901568009603,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015594215996384998,
+      "loss": 1.4291,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8930902543326581,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015592709745443592,
+      "loss": 1.1895,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8933903518643559,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015591203494502183,
+      "loss": 1.2693,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8936904493960537,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015589697243560777,
+      "loss": 1.5826,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8939905469277515,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001558819099261937,
+      "loss": 1.2347,
+      "step": 2979
+    },
+    {
+      "epoch": 0.8942906444594493,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015586684741677963,
+      "loss": 1.428,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8945907419911471,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001558517849073656,
+      "loss": 1.491,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8948908395228449,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001558367223979515,
+      "loss": 1.3497,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8951909370545428,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015582165988853742,
+      "loss": 1.2881,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8954910345862406,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001558065973791234,
+      "loss": 1.211,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8957911321179384,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001557915348697093,
+      "loss": 1.3923,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8960912296496362,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015577647236029521,
+      "loss": 1.2636,
+      "step": 2986
+    },
+    {
+      "epoch": 0.896391327181334,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015576140985088118,
+      "loss": 1.2306,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8966914247130318,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001557463473414671,
+      "loss": 1.3334,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8969915222447296,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015573128483205303,
+      "loss": 1.2956,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8972916197764274,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015571622232263895,
+      "loss": 1.2826,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8975917173081251,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001557011598132249,
+      "loss": 1.1142,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8978918148398229,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00015568609730381083,
+      "loss": 1.435,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8981919123715207,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015567103479439674,
+      "loss": 1.2286,
+      "step": 2993
+    },
+    {
+      "epoch": 0.8984920099032185,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015565597228498268,
+      "loss": 1.4797,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8987921074349163,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015564090977556862,
+      "loss": 1.2522,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8990922049666141,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015562584726615454,
+      "loss": 1.2248,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8993923024983119,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001556107847567405,
+      "loss": 1.3291,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8996924000300097,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015559572224732642,
+      "loss": 1.472,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8999924975617075,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015558065973791233,
+      "loss": 1.2916,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9002925950934053,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001555655972284983,
+      "loss": 1.0412,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9005926926251031,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001555505347190842,
+      "loss": 1.275,
+      "step": 3001
+    },
+    {
+      "epoch": 0.900892790156801,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015553547220967012,
+      "loss": 1.3011,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9011928876884988,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015552040970025606,
+      "loss": 1.2443,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9014929852201966,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.000155505347190842,
+      "loss": 1.2414,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9017930827518944,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015549028468142792,
+      "loss": 1.3269,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9020931802835922,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015547522217201386,
+      "loss": 1.282,
+      "step": 3006
+    },
+    {
+      "epoch": 0.90239327781529,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001554601596625998,
+      "loss": 1.4143,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9026933753469878,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015544509715318574,
+      "loss": 1.3809,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9029934728786856,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015543003464377165,
+      "loss": 1.2346,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9032935704103834,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001554149721343576,
+      "loss": 1.3283,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9035936679420812,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015539990962494353,
+      "loss": 1.2195,
+      "step": 3011
+    },
+    {
+      "epoch": 0.903893765473779,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015538484711552944,
+      "loss": 1.0746,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9041938630054768,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015536978460611539,
+      "loss": 1.2805,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9044939605371746,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015535472209670133,
+      "loss": 1.406,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9047940580688724,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015533965958728724,
+      "loss": 1.2393,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9050941556005702,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015532459707787318,
+      "loss": 1.186,
+      "step": 3016
+    },
+    {
+      "epoch": 0.905394253132268,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015530953456845912,
+      "loss": 1.3654,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9056943506639658,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015529447205904503,
+      "loss": 1.3558,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9059944481956635,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015527940954963097,
+      "loss": 1.4617,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9062945457273613,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001552643470402169,
+      "loss": 1.2195,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9065946432590591,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015524928453080283,
+      "loss": 1.3107,
+      "step": 3021
+    },
+    {
+      "epoch": 0.906894740790757,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015523422202138877,
+      "loss": 1.1979,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9071948383224548,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001552191595119747,
+      "loss": 1.2048,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9074949358541526,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015520409700256065,
+      "loss": 1.233,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9077950333858504,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015518903449314656,
+      "loss": 1.3547,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9080951309175482,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001551739719837325,
+      "loss": 1.2749,
+      "step": 3026
+    },
+    {
+      "epoch": 0.908395228449246,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015515890947431844,
+      "loss": 1.2458,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9086953259809438,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015514384696490435,
+      "loss": 1.2547,
+      "step": 3028
+    },
+    {
+      "epoch": 0.9089954235126416,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001551287844554903,
+      "loss": 1.3921,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9092955210443394,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015511372194607623,
+      "loss": 1.1641,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9095956185760372,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015509865943666215,
+      "loss": 1.5032,
+      "step": 3031
+    },
+    {
+      "epoch": 0.909895716107735,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001550835969272481,
+      "loss": 1.2456,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9101958136394328,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015506853441783403,
+      "loss": 1.3433,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9104959111711306,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015505347190841994,
+      "loss": 1.3368,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9107960087028284,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015503840939900588,
+      "loss": 1.271,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9110961062345262,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015502334688959182,
+      "loss": 1.3345,
+      "step": 3036
+    },
+    {
+      "epoch": 0.911396203766224,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015500828438017774,
+      "loss": 1.4003,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9116963012979218,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015499322187076368,
+      "loss": 1.2034,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9119963988296196,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00015497815936134962,
+      "loss": 1.2672,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9122964963613174,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015496309685193553,
+      "loss": 1.2619,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9125965938930153,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015494803434252147,
+      "loss": 1.3641,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9128966914247131,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001549329718331074,
+      "loss": 1.2224,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9131967889564109,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015491790932369335,
+      "loss": 1.3068,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9134968864881087,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015490284681427926,
+      "loss": 1.3287,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9137969840198065,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001548877843048652,
+      "loss": 1.2375,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9140970815515043,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015487272179545114,
+      "loss": 1.3555,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9143971790832021,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015485765928603706,
+      "loss": 1.3548,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9146972766148999,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015484259677662297,
+      "loss": 1.3697,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9149973741465977,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00015482753426720894,
+      "loss": 1.2677,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9152974716782954,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015481247175779485,
+      "loss": 1.1984,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9155975692099932,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001547974092483808,
+      "loss": 1.4955,
+      "step": 3051
+    },
+    {
+      "epoch": 0.915897666741691,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015478234673896673,
+      "loss": 1.3294,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9161977642733888,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015476728422955264,
+      "loss": 1.2205,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9164978618050866,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015475222172013859,
+      "loss": 1.3841,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9167979593367844,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015473715921072453,
+      "loss": 1.2115,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9170980568684822,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015472209670131044,
+      "loss": 1.3146,
+      "step": 3056
+    },
+    {
+      "epoch": 0.91739815440018,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015470703419189638,
+      "loss": 1.2619,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9176982519318778,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00015469197168248232,
+      "loss": 1.3532,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9179983494635756,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015467690917306823,
+      "loss": 1.3594,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9182984469952734,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00015466184666365417,
+      "loss": 1.5091,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9185985445269713,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015464678415424009,
+      "loss": 1.228,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9188986420586691,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015463172164482605,
+      "loss": 1.5384,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9191987395903669,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015461665913541197,
+      "loss": 1.2112,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9194988371220647,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015460159662599788,
+      "loss": 1.277,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9197989346537625,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015458653411658385,
+      "loss": 1.3168,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9200990321854603,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015457147160716976,
+      "loss": 1.2745,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9203991297171581,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015455640909775567,
+      "loss": 1.3328,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9206992272488559,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015454134658834164,
+      "loss": 1.2191,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9209993247805537,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015452628407892755,
+      "loss": 1.4223,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9212994223122515,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001545112215695135,
+      "loss": 1.1868,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9215995198439493,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015449615906009943,
+      "loss": 1.3298,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9218996173756471,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015448109655068535,
+      "loss": 1.4443,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9221997149073449,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001544660340412713,
+      "loss": 1.1642,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9224998124390427,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001544509715318572,
+      "loss": 1.4334,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9227999099707405,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015443590902244314,
+      "loss": 1.2557,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9231000075024383,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015442084651302908,
+      "loss": 1.308,
+      "step": 3076
+    },
+    {
+      "epoch": 0.923400105034136,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.000154405784003615,
+      "loss": 1.2138,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9237002025658338,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015439072149420096,
+      "loss": 1.3744,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9240003000975316,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015437565898478688,
+      "loss": 1.1994,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9243003976292296,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001543605964753728,
+      "loss": 1.3968,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9246004951609273,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015434553396595876,
+      "loss": 1.2832,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9249005926926251,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015433047145654467,
+      "loss": 1.3684,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9252006902243229,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015431540894713058,
+      "loss": 1.2826,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9255007877560207,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015430034643771652,
+      "loss": 1.2819,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9258008852877185,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015428528392830246,
+      "loss": 1.4134,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9261009828194163,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015427022141888838,
+      "loss": 1.249,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9264010803511141,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015425515890947432,
+      "loss": 1.2146,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9267011778828119,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015424009640006026,
+      "loss": 1.3082,
+      "step": 3088
+    },
+    {
+      "epoch": 0.9270012754145097,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001542250338906462,
+      "loss": 1.5657,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9273013729462075,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001542099713812321,
+      "loss": 1.2857,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9276014704779053,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015419490887181805,
+      "loss": 1.3934,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9279015680096031,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.000154179846362404,
+      "loss": 1.4583,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9282016655413009,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001541647838529899,
+      "loss": 1.244,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9285017630729987,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00015414972134357585,
+      "loss": 1.7227,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9288018606046965,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015413465883416179,
+      "loss": 1.1422,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9291019581363943,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001541195963247477,
+      "loss": 1.2743,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9294020556680921,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015410453381533364,
+      "loss": 1.2442,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9297021531997899,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015408947130591958,
+      "loss": 1.3463,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9300022507314877,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001540744087965055,
+      "loss": 1.291,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9303023482631856,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015405934628709143,
+      "loss": 1.4495,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9306024457948834,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015404428377767737,
+      "loss": 1.181,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9309025433265812,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015402922126826329,
+      "loss": 1.2473,
+      "step": 3102
+    },
+    {
+      "epoch": 0.931202640858279,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015401415875884923,
+      "loss": 1.4011,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9315027383899768,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015399909624943517,
+      "loss": 1.1944,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9318028359216746,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001539840337400211,
+      "loss": 1.3614,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9321029334533724,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015396897123060702,
+      "loss": 1.2566,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9324030309850702,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015395390872119296,
+      "loss": 1.2624,
+      "step": 3107
+    },
+    {
+      "epoch": 0.932703128516768,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001539388462117789,
+      "loss": 1.223,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9330032260484658,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015392378370236481,
+      "loss": 1.3454,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9333033235801635,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015390872119295075,
+      "loss": 1.3328,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9336034211118613,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001538936586835367,
+      "loss": 1.1984,
+      "step": 3111
+    },
+    {
+      "epoch": 0.9339035186435591,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001538785961741226,
+      "loss": 1.1032,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9342036161752569,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015386353366470855,
+      "loss": 1.3246,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9345037137069547,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001538484711552945,
+      "loss": 1.6474,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9348038112386525,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001538334086458804,
+      "loss": 1.2438,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9351039087703503,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015381834613646634,
+      "loss": 1.2052,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9354040063020481,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015380328362705228,
+      "loss": 1.1318,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9357041038337459,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001537882211176382,
+      "loss": 1.2239,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9360042013654438,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015377315860822414,
+      "loss": 1.3496,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9363042988971416,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015375809609881008,
+      "loss": 1.1736,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9366043964288394,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.000153743033589396,
+      "loss": 1.3692,
+      "step": 3121
+    },
+    {
+      "epoch": 0.9369044939605372,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00015372797107998193,
+      "loss": 1.4884,
+      "step": 3122
+    },
+    {
+      "epoch": 0.937204591492235,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00015371290857056787,
+      "loss": 1.3912,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9375046890239328,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001536978460611538,
+      "loss": 1.2286,
+      "step": 3124
+    },
+    {
+      "epoch": 0.9378047865556306,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015368278355173972,
+      "loss": 1.2853,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9381048840873284,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015366772104232566,
+      "loss": 1.2344,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9384049816190262,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.0001536526585329116,
+      "loss": 1.4996,
+      "step": 3127
+    },
+    {
+      "epoch": 0.938705079150724,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015363759602349752,
+      "loss": 1.3603,
+      "step": 3128
+    },
+    {
+      "epoch": 0.9390051766824218,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00015362253351408346,
+      "loss": 1.6359,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9393052742141196,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001536074710046694,
+      "loss": 1.4332,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9396053717458174,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001535924084952553,
+      "loss": 1.1578,
+      "step": 3131
+    },
+    {
+      "epoch": 0.9399054692775152,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015357734598584125,
+      "loss": 1.1858,
+      "step": 3132
+    },
+    {
+      "epoch": 0.940205566809213,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001535622834764272,
+      "loss": 1.3674,
+      "step": 3133
+    },
+    {
+      "epoch": 0.9405056643409108,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001535472209670131,
+      "loss": 1.4128,
+      "step": 3134
+    },
+    {
+      "epoch": 0.9408057618726086,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015353215845759905,
+      "loss": 1.4885,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9411058594043064,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015351709594818499,
+      "loss": 1.3787,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9414059569360042,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001535020334387709,
+      "loss": 1.6679,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9417060544677021,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015348697092935684,
+      "loss": 1.3852,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9420061519993999,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015347190841994278,
+      "loss": 1.31,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9423062495310977,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001534568459105287,
+      "loss": 1.1159,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9426063470627954,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015344178340111463,
+      "loss": 1.1408,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9429064445944932,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015342672089170055,
+      "loss": 1.2693,
+      "step": 3142
+    },
+    {
+      "epoch": 0.943206542126191,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001534116583822865,
+      "loss": 1.359,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9435066396578888,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015339659587287243,
+      "loss": 1.3865,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9438067371895866,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015338153336345834,
+      "loss": 1.3032,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9441068347212844,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001533664708540443,
+      "loss": 1.4202,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9444069322529822,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015335140834463022,
+      "loss": 1.2584,
+      "step": 3147
+    },
+    {
+      "epoch": 0.94470702978468,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015333634583521613,
+      "loss": 1.2318,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9450071273163778,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001533212833258021,
+      "loss": 1.4184,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9453072248480756,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015330622081638801,
+      "loss": 1.3873,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9456073223797734,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015329115830697395,
+      "loss": 1.2761,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9459074199114712,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001532760957975599,
+      "loss": 1.295,
+      "step": 3152
+    },
+    {
+      "epoch": 0.946207517443169,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001532610332881458,
+      "loss": 1.2752,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9465076149748668,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00015324597077873175,
+      "loss": 1.3081,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9468077125065646,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015323090826931766,
+      "loss": 1.2913,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9471078100382624,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001532158457599036,
+      "loss": 1.2706,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9474079075699602,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015320078325048954,
+      "loss": 1.4667,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9477080051016581,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.00015318572074107546,
+      "loss": 1.4606,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9480081026333559,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015317065823166142,
+      "loss": 1.433,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9483082001650537,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015315559572224734,
+      "loss": 1.245,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9486082976967515,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015314053321283325,
+      "loss": 1.3795,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9489083952284493,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015312547070341922,
+      "loss": 1.2142,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9492084927601471,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015311040819400513,
+      "loss": 1.4043,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9495085902918449,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015309534568459104,
+      "loss": 1.3198,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9498086878235427,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.000153080283175177,
+      "loss": 1.2182,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9501087853552405,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015306522066576292,
+      "loss": 1.4412,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9504088828869383,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015305015815634884,
+      "loss": 1.4214,
+      "step": 3167
+    },
+    {
+      "epoch": 0.950708980418636,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015303509564693478,
+      "loss": 1.1593,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9510090779503338,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015302003313752072,
+      "loss": 1.3052,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9513091754820316,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015300497062810666,
+      "loss": 1.4334,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9516092730137294,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00015298990811869257,
+      "loss": 1.2297,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9519093705454272,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001529748456092785,
+      "loss": 1.3714,
+      "step": 3172
+    },
+    {
+      "epoch": 0.952209468077125,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00015295978309986445,
+      "loss": 1.2986,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9525095656088228,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015294472059045036,
+      "loss": 1.1989,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9528096631405206,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001529296580810363,
+      "loss": 1.3853,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9531097606722184,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015291459557162225,
+      "loss": 1.207,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9534098582039163,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015289953306220816,
+      "loss": 1.3292,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9537099557356141,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001528844705527941,
+      "loss": 1.1768,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9540100532673119,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015286940804338004,
+      "loss": 1.306,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9543101507990097,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015285434553396595,
+      "loss": 1.0995,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9546102483307075,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001528392830245519,
+      "loss": 1.2927,
+      "step": 3181
+    },
+    {
+      "epoch": 0.9549103458624053,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015282422051513783,
+      "loss": 1.3044,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9552104433941031,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015280915800572375,
+      "loss": 1.2345,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9555105409258009,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001527940954963097,
+      "loss": 1.2689,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9558106384574987,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015277903298689563,
+      "loss": 1.1211,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9561107359891965,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015276397047748157,
+      "loss": 1.3926,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9564108335208943,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015274890796806748,
+      "loss": 1.276,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9567109310525921,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015273384545865342,
+      "loss": 1.3093,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9570110285842899,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015271878294923936,
+      "loss": 1.5502,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9573111261159877,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015270372043982527,
+      "loss": 1.3144,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9576112236476855,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015268865793041121,
+      "loss": 1.3148,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9579113211793833,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015267359542099715,
+      "loss": 1.1904,
+      "step": 3192
+    },
+    {
+      "epoch": 0.9582114187110811,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015265853291158307,
+      "loss": 1.2793,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9585115162427789,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.000152643470402169,
+      "loss": 1.3894,
+      "step": 3194
+    },
+    {
+      "epoch": 0.9588116137744767,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015262840789275495,
+      "loss": 1.3387,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9591117113061745,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015261334538334086,
+      "loss": 1.3235,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9594118088378724,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001525982828739268,
+      "loss": 1.4606,
+      "step": 3197
+    },
+    {
+      "epoch": 0.9597119063695702,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015258322036451274,
+      "loss": 1.2038,
+      "step": 3198
+    },
+    {
+      "epoch": 0.960012003901268,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015256815785509866,
+      "loss": 1.3242,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9603121014329657,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001525530953456846,
+      "loss": 1.2614,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9606121989646635,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015253803283627054,
+      "loss": 1.2792,
+      "step": 3201
+    },
+    {
+      "epoch": 0.9609122964963613,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015252297032685645,
+      "loss": 1.4208,
+      "step": 3202
+    },
+    {
+      "epoch": 0.9612123940280591,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001525079078174424,
+      "loss": 1.1942,
+      "step": 3203
+    },
+    {
+      "epoch": 0.9615124915597569,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00015249284530802833,
+      "loss": 1.1688,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9618125890914547,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015247778279861427,
+      "loss": 1.2786,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9621126866231525,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00015246272028920018,
+      "loss": 1.4389,
+      "step": 3206
+    },
+    {
+      "epoch": 0.9624127841548503,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015244765777978612,
+      "loss": 1.294,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9627128816865481,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015243259527037206,
+      "loss": 1.1931,
+      "step": 3208
+    },
+    {
+      "epoch": 0.9630129792182459,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015241753276095798,
+      "loss": 1.2546,
+      "step": 3209
+    },
+    {
+      "epoch": 0.9633130767499437,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015240247025154392,
+      "loss": 1.2928,
+      "step": 3210
+    },
+    {
+      "epoch": 0.9636131742816415,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015238740774212986,
+      "loss": 1.3797,
+      "step": 3211
+    },
+    {
+      "epoch": 0.9639132718133393,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015237234523271577,
+      "loss": 1.4895,
+      "step": 3212
+    },
+    {
+      "epoch": 0.9642133693450371,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001523572827233017,
+      "loss": 1.4223,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9645134668767349,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015234222021388765,
+      "loss": 1.243,
+      "step": 3214
+    },
+    {
+      "epoch": 0.9648135644084327,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015232715770447356,
+      "loss": 1.2401,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9651136619401306,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001523120951950595,
+      "loss": 1.3818,
+      "step": 3216
+    },
+    {
+      "epoch": 0.9654137594718284,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015229703268564545,
+      "loss": 1.4468,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9657138570035262,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015228197017623136,
+      "loss": 1.4207,
+      "step": 3218
+    },
+    {
+      "epoch": 0.966013954535224,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001522669076668173,
+      "loss": 1.3059,
+      "step": 3219
+    },
+    {
+      "epoch": 0.9663140520669218,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015225184515740324,
+      "loss": 1.4666,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9666141495986196,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015223678264798915,
+      "loss": 1.2132,
+      "step": 3221
+    },
+    {
+      "epoch": 0.9669142471303174,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001522217201385751,
+      "loss": 1.4499,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9672143446620152,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015220665762916103,
+      "loss": 1.2935,
+      "step": 3223
+    },
+    {
+      "epoch": 0.967514442193713,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00015219159511974697,
+      "loss": 1.257,
+      "step": 3224
+    },
+    {
+      "epoch": 0.9678145397254108,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001521765326103329,
+      "loss": 1.2961,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9681146372571086,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001521614701009188,
+      "loss": 1.275,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9684147347888064,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015214640759150477,
+      "loss": 1.3071,
+      "step": 3227
+    },
+    {
+      "epoch": 0.9687148323205041,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00015213134508209068,
+      "loss": 1.3322,
+      "step": 3228
+    },
+    {
+      "epoch": 0.969014929852202,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001521162825726766,
+      "loss": 1.2073,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9693150273838997,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015210122006326256,
+      "loss": 1.2727,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9696151249155975,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015208615755384847,
+      "loss": 1.4803,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9699152224472953,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015207109504443441,
+      "loss": 1.2938,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9702153199789931,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015205603253502035,
+      "loss": 1.3457,
+      "step": 3233
+    },
+    {
+      "epoch": 0.9705154175106909,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015204097002560627,
+      "loss": 1.4359,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9708155150423887,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001520259075161922,
+      "loss": 1.2108,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9711156125740866,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00015201084500677815,
+      "loss": 1.2169,
+      "step": 3236
+    },
+    {
+      "epoch": 0.9714157101057844,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015199578249736406,
+      "loss": 1.2122,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9717158076374822,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015198071998795,
+      "loss": 1.2408,
+      "step": 3238
+    },
+    {
+      "epoch": 0.97201590516918,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00015196565747853592,
+      "loss": 1.1479,
+      "step": 3239
+    },
+    {
+      "epoch": 0.9723160027008778,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015195059496912188,
+      "loss": 1.2285,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9726161002325756,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001519355324597078,
+      "loss": 1.3461,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9729161977642734,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001519204699502937,
+      "loss": 1.3693,
+      "step": 3242
+    },
+    {
+      "epoch": 0.9732162952959712,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015190540744087968,
+      "loss": 1.2446,
+      "step": 3243
+    },
+    {
+      "epoch": 0.973516392827669,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001518903449314656,
+      "loss": 1.3235,
+      "step": 3244
+    },
+    {
+      "epoch": 0.9738164903593668,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001518752824220515,
+      "loss": 1.2744,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9741165878910646,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00015186021991263747,
+      "loss": 1.3303,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9744166854227624,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015184515740322338,
+      "loss": 1.2841,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9747167829544602,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001518300948938093,
+      "loss": 1.2586,
+      "step": 3248
+    },
+    {
+      "epoch": 0.975016880486158,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00015181503238439524,
+      "loss": 1.188,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9753169780178558,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00015179996987498118,
+      "loss": 1.0491,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9756170755495536,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00015178490736556712,
+      "loss": 1.2157,
+      "step": 3251
+    },
+    {
+      "epoch": 0.9759171730812514,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00015176984485615303,
+      "loss": 1.1488,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9762172706129492,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015175478234673897,
+      "loss": 1.2419,
+      "step": 3253
+    },
+    {
+      "epoch": 0.976517368144647,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001517397198373249,
+      "loss": 1.1383,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9768174656763449,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015172465732791082,
+      "loss": 1.3823,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9771175632080427,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015170959481849677,
+      "loss": 1.3467,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9774176607397405,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001516945323090827,
+      "loss": 1.3585,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9777177582714383,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015167946979966862,
+      "loss": 1.2789,
+      "step": 3258
+    },
+    {
+      "epoch": 0.978017855803136,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015166440729025459,
+      "loss": 1.2802,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9783179533348338,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001516493447808405,
+      "loss": 1.3572,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9786180508665316,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001516342822714264,
+      "loss": 1.2226,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9789181483982294,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015161921976201235,
+      "loss": 1.4036,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9792182459299272,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001516041572525983,
+      "loss": 1.3359,
+      "step": 3263
+    },
+    {
+      "epoch": 0.979518343461625,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001515890947431842,
+      "loss": 1.2354,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9798184409933228,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015157403223377015,
+      "loss": 1.2926,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9801185385250206,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001515589697243561,
+      "loss": 1.312,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9804186360567184,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015154390721494203,
+      "loss": 1.2045,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9807187335884162,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015152884470552794,
+      "loss": 1.2783,
+      "step": 3268
+    },
+    {
+      "epoch": 0.981018831120114,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015151378219611388,
+      "loss": 1.2645,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9813189286518118,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015149871968669982,
+      "loss": 1.2277,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9816190261835096,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015148365717728573,
+      "loss": 1.2894,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9819191237152074,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00015146859466787167,
+      "loss": 1.2568,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9822192212469052,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015145353215845761,
+      "loss": 1.3047,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9825193187786031,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015143846964904353,
+      "loss": 1.2299,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9828194163103009,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00015142340713962947,
+      "loss": 1.3297,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9831195138419987,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001514083446302154,
+      "loss": 1.3538,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9834196113736965,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015139328212080132,
+      "loss": 1.262,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9837197089053943,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015137821961138726,
+      "loss": 1.4108,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9840198064370921,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001513631571019732,
+      "loss": 1.3683,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9843199039687899,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00015134809459255912,
+      "loss": 1.2118,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9846200015004877,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00015133303208314506,
+      "loss": 1.3833,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9849200990321855,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.000151317969573731,
+      "loss": 1.4059,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9852201965638833,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001513029070643169,
+      "loss": 1.3468,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9855202940955811,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00015128784455490285,
+      "loss": 1.1131,
+      "step": 3284
+    },
+    {
+      "epoch": 0.9858203916272789,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001512727820454888,
+      "loss": 1.253,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9861204891589767,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00015125771953607473,
+      "loss": 1.4786,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9864205866906745,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015124265702666064,
+      "loss": 1.2408,
+      "step": 3287
+    },
+    {
+      "epoch": 0.9867206842223722,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00015122759451724658,
+      "loss": 1.4596,
+      "step": 3288
+    },
+    {
+      "epoch": 0.98702078175407,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015121253200783252,
+      "loss": 1.3259,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9873208792857678,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00015119746949841844,
+      "loss": 1.2714,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9876209768174656,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00015118240698900438,
+      "loss": 1.3313,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9879210743491634,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015116734447959032,
+      "loss": 1.3759,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9882211718808612,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015115228197017623,
+      "loss": 1.2907,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9885212694125591,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015113721946076217,
+      "loss": 1.3222,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9888213669442569,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001511221569513481,
+      "loss": 1.2332,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9891214644759547,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015110709444193402,
+      "loss": 1.3737,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9894215620076525,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015109203193251997,
+      "loss": 1.241,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9897216595393503,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001510769694231059,
+      "loss": 1.2238,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9900217570710481,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00015106190691369182,
+      "loss": 1.1648,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9903218546027459,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015104684440427776,
+      "loss": 1.1457,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9906219521344437,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001510317818948637,
+      "loss": 1.3547,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9909220496661415,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001510167193854496,
+      "loss": 1.1766,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9912221471978393,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00015100165687603555,
+      "loss": 1.2454,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9915222447295371,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001509865943666215,
+      "loss": 1.3926,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9918223422612349,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015097153185720743,
+      "loss": 1.2143,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9921224397929327,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015095646934779335,
+      "loss": 1.223,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9924225373246305,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001509414068383793,
+      "loss": 1.4326,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9927226348563283,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015092634432896523,
+      "loss": 1.2948,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9930227323880261,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015091128181955114,
+      "loss": 1.2289,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9933228299197239,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015089621931013705,
+      "loss": 1.2895,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9936229274514217,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015088115680072302,
+      "loss": 1.2653,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9939230249831195,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015086609429130893,
+      "loss": 1.3337,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9942231225148174,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00015085103178189487,
+      "loss": 1.3752,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9945232200465152,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015083596927248081,
+      "loss": 1.1749,
+      "step": 3314
+    },
+    {
+      "epoch": 0.994823317578213,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015082090676306673,
+      "loss": 1.2589,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9951234151099108,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015080584425365267,
+      "loss": 1.2754,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9954235126416086,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001507907817442386,
+      "loss": 1.2801,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9957236101733064,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00015077571923482452,
+      "loss": 1.1689,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9960237077050041,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015076065672541046,
+      "loss": 1.3982,
+      "step": 3319
+    },
+    {
+      "epoch": 0.9963238052367019,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00015074559421599638,
+      "loss": 1.3245,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9966239027683997,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015073053170658234,
+      "loss": 1.2021,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9969240003000975,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015071546919716826,
+      "loss": 1.1952,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9972240978317953,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015070040668775417,
+      "loss": 1.334,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9975241953634931,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015068534417834014,
+      "loss": 1.2942,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9978242928951909,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015067028166892605,
+      "loss": 1.2871,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9981243904268887,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015065521915951196,
+      "loss": 1.3793,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9984244879585865,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015064015665009793,
+      "loss": 1.2587,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9987245854902843,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015062509414068384,
+      "loss": 1.3355,
+      "step": 3328
+    },
+    {
+      "epoch": 0.9990246830219821,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015061003163126976,
+      "loss": 1.3061,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9993247805536799,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015059496912185572,
+      "loss": 1.4642,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9996248780853777,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015057990661244164,
+      "loss": 1.2893,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9999249756170755,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00015056484410302758,
+      "loss": 1.3543,
+      "step": 3332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001505497815936135,
+      "loss": 1.243,
+      "step": 3333
+    },
+    {
+      "epoch": 1.0003000975316978,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00015053471908419943,
+      "loss": 1.2039,
+      "step": 3334
+    },
+    {
+      "epoch": 1.0006001950633956,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015051965657478537,
+      "loss": 1.0527,
+      "step": 3335
+    },
+    {
+      "epoch": 1.0009002925950934,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00015050459406537128,
+      "loss": 1.0187,
+      "step": 3336
+    },
+    {
+      "epoch": 1.0012003901267912,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015048953155595723,
+      "loss": 0.9977,
+      "step": 3337
+    },
+    {
+      "epoch": 1.001500487658489,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00015047446904654317,
+      "loss": 1.0457,
+      "step": 3338
+    },
+    {
+      "epoch": 1.0018005851901868,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015045940653712908,
+      "loss": 1.1836,
+      "step": 3339
+    },
+    {
+      "epoch": 1.0021006827218846,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00015044434402771505,
+      "loss": 1.1175,
+      "step": 3340
+    },
+    {
+      "epoch": 1.0024007802535824,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00015042928151830096,
+      "loss": 1.1357,
+      "step": 3341
+    },
+    {
+      "epoch": 1.0027008777852802,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00015041421900888687,
+      "loss": 0.9933,
+      "step": 3342
+    },
+    {
+      "epoch": 1.003000975316978,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00015039915649947284,
+      "loss": 1.0579,
+      "step": 3343
+    },
+    {
+      "epoch": 1.0033010728486758,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015038409399005875,
+      "loss": 1.0267,
+      "step": 3344
+    },
+    {
+      "epoch": 1.0036011703803736,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00015036903148064467,
+      "loss": 1.1227,
+      "step": 3345
+    },
+    {
+      "epoch": 1.0039012679120713,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001503539689712306,
+      "loss": 1.0639,
+      "step": 3346
+    },
+    {
+      "epoch": 1.0042013654437691,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00015033890646181655,
+      "loss": 0.9731,
+      "step": 3347
+    },
+    {
+      "epoch": 1.004501462975467,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001503238439524025,
+      "loss": 1.0959,
+      "step": 3348
+    },
+    {
+      "epoch": 1.0048015605071647,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001503087814429884,
+      "loss": 1.0443,
+      "step": 3349
+    },
+    {
+      "epoch": 1.0051016580388625,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00015029371893357434,
+      "loss": 1.0437,
+      "step": 3350
+    },
+    {
+      "epoch": 1.0054017555705603,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00015027865642416028,
+      "loss": 1.0881,
+      "step": 3351
+    },
+    {
+      "epoch": 1.0057018531022581,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001502635939147462,
+      "loss": 1.0651,
+      "step": 3352
+    },
+    {
+      "epoch": 1.0060019506339561,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00015024853140533213,
+      "loss": 1.0426,
+      "step": 3353
+    },
+    {
+      "epoch": 1.006302048165654,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00015023346889591807,
+      "loss": 1.145,
+      "step": 3354
+    },
+    {
+      "epoch": 1.0066021456973517,
+      "grad_norm": 1.125,
+      "learning_rate": 0.000150218406386504,
+      "loss": 1.0634,
+      "step": 3355
+    },
+    {
+      "epoch": 1.0069022432290495,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015020334387708993,
+      "loss": 0.9475,
+      "step": 3356
+    },
+    {
+      "epoch": 1.0072023407607473,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015018828136767587,
+      "loss": 0.9641,
+      "step": 3357
+    },
+    {
+      "epoch": 1.0075024382924451,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00015017321885826178,
+      "loss": 1.056,
+      "step": 3358
+    },
+    {
+      "epoch": 1.007802535824143,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00015015815634884772,
+      "loss": 1.0762,
+      "step": 3359
+    },
+    {
+      "epoch": 1.0081026333558407,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00015014309383943366,
+      "loss": 1.0712,
+      "step": 3360
+    },
+    {
+      "epoch": 1.0084027308875385,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00015012803133001958,
+      "loss": 1.1456,
+      "step": 3361
+    },
+    {
+      "epoch": 1.0087028284192363,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00015011296882060552,
+      "loss": 1.1155,
+      "step": 3362
+    },
+    {
+      "epoch": 1.009002925950934,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00015009790631119146,
+      "loss": 1.1318,
+      "step": 3363
+    },
+    {
+      "epoch": 1.009303023482632,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00015008284380177737,
+      "loss": 1.0007,
+      "step": 3364
+    },
+    {
+      "epoch": 1.0096031210143297,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001500677812923633,
+      "loss": 1.0908,
+      "step": 3365
+    },
+    {
+      "epoch": 1.0099032185460275,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00015005271878294925,
+      "loss": 1.0801,
+      "step": 3366
+    },
+    {
+      "epoch": 1.0102033160777253,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001500376562735352,
+      "loss": 1.1358,
+      "step": 3367
+    },
+    {
+      "epoch": 1.010503413609423,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001500225937641211,
+      "loss": 0.9545,
+      "step": 3368
+    },
+    {
+      "epoch": 1.0108035111411209,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015000753125470704,
+      "loss": 1.233,
+      "step": 3369
+    },
+    {
+      "epoch": 1.0111036086728187,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00014999246874529298,
+      "loss": 0.976,
+      "step": 3370
+    },
+    {
+      "epoch": 1.0114037062045165,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001499774062358789,
+      "loss": 0.955,
+      "step": 3371
+    },
+    {
+      "epoch": 1.0117038037362143,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014996234372646484,
+      "loss": 1.0807,
+      "step": 3372
+    },
+    {
+      "epoch": 1.012003901267912,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014994728121705078,
+      "loss": 1.0801,
+      "step": 3373
+    },
+    {
+      "epoch": 1.0123039987996099,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001499322187076367,
+      "loss": 1.2236,
+      "step": 3374
+    },
+    {
+      "epoch": 1.0126040963313077,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014991715619822263,
+      "loss": 1.028,
+      "step": 3375
+    },
+    {
+      "epoch": 1.0129041938630055,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014990209368880857,
+      "loss": 1.0801,
+      "step": 3376
+    },
+    {
+      "epoch": 1.0132042913947032,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014988703117939448,
+      "loss": 1.0792,
+      "step": 3377
+    },
+    {
+      "epoch": 1.013504388926401,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014987196866998043,
+      "loss": 0.9272,
+      "step": 3378
+    },
+    {
+      "epoch": 1.0138044864580988,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014985690616056637,
+      "loss": 1.0352,
+      "step": 3379
+    },
+    {
+      "epoch": 1.0141045839897966,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014984184365115228,
+      "loss": 1.2585,
+      "step": 3380
+    },
+    {
+      "epoch": 1.0144046815214944,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014982678114173822,
+      "loss": 1.129,
+      "step": 3381
+    },
+    {
+      "epoch": 1.0147047790531922,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00014981171863232416,
+      "loss": 1.0665,
+      "step": 3382
+    },
+    {
+      "epoch": 1.01500487658489,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001497966561229101,
+      "loss": 0.9684,
+      "step": 3383
+    },
+    {
+      "epoch": 1.0153049741165878,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.000149781593613496,
+      "loss": 0.847,
+      "step": 3384
+    },
+    {
+      "epoch": 1.0156050716482856,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014976653110408195,
+      "loss": 1.0456,
+      "step": 3385
+    },
+    {
+      "epoch": 1.0159051691799834,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001497514685946679,
+      "loss": 1.157,
+      "step": 3386
+    },
+    {
+      "epoch": 1.0162052667116812,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001497364060852538,
+      "loss": 1.0621,
+      "step": 3387
+    },
+    {
+      "epoch": 1.016505364243379,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00014972134357583975,
+      "loss": 1.0306,
+      "step": 3388
+    },
+    {
+      "epoch": 1.0168054617750768,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001497062810664257,
+      "loss": 1.1093,
+      "step": 3389
+    },
+    {
+      "epoch": 1.0171055593067746,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001496912185570116,
+      "loss": 1.1033,
+      "step": 3390
+    },
+    {
+      "epoch": 1.0174056568384726,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014967615604759751,
+      "loss": 1.1307,
+      "step": 3391
+    },
+    {
+      "epoch": 1.0177057543701704,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00014966109353818348,
+      "loss": 1.1558,
+      "step": 3392
+    },
+    {
+      "epoch": 1.0180058519018682,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001496460310287694,
+      "loss": 1.1393,
+      "step": 3393
+    },
+    {
+      "epoch": 1.018305949433566,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00014963096851935533,
+      "loss": 0.9266,
+      "step": 3394
+    },
+    {
+      "epoch": 1.0186060469652638,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014961590600994127,
+      "loss": 1.2388,
+      "step": 3395
+    },
+    {
+      "epoch": 1.0189061444969616,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001496008435005272,
+      "loss": 1.2517,
+      "step": 3396
+    },
+    {
+      "epoch": 1.0192062420286594,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014958578099111313,
+      "loss": 1.0354,
+      "step": 3397
+    },
+    {
+      "epoch": 1.0195063395603572,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014957071848169907,
+      "loss": 1.1143,
+      "step": 3398
+    },
+    {
+      "epoch": 1.019806437092055,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014955565597228498,
+      "loss": 1.0618,
+      "step": 3399
+    },
+    {
+      "epoch": 1.0201065346237528,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014954059346287092,
+      "loss": 0.9842,
+      "step": 3400
+    },
+    {
+      "epoch": 1.0204066321554506,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014952553095345686,
+      "loss": 1.0785,
+      "step": 3401
+    },
+    {
+      "epoch": 1.0207067296871484,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001495104684440428,
+      "loss": 0.8975,
+      "step": 3402
+    },
+    {
+      "epoch": 1.0210068272188462,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00014949540593462872,
+      "loss": 1.1161,
+      "step": 3403
+    },
+    {
+      "epoch": 1.021306924750544,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00014948034342521463,
+      "loss": 1.0765,
+      "step": 3404
+    },
+    {
+      "epoch": 1.0216070222822418,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001494652809158006,
+      "loss": 1.1039,
+      "step": 3405
+    },
+    {
+      "epoch": 1.0219071198139396,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001494502184063865,
+      "loss": 1.1842,
+      "step": 3406
+    },
+    {
+      "epoch": 1.0222072173456374,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014943515589697242,
+      "loss": 0.9892,
+      "step": 3407
+    },
+    {
+      "epoch": 1.0225073148773351,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001494200933875584,
+      "loss": 1.0832,
+      "step": 3408
+    },
+    {
+      "epoch": 1.022807412409033,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001494050308781443,
+      "loss": 0.8715,
+      "step": 3409
+    },
+    {
+      "epoch": 1.0231075099407307,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014938996836873024,
+      "loss": 0.908,
+      "step": 3410
+    },
+    {
+      "epoch": 1.0234076074724285,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014937490585931618,
+      "loss": 0.9814,
+      "step": 3411
+    },
+    {
+      "epoch": 1.0237077050041263,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001493598433499021,
+      "loss": 1.1114,
+      "step": 3412
+    },
+    {
+      "epoch": 1.0240078025358241,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014934478084048804,
+      "loss": 1.096,
+      "step": 3413
+    },
+    {
+      "epoch": 1.024307900067522,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014932971833107395,
+      "loss": 1.0915,
+      "step": 3414
+    },
+    {
+      "epoch": 1.0246079975992197,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001493146558216599,
+      "loss": 0.9262,
+      "step": 3415
+    },
+    {
+      "epoch": 1.0249080951309175,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014929959331224583,
+      "loss": 1.3882,
+      "step": 3416
+    },
+    {
+      "epoch": 1.0252081926626153,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014928453080283174,
+      "loss": 1.1681,
+      "step": 3417
+    },
+    {
+      "epoch": 1.025508290194313,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014926946829341769,
+      "loss": 1.0202,
+      "step": 3418
+    },
+    {
+      "epoch": 1.025808387726011,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00014925440578400363,
+      "loss": 1.1708,
+      "step": 3419
+    },
+    {
+      "epoch": 1.0261084852577087,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014923934327458954,
+      "loss": 1.1651,
+      "step": 3420
+    },
+    {
+      "epoch": 1.0264085827894065,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001492242807651755,
+      "loss": 1.0867,
+      "step": 3421
+    },
+    {
+      "epoch": 1.0267086803211043,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014920921825576142,
+      "loss": 1.0418,
+      "step": 3422
+    },
+    {
+      "epoch": 1.027008777852802,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00014919415574634733,
+      "loss": 0.9849,
+      "step": 3423
+    },
+    {
+      "epoch": 1.0273088753844999,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001491790932369333,
+      "loss": 1.2761,
+      "step": 3424
+    },
+    {
+      "epoch": 1.0276089729161977,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001491640307275192,
+      "loss": 1.2147,
+      "step": 3425
+    },
+    {
+      "epoch": 1.0279090704478955,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00014914896821810513,
+      "loss": 1.1427,
+      "step": 3426
+    },
+    {
+      "epoch": 1.0282091679795933,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014913390570869107,
+      "loss": 0.9519,
+      "step": 3427
+    },
+    {
+      "epoch": 1.028509265511291,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000149118843199277,
+      "loss": 0.9588,
+      "step": 3428
+    },
+    {
+      "epoch": 1.0288093630429889,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014910378068986295,
+      "loss": 1.1389,
+      "step": 3429
+    },
+    {
+      "epoch": 1.0291094605746869,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014908871818044886,
+      "loss": 1.0575,
+      "step": 3430
+    },
+    {
+      "epoch": 1.0294095581063847,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001490736556710348,
+      "loss": 0.9395,
+      "step": 3431
+    },
+    {
+      "epoch": 1.0297096556380825,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00014905859316162074,
+      "loss": 1.1282,
+      "step": 3432
+    },
+    {
+      "epoch": 1.0300097531697803,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00014904353065220665,
+      "loss": 0.9917,
+      "step": 3433
+    },
+    {
+      "epoch": 1.030309850701478,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001490284681427926,
+      "loss": 1.0631,
+      "step": 3434
+    },
+    {
+      "epoch": 1.0306099482331759,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00014901340563337853,
+      "loss": 1.0032,
+      "step": 3435
+    },
+    {
+      "epoch": 1.0309100457648737,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014899834312396445,
+      "loss": 1.0844,
+      "step": 3436
+    },
+    {
+      "epoch": 1.0312101432965715,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014898328061455042,
+      "loss": 1.1862,
+      "step": 3437
+    },
+    {
+      "epoch": 1.0315102408282693,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014896821810513633,
+      "loss": 1.0904,
+      "step": 3438
+    },
+    {
+      "epoch": 1.031810338359967,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014895315559572224,
+      "loss": 1.1538,
+      "step": 3439
+    },
+    {
+      "epoch": 1.0321104358916648,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014893809308630818,
+      "loss": 1.0375,
+      "step": 3440
+    },
+    {
+      "epoch": 1.0324105334233626,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014892303057689412,
+      "loss": 1.0437,
+      "step": 3441
+    },
+    {
+      "epoch": 1.0327106309550604,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014890796806748004,
+      "loss": 1.0654,
+      "step": 3442
+    },
+    {
+      "epoch": 1.0330107284867582,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014889290555806598,
+      "loss": 1.0325,
+      "step": 3443
+    },
+    {
+      "epoch": 1.033310826018456,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014887784304865192,
+      "loss": 1.1188,
+      "step": 3444
+    },
+    {
+      "epoch": 1.0336109235501538,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014886278053923783,
+      "loss": 1.1427,
+      "step": 3445
+    },
+    {
+      "epoch": 1.0339110210818516,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014884771802982377,
+      "loss": 1.0513,
+      "step": 3446
+    },
+    {
+      "epoch": 1.0342111186135494,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001488326555204097,
+      "loss": 0.9407,
+      "step": 3447
+    },
+    {
+      "epoch": 1.0345112161452472,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00014881759301099565,
+      "loss": 1.1964,
+      "step": 3448
+    },
+    {
+      "epoch": 1.034811313676945,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00014880253050158156,
+      "loss": 0.9971,
+      "step": 3449
+    },
+    {
+      "epoch": 1.0351114112086428,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001487874679921675,
+      "loss": 1.0553,
+      "step": 3450
+    },
+    {
+      "epoch": 1.0354115087403406,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014877240548275344,
+      "loss": 1.1559,
+      "step": 3451
+    },
+    {
+      "epoch": 1.0357116062720384,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014875734297333936,
+      "loss": 1.1931,
+      "step": 3452
+    },
+    {
+      "epoch": 1.0360117038037362,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001487422804639253,
+      "loss": 1.234,
+      "step": 3453
+    },
+    {
+      "epoch": 1.036311801335434,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014872721795451124,
+      "loss": 1.2856,
+      "step": 3454
+    },
+    {
+      "epoch": 1.0366118988671318,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014871215544509715,
+      "loss": 1.0562,
+      "step": 3455
+    },
+    {
+      "epoch": 1.0369119963988296,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001486970929356831,
+      "loss": 1.1136,
+      "step": 3456
+    },
+    {
+      "epoch": 1.0372120939305274,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014868203042626903,
+      "loss": 1.1763,
+      "step": 3457
+    },
+    {
+      "epoch": 1.0375121914622252,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00014866696791685494,
+      "loss": 0.9996,
+      "step": 3458
+    },
+    {
+      "epoch": 1.037812288993923,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014865190540744089,
+      "loss": 0.9622,
+      "step": 3459
+    },
+    {
+      "epoch": 1.0381123865256208,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00014863684289802683,
+      "loss": 1.016,
+      "step": 3460
+    },
+    {
+      "epoch": 1.0384124840573186,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014862178038861274,
+      "loss": 1.0539,
+      "step": 3461
+    },
+    {
+      "epoch": 1.0387125815890164,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014860671787919868,
+      "loss": 0.8954,
+      "step": 3462
+    },
+    {
+      "epoch": 1.0390126791207142,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014859165536978462,
+      "loss": 1.0173,
+      "step": 3463
+    },
+    {
+      "epoch": 1.039312776652412,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014857659286037056,
+      "loss": 1.0283,
+      "step": 3464
+    },
+    {
+      "epoch": 1.0396128741841097,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014856153035095647,
+      "loss": 0.9829,
+      "step": 3465
+    },
+    {
+      "epoch": 1.0399129717158075,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001485464678415424,
+      "loss": 0.9507,
+      "step": 3466
+    },
+    {
+      "epoch": 1.0402130692475053,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014853140533212835,
+      "loss": 1.2274,
+      "step": 3467
+    },
+    {
+      "epoch": 1.0405131667792031,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014851634282271427,
+      "loss": 1.1462,
+      "step": 3468
+    },
+    {
+      "epoch": 1.0408132643109012,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001485012803133002,
+      "loss": 0.9688,
+      "step": 3469
+    },
+    {
+      "epoch": 1.041113361842599,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014848621780388615,
+      "loss": 1.0579,
+      "step": 3470
+    },
+    {
+      "epoch": 1.0414134593742967,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014847115529447206,
+      "loss": 1.0185,
+      "step": 3471
+    },
+    {
+      "epoch": 1.0417135569059945,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.000148456092785058,
+      "loss": 1.1523,
+      "step": 3472
+    },
+    {
+      "epoch": 1.0420136544376923,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00014844103027564394,
+      "loss": 1.0728,
+      "step": 3473
+    },
+    {
+      "epoch": 1.0423137519693901,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014842596776622985,
+      "loss": 1.1162,
+      "step": 3474
+    },
+    {
+      "epoch": 1.042613849501088,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001484109052568158,
+      "loss": 1.0803,
+      "step": 3475
+    },
+    {
+      "epoch": 1.0429139470327857,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014839584274740173,
+      "loss": 0.9683,
+      "step": 3476
+    },
+    {
+      "epoch": 1.0432140445644835,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014838078023798765,
+      "loss": 1.0315,
+      "step": 3477
+    },
+    {
+      "epoch": 1.0435141420961813,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001483657177285736,
+      "loss": 1.2398,
+      "step": 3478
+    },
+    {
+      "epoch": 1.0438142396278791,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014835065521915953,
+      "loss": 1.1138,
+      "step": 3479
+    },
+    {
+      "epoch": 1.044114337159577,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014833559270974544,
+      "loss": 1.0138,
+      "step": 3480
+    },
+    {
+      "epoch": 1.0444144346912747,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00014832053020033138,
+      "loss": 1.1907,
+      "step": 3481
+    },
+    {
+      "epoch": 1.0447145322229725,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00014830546769091732,
+      "loss": 0.8607,
+      "step": 3482
+    },
+    {
+      "epoch": 1.0450146297546703,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014829040518150326,
+      "loss": 0.9037,
+      "step": 3483
+    },
+    {
+      "epoch": 1.045314727286368,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014827534267208918,
+      "loss": 0.8543,
+      "step": 3484
+    },
+    {
+      "epoch": 1.045614824818066,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001482602801626751,
+      "loss": 0.9897,
+      "step": 3485
+    },
+    {
+      "epoch": 1.0459149223497637,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014824521765326106,
+      "loss": 1.0463,
+      "step": 3486
+    },
+    {
+      "epoch": 1.0462150198814615,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014823015514384697,
+      "loss": 1.0342,
+      "step": 3487
+    },
+    {
+      "epoch": 1.0465151174131593,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014821509263443288,
+      "loss": 1.0641,
+      "step": 3488
+    },
+    {
+      "epoch": 1.046815214944857,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014820003012501885,
+      "loss": 1.0872,
+      "step": 3489
+    },
+    {
+      "epoch": 1.0471153124765549,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014818496761560476,
+      "loss": 0.9641,
+      "step": 3490
+    },
+    {
+      "epoch": 1.0474154100082527,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001481699051061907,
+      "loss": 1.0293,
+      "step": 3491
+    },
+    {
+      "epoch": 1.0477155075399505,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00014815484259677664,
+      "loss": 1.2783,
+      "step": 3492
+    },
+    {
+      "epoch": 1.0480156050716483,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014813978008736256,
+      "loss": 0.9915,
+      "step": 3493
+    },
+    {
+      "epoch": 1.048315702603346,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001481247175779485,
+      "loss": 1.1167,
+      "step": 3494
+    },
+    {
+      "epoch": 1.0486158001350439,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014810965506853444,
+      "loss": 1.013,
+      "step": 3495
+    },
+    {
+      "epoch": 1.0489158976667416,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014809459255912035,
+      "loss": 0.9687,
+      "step": 3496
+    },
+    {
+      "epoch": 1.0492159951984394,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001480795300497063,
+      "loss": 1.0029,
+      "step": 3497
+    },
+    {
+      "epoch": 1.0495160927301372,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001480644675402922,
+      "loss": 1.0085,
+      "step": 3498
+    },
+    {
+      "epoch": 1.049816190261835,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014804940503087815,
+      "loss": 1.013,
+      "step": 3499
+    },
+    {
+      "epoch": 1.0501162877935328,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014803434252146409,
+      "loss": 1.2271,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0504163853252306,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014801928001205,
+      "loss": 1.0194,
+      "step": 3501
+    },
+    {
+      "epoch": 1.0507164828569284,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014800421750263597,
+      "loss": 1.0262,
+      "step": 3502
+    },
+    {
+      "epoch": 1.0510165803886262,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014798915499322188,
+      "loss": 0.9579,
+      "step": 3503
+    },
+    {
+      "epoch": 1.051316677920324,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001479740924838078,
+      "loss": 1.0026,
+      "step": 3504
+    },
+    {
+      "epoch": 1.0516167754520218,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014795902997439376,
+      "loss": 1.0568,
+      "step": 3505
+    },
+    {
+      "epoch": 1.0519168729837196,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014794396746497967,
+      "loss": 1.0042,
+      "step": 3506
+    },
+    {
+      "epoch": 1.0522169705154174,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014792890495556559,
+      "loss": 1.0494,
+      "step": 3507
+    },
+    {
+      "epoch": 1.0525170680471154,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014791384244615155,
+      "loss": 0.962,
+      "step": 3508
+    },
+    {
+      "epoch": 1.0528171655788132,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014789877993673747,
+      "loss": 0.9943,
+      "step": 3509
+    },
+    {
+      "epoch": 1.053117263110511,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001478837174273234,
+      "loss": 1.19,
+      "step": 3510
+    },
+    {
+      "epoch": 1.0534173606422088,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014786865491790932,
+      "loss": 1.111,
+      "step": 3511
+    },
+    {
+      "epoch": 1.0537174581739066,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014785359240849526,
+      "loss": 0.9804,
+      "step": 3512
+    },
+    {
+      "epoch": 1.0540175557056044,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001478385298990812,
+      "loss": 0.9441,
+      "step": 3513
+    },
+    {
+      "epoch": 1.0543176532373022,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00014782346738966711,
+      "loss": 1.0781,
+      "step": 3514
+    },
+    {
+      "epoch": 1.054617750769,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014780840488025305,
+      "loss": 0.9821,
+      "step": 3515
+    },
+    {
+      "epoch": 1.0549178483006978,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.000147793342370839,
+      "loss": 1.0597,
+      "step": 3516
+    },
+    {
+      "epoch": 1.0552179458323956,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001477782798614249,
+      "loss": 1.0673,
+      "step": 3517
+    },
+    {
+      "epoch": 1.0555180433640934,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014776321735201088,
+      "loss": 1.031,
+      "step": 3518
+    },
+    {
+      "epoch": 1.0558181408957912,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001477481548425968,
+      "loss": 1.1664,
+      "step": 3519
+    },
+    {
+      "epoch": 1.056118238427489,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001477330923331827,
+      "loss": 1.1571,
+      "step": 3520
+    },
+    {
+      "epoch": 1.0564183359591868,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014771802982376864,
+      "loss": 1.0245,
+      "step": 3521
+    },
+    {
+      "epoch": 1.0567184334908846,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014770296731435458,
+      "loss": 1.0832,
+      "step": 3522
+    },
+    {
+      "epoch": 1.0570185310225824,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001476879048049405,
+      "loss": 1.025,
+      "step": 3523
+    },
+    {
+      "epoch": 1.0573186285542802,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014767284229552644,
+      "loss": 1.0928,
+      "step": 3524
+    },
+    {
+      "epoch": 1.057618726085978,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014765777978611238,
+      "loss": 1.1428,
+      "step": 3525
+    },
+    {
+      "epoch": 1.0579188236176758,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001476427172766983,
+      "loss": 1.117,
+      "step": 3526
+    },
+    {
+      "epoch": 1.0582189211493735,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00014762765476728423,
+      "loss": 1.2829,
+      "step": 3527
+    },
+    {
+      "epoch": 1.0585190186810713,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014761259225787017,
+      "loss": 1.2026,
+      "step": 3528
+    },
+    {
+      "epoch": 1.0588191162127691,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001475975297484561,
+      "loss": 1.0769,
+      "step": 3529
+    },
+    {
+      "epoch": 1.059119213744467,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014758246723904202,
+      "loss": 1.1914,
+      "step": 3530
+    },
+    {
+      "epoch": 1.0594193112761647,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014756740472962796,
+      "loss": 1.0964,
+      "step": 3531
+    },
+    {
+      "epoch": 1.0597194088078625,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001475523422202139,
+      "loss": 1.0458,
+      "step": 3532
+    },
+    {
+      "epoch": 1.0600195063395603,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014753727971079982,
+      "loss": 0.9964,
+      "step": 3533
+    },
+    {
+      "epoch": 1.0603196038712581,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014752221720138576,
+      "loss": 1.0697,
+      "step": 3534
+    },
+    {
+      "epoch": 1.060619701402956,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001475071546919717,
+      "loss": 1.0667,
+      "step": 3535
+    },
+    {
+      "epoch": 1.0609197989346537,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001474920921825576,
+      "loss": 0.9745,
+      "step": 3536
+    },
+    {
+      "epoch": 1.0612198964663515,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014747702967314355,
+      "loss": 1.2627,
+      "step": 3537
+    },
+    {
+      "epoch": 1.0615199939980493,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001474619671637295,
+      "loss": 1.038,
+      "step": 3538
+    },
+    {
+      "epoch": 1.061820091529747,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001474469046543154,
+      "loss": 1.0764,
+      "step": 3539
+    },
+    {
+      "epoch": 1.062120189061445,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014743184214490135,
+      "loss": 1.0228,
+      "step": 3540
+    },
+    {
+      "epoch": 1.0624202865931427,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014741677963548729,
+      "loss": 1.1635,
+      "step": 3541
+    },
+    {
+      "epoch": 1.0627203841248405,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001474017171260732,
+      "loss": 1.0164,
+      "step": 3542
+    },
+    {
+      "epoch": 1.0630204816565383,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00014738665461665914,
+      "loss": 1.0282,
+      "step": 3543
+    },
+    {
+      "epoch": 1.063320579188236,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014737159210724508,
+      "loss": 1.0236,
+      "step": 3544
+    },
+    {
+      "epoch": 1.0636206767199339,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014735652959783102,
+      "loss": 1.2436,
+      "step": 3545
+    },
+    {
+      "epoch": 1.0639207742516317,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014734146708841693,
+      "loss": 1.0388,
+      "step": 3546
+    },
+    {
+      "epoch": 1.0642208717833297,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014732640457900287,
+      "loss": 1.0027,
+      "step": 3547
+    },
+    {
+      "epoch": 1.0645209693150275,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001473113420695888,
+      "loss": 0.8788,
+      "step": 3548
+    },
+    {
+      "epoch": 1.0648210668467253,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014729627956017473,
+      "loss": 0.9788,
+      "step": 3549
+    },
+    {
+      "epoch": 1.065121164378423,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014728121705076067,
+      "loss": 1.082,
+      "step": 3550
+    },
+    {
+      "epoch": 1.0654212619101209,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001472661545413466,
+      "loss": 1.0243,
+      "step": 3551
+    },
+    {
+      "epoch": 1.0657213594418187,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014725109203193252,
+      "loss": 1.048,
+      "step": 3552
+    },
+    {
+      "epoch": 1.0660214569735165,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00014723602952251846,
+      "loss": 0.8532,
+      "step": 3553
+    },
+    {
+      "epoch": 1.0663215545052143,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001472209670131044,
+      "loss": 1.2176,
+      "step": 3554
+    },
+    {
+      "epoch": 1.066621652036912,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014720590450369031,
+      "loss": 1.1672,
+      "step": 3555
+    },
+    {
+      "epoch": 1.0669217495686099,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014719084199427625,
+      "loss": 1.1527,
+      "step": 3556
+    },
+    {
+      "epoch": 1.0672218471003077,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001471757794848622,
+      "loss": 1.0643,
+      "step": 3557
+    },
+    {
+      "epoch": 1.0675219446320054,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001471607169754481,
+      "loss": 0.9624,
+      "step": 3558
+    },
+    {
+      "epoch": 1.0678220421637032,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014714565446603405,
+      "loss": 1.0716,
+      "step": 3559
+    },
+    {
+      "epoch": 1.068122139695401,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00014713059195662,
+      "loss": 1.1204,
+      "step": 3560
+    },
+    {
+      "epoch": 1.0684222372270988,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001471155294472059,
+      "loss": 1.0464,
+      "step": 3561
+    },
+    {
+      "epoch": 1.0687223347587966,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00014710046693779184,
+      "loss": 0.8531,
+      "step": 3562
+    },
+    {
+      "epoch": 1.0690224322904944,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014708540442837778,
+      "loss": 0.988,
+      "step": 3563
+    },
+    {
+      "epoch": 1.0693225298221922,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014707034191896372,
+      "loss": 1.1896,
+      "step": 3564
+    },
+    {
+      "epoch": 1.06962262735389,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014705527940954964,
+      "loss": 0.9547,
+      "step": 3565
+    },
+    {
+      "epoch": 1.0699227248855878,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014704021690013558,
+      "loss": 0.9893,
+      "step": 3566
+    },
+    {
+      "epoch": 1.0702228224172856,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00014702515439072152,
+      "loss": 1.1418,
+      "step": 3567
+    },
+    {
+      "epoch": 1.0705229199489834,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00014701009188130743,
+      "loss": 1.0344,
+      "step": 3568
+    },
+    {
+      "epoch": 1.0708230174806812,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014699502937189334,
+      "loss": 1.0847,
+      "step": 3569
+    },
+    {
+      "epoch": 1.071123115012379,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001469799668624793,
+      "loss": 1.082,
+      "step": 3570
+    },
+    {
+      "epoch": 1.0714232125440768,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014696490435306522,
+      "loss": 0.914,
+      "step": 3571
+    },
+    {
+      "epoch": 1.0717233100757746,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00014694984184365116,
+      "loss": 1.0121,
+      "step": 3572
+    },
+    {
+      "epoch": 1.0720234076074724,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001469347793342371,
+      "loss": 0.9822,
+      "step": 3573
+    },
+    {
+      "epoch": 1.0723235051391702,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014691971682482302,
+      "loss": 1.0387,
+      "step": 3574
+    },
+    {
+      "epoch": 1.072623602670868,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014690465431540896,
+      "loss": 1.1467,
+      "step": 3575
+    },
+    {
+      "epoch": 1.0729237002025658,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001468895918059949,
+      "loss": 1.0692,
+      "step": 3576
+    },
+    {
+      "epoch": 1.0732237977342636,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001468745292965808,
+      "loss": 1.0281,
+      "step": 3577
+    },
+    {
+      "epoch": 1.0735238952659614,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014685946678716675,
+      "loss": 1.078,
+      "step": 3578
+    },
+    {
+      "epoch": 1.0738239927976592,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001468444042777527,
+      "loss": 1.1216,
+      "step": 3579
+    },
+    {
+      "epoch": 1.074124090329357,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001468293417683386,
+      "loss": 1.0332,
+      "step": 3580
+    },
+    {
+      "epoch": 1.0744241878610548,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00014681427925892455,
+      "loss": 1.0282,
+      "step": 3581
+    },
+    {
+      "epoch": 1.0747242853927526,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014679921674951046,
+      "loss": 1.0961,
+      "step": 3582
+    },
+    {
+      "epoch": 1.0750243829244504,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014678415424009643,
+      "loss": 1.1022,
+      "step": 3583
+    },
+    {
+      "epoch": 1.0753244804561481,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014676909173068234,
+      "loss": 1.0418,
+      "step": 3584
+    },
+    {
+      "epoch": 1.075624577987846,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014675402922126825,
+      "loss": 1.1828,
+      "step": 3585
+    },
+    {
+      "epoch": 1.075924675519544,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014673896671185422,
+      "loss": 1.1665,
+      "step": 3586
+    },
+    {
+      "epoch": 1.0762247730512418,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014672390420244013,
+      "loss": 1.1,
+      "step": 3587
+    },
+    {
+      "epoch": 1.0765248705829396,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014670884169302605,
+      "loss": 1.1869,
+      "step": 3588
+    },
+    {
+      "epoch": 1.0768249681146373,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014669377918361201,
+      "loss": 1.0397,
+      "step": 3589
+    },
+    {
+      "epoch": 1.0771250656463351,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014667871667419793,
+      "loss": 1.0235,
+      "step": 3590
+    },
+    {
+      "epoch": 1.077425163178033,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014666365416478387,
+      "loss": 1.1844,
+      "step": 3591
+    },
+    {
+      "epoch": 1.0777252607097307,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014664859165536978,
+      "loss": 1.0008,
+      "step": 3592
+    },
+    {
+      "epoch": 1.0780253582414285,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00014663352914595572,
+      "loss": 1.1219,
+      "step": 3593
+    },
+    {
+      "epoch": 1.0783254557731263,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014661846663654166,
+      "loss": 1.1054,
+      "step": 3594
+    },
+    {
+      "epoch": 1.0786255533048241,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014660340412712757,
+      "loss": 1.0188,
+      "step": 3595
+    },
+    {
+      "epoch": 1.078925650836522,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014658834161771351,
+      "loss": 1.2891,
+      "step": 3596
+    },
+    {
+      "epoch": 1.0792257483682197,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00014657327910829945,
+      "loss": 1.2174,
+      "step": 3597
+    },
+    {
+      "epoch": 1.0795258458999175,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014655821659888537,
+      "loss": 0.9183,
+      "step": 3598
+    },
+    {
+      "epoch": 1.0798259434316153,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00014654315408947134,
+      "loss": 1.3557,
+      "step": 3599
+    },
+    {
+      "epoch": 1.080126040963313,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014652809158005725,
+      "loss": 1.0142,
+      "step": 3600
+    },
+    {
+      "epoch": 1.080426138495011,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014651302907064316,
+      "loss": 1.1377,
+      "step": 3601
+    },
+    {
+      "epoch": 1.0807262360267087,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014649796656122913,
+      "loss": 0.9733,
+      "step": 3602
+    },
+    {
+      "epoch": 1.0810263335584065,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014648290405181504,
+      "loss": 1.0548,
+      "step": 3603
+    },
+    {
+      "epoch": 1.0813264310901043,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014646784154240096,
+      "loss": 1.0544,
+      "step": 3604
+    },
+    {
+      "epoch": 1.081626528621802,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001464527790329869,
+      "loss": 0.9851,
+      "step": 3605
+    },
+    {
+      "epoch": 1.0819266261534999,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014643771652357284,
+      "loss": 0.9255,
+      "step": 3606
+    },
+    {
+      "epoch": 1.0822267236851977,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014642265401415875,
+      "loss": 1.0809,
+      "step": 3607
+    },
+    {
+      "epoch": 1.0825268212168955,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001464075915047447,
+      "loss": 1.2343,
+      "step": 3608
+    },
+    {
+      "epoch": 1.0828269187485933,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014639252899533063,
+      "loss": 1.1921,
+      "step": 3609
+    },
+    {
+      "epoch": 1.083127016280291,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014637746648591657,
+      "loss": 1.0331,
+      "step": 3610
+    },
+    {
+      "epoch": 1.0834271138119889,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00014636240397650248,
+      "loss": 1.0536,
+      "step": 3611
+    },
+    {
+      "epoch": 1.0837272113436867,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014634734146708842,
+      "loss": 1.0511,
+      "step": 3612
+    },
+    {
+      "epoch": 1.0840273088753845,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014633227895767436,
+      "loss": 0.9657,
+      "step": 3613
+    },
+    {
+      "epoch": 1.0843274064070823,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014631721644826028,
+      "loss": 1.0773,
+      "step": 3614
+    },
+    {
+      "epoch": 1.08462750393878,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014630215393884622,
+      "loss": 0.9361,
+      "step": 3615
+    },
+    {
+      "epoch": 1.0849276014704778,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014628709142943216,
+      "loss": 1.0562,
+      "step": 3616
+    },
+    {
+      "epoch": 1.0852276990021756,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014627202892001807,
+      "loss": 0.9882,
+      "step": 3617
+    },
+    {
+      "epoch": 1.0855277965338734,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.000146256966410604,
+      "loss": 1.03,
+      "step": 3618
+    },
+    {
+      "epoch": 1.0858278940655712,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014624190390118995,
+      "loss": 0.9585,
+      "step": 3619
+    },
+    {
+      "epoch": 1.086127991597269,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014622684139177586,
+      "loss": 1.0982,
+      "step": 3620
+    },
+    {
+      "epoch": 1.0864280891289668,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001462117788823618,
+      "loss": 0.921,
+      "step": 3621
+    },
+    {
+      "epoch": 1.0867281866606646,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014619671637294775,
+      "loss": 0.9135,
+      "step": 3622
+    },
+    {
+      "epoch": 1.0870282841923624,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014618165386353366,
+      "loss": 1.0046,
+      "step": 3623
+    },
+    {
+      "epoch": 1.0873283817240602,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001461665913541196,
+      "loss": 1.1094,
+      "step": 3624
+    },
+    {
+      "epoch": 1.0876284792557582,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014615152884470554,
+      "loss": 1.0397,
+      "step": 3625
+    },
+    {
+      "epoch": 1.087928576787456,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00014613646633529148,
+      "loss": 1.1634,
+      "step": 3626
+    },
+    {
+      "epoch": 1.0882286743191538,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001461214038258774,
+      "loss": 0.9921,
+      "step": 3627
+    },
+    {
+      "epoch": 1.0885287718508516,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014610634131646333,
+      "loss": 1.1607,
+      "step": 3628
+    },
+    {
+      "epoch": 1.0888288693825494,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014609127880704927,
+      "loss": 1.0174,
+      "step": 3629
+    },
+    {
+      "epoch": 1.0891289669142472,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001460762162976352,
+      "loss": 0.9309,
+      "step": 3630
+    },
+    {
+      "epoch": 1.089429064445945,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014606115378822113,
+      "loss": 1.0422,
+      "step": 3631
+    },
+    {
+      "epoch": 1.0897291619776428,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014604609127880707,
+      "loss": 1.0856,
+      "step": 3632
+    },
+    {
+      "epoch": 1.0900292595093406,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014603102876939298,
+      "loss": 1.1403,
+      "step": 3633
+    },
+    {
+      "epoch": 1.0903293570410384,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014601596625997892,
+      "loss": 1.2238,
+      "step": 3634
+    },
+    {
+      "epoch": 1.0906294545727362,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014600090375056486,
+      "loss": 1.0562,
+      "step": 3635
+    },
+    {
+      "epoch": 1.090929552104434,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00014598584124115077,
+      "loss": 0.9922,
+      "step": 3636
+    },
+    {
+      "epoch": 1.0912296496361318,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014597077873173671,
+      "loss": 1.1439,
+      "step": 3637
+    },
+    {
+      "epoch": 1.0915297471678296,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014595571622232265,
+      "loss": 1.2576,
+      "step": 3638
+    },
+    {
+      "epoch": 1.0918298446995274,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014594065371290857,
+      "loss": 1.1133,
+      "step": 3639
+    },
+    {
+      "epoch": 1.0921299422312252,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001459255912034945,
+      "loss": 1.2375,
+      "step": 3640
+    },
+    {
+      "epoch": 1.092430039762923,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014591052869408045,
+      "loss": 1.1903,
+      "step": 3641
+    },
+    {
+      "epoch": 1.0927301372946208,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014589546618466636,
+      "loss": 1.0328,
+      "step": 3642
+    },
+    {
+      "epoch": 1.0930302348263186,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001458804036752523,
+      "loss": 1.092,
+      "step": 3643
+    },
+    {
+      "epoch": 1.0933303323580164,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014586534116583824,
+      "loss": 1.1112,
+      "step": 3644
+    },
+    {
+      "epoch": 1.0936304298897142,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014585027865642418,
+      "loss": 1.0486,
+      "step": 3645
+    },
+    {
+      "epoch": 1.093930527421412,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001458352161470101,
+      "loss": 1.2153,
+      "step": 3646
+    },
+    {
+      "epoch": 1.0942306249531097,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014582015363759604,
+      "loss": 1.1161,
+      "step": 3647
+    },
+    {
+      "epoch": 1.0945307224848075,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014580509112818198,
+      "loss": 1.0963,
+      "step": 3648
+    },
+    {
+      "epoch": 1.0948308200165053,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001457900286187679,
+      "loss": 1.0737,
+      "step": 3649
+    },
+    {
+      "epoch": 1.0951309175482031,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001457749661093538,
+      "loss": 1.1673,
+      "step": 3650
+    },
+    {
+      "epoch": 1.095431015079901,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014575990359993977,
+      "loss": 1.0833,
+      "step": 3651
+    },
+    {
+      "epoch": 1.0957311126115987,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014574484109052568,
+      "loss": 1.0062,
+      "step": 3652
+    },
+    {
+      "epoch": 1.0960312101432965,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014572977858111162,
+      "loss": 1.1946,
+      "step": 3653
+    },
+    {
+      "epoch": 1.0963313076749943,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014571471607169756,
+      "loss": 1.1911,
+      "step": 3654
+    },
+    {
+      "epoch": 1.0966314052066921,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014569965356228348,
+      "loss": 1.0343,
+      "step": 3655
+    },
+    {
+      "epoch": 1.09693150273839,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00014568459105286942,
+      "loss": 1.089,
+      "step": 3656
+    },
+    {
+      "epoch": 1.0972316002700877,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014566952854345536,
+      "loss": 1.1031,
+      "step": 3657
+    },
+    {
+      "epoch": 1.0975316978017855,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00014565446603404127,
+      "loss": 1.2519,
+      "step": 3658
+    },
+    {
+      "epoch": 1.0978317953334833,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001456394035246272,
+      "loss": 1.1629,
+      "step": 3659
+    },
+    {
+      "epoch": 1.098131892865181,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014562434101521315,
+      "loss": 1.042,
+      "step": 3660
+    },
+    {
+      "epoch": 1.0984319903968789,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014560927850579907,
+      "loss": 1.1369,
+      "step": 3661
+    },
+    {
+      "epoch": 1.0987320879285767,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.000145594215996385,
+      "loss": 1.1404,
+      "step": 3662
+    },
+    {
+      "epoch": 1.0990321854602745,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014557915348697092,
+      "loss": 0.9866,
+      "step": 3663
+    },
+    {
+      "epoch": 1.0993322829919725,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014556409097755689,
+      "loss": 0.9661,
+      "step": 3664
+    },
+    {
+      "epoch": 1.0996323805236703,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001455490284681428,
+      "loss": 1.0398,
+      "step": 3665
+    },
+    {
+      "epoch": 1.099932478055368,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001455339659587287,
+      "loss": 1.0937,
+      "step": 3666
+    },
+    {
+      "epoch": 1.1002325755870659,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014551890344931468,
+      "loss": 1.0013,
+      "step": 3667
+    },
+    {
+      "epoch": 1.1005326731187637,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001455038409399006,
+      "loss": 1.236,
+      "step": 3668
+    },
+    {
+      "epoch": 1.1008327706504615,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001454887784304865,
+      "loss": 1.0935,
+      "step": 3669
+    },
+    {
+      "epoch": 1.1011328681821593,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014547371592107247,
+      "loss": 1.1182,
+      "step": 3670
+    },
+    {
+      "epoch": 1.101432965713857,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001454586534116584,
+      "loss": 1.0679,
+      "step": 3671
+    },
+    {
+      "epoch": 1.1017330632455549,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00014544359090224433,
+      "loss": 1.0782,
+      "step": 3672
+    },
+    {
+      "epoch": 1.1020331607772527,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014542852839283027,
+      "loss": 0.9702,
+      "step": 3673
+    },
+    {
+      "epoch": 1.1023332583089505,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014541346588341618,
+      "loss": 1.0159,
+      "step": 3674
+    },
+    {
+      "epoch": 1.1026333558406483,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00014539840337400212,
+      "loss": 1.1441,
+      "step": 3675
+    },
+    {
+      "epoch": 1.102933453372346,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014538334086458803,
+      "loss": 1.1535,
+      "step": 3676
+    },
+    {
+      "epoch": 1.1032335509040438,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014536827835517397,
+      "loss": 1.1163,
+      "step": 3677
+    },
+    {
+      "epoch": 1.1035336484357416,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014535321584575991,
+      "loss": 0.946,
+      "step": 3678
+    },
+    {
+      "epoch": 1.1038337459674394,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014533815333634583,
+      "loss": 1.2297,
+      "step": 3679
+    },
+    {
+      "epoch": 1.1041338434991372,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001453230908269318,
+      "loss": 0.9837,
+      "step": 3680
+    },
+    {
+      "epoch": 1.104433941030835,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001453080283175177,
+      "loss": 1.261,
+      "step": 3681
+    },
+    {
+      "epoch": 1.1047340385625328,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014529296580810362,
+      "loss": 1.0111,
+      "step": 3682
+    },
+    {
+      "epoch": 1.1050341360942306,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001452779032986896,
+      "loss": 0.9702,
+      "step": 3683
+    },
+    {
+      "epoch": 1.1053342336259284,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001452628407892755,
+      "loss": 1.0692,
+      "step": 3684
+    },
+    {
+      "epoch": 1.1056343311576262,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014524777827986142,
+      "loss": 0.9896,
+      "step": 3685
+    },
+    {
+      "epoch": 1.105934428689324,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014523271577044738,
+      "loss": 1.1111,
+      "step": 3686
+    },
+    {
+      "epoch": 1.1062345262210218,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001452176532610333,
+      "loss": 0.89,
+      "step": 3687
+    },
+    {
+      "epoch": 1.1065346237527196,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001452025907516192,
+      "loss": 1.1751,
+      "step": 3688
+    },
+    {
+      "epoch": 1.1068347212844174,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00014518752824220515,
+      "loss": 1.2406,
+      "step": 3689
+    },
+    {
+      "epoch": 1.1071348188161152,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001451724657327911,
+      "loss": 0.9914,
+      "step": 3690
+    },
+    {
+      "epoch": 1.107434916347813,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014515740322337703,
+      "loss": 0.9851,
+      "step": 3691
+    },
+    {
+      "epoch": 1.1077350138795108,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014514234071396294,
+      "loss": 1.1031,
+      "step": 3692
+    },
+    {
+      "epoch": 1.1080351114112086,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014512727820454888,
+      "loss": 1.1329,
+      "step": 3693
+    },
+    {
+      "epoch": 1.1083352089429064,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014511221569513482,
+      "loss": 0.9897,
+      "step": 3694
+    },
+    {
+      "epoch": 1.1086353064746042,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014509715318572074,
+      "loss": 1.1678,
+      "step": 3695
+    },
+    {
+      "epoch": 1.108935404006302,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00014508209067630668,
+      "loss": 0.9753,
+      "step": 3696
+    },
+    {
+      "epoch": 1.1092355015379998,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014506702816689262,
+      "loss": 1.0663,
+      "step": 3697
+    },
+    {
+      "epoch": 1.1095355990696976,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014505196565747853,
+      "loss": 1.0335,
+      "step": 3698
+    },
+    {
+      "epoch": 1.1098356966013954,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014503690314806447,
+      "loss": 1.006,
+      "step": 3699
+    },
+    {
+      "epoch": 1.1101357941330932,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001450218406386504,
+      "loss": 1.1069,
+      "step": 3700
+    },
+    {
+      "epoch": 1.1104358916647912,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00014500677812923633,
+      "loss": 1.1503,
+      "step": 3701
+    },
+    {
+      "epoch": 1.1107359891964887,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014499171561982227,
+      "loss": 1.0943,
+      "step": 3702
+    },
+    {
+      "epoch": 1.1110360867281868,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001449766531104082,
+      "loss": 1.0965,
+      "step": 3703
+    },
+    {
+      "epoch": 1.1113361842598846,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014496159060099412,
+      "loss": 1.0757,
+      "step": 3704
+    },
+    {
+      "epoch": 1.1116362817915824,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014494652809158006,
+      "loss": 1.0724,
+      "step": 3705
+    },
+    {
+      "epoch": 1.1119363793232802,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000144931465582166,
+      "loss": 1.1526,
+      "step": 3706
+    },
+    {
+      "epoch": 1.112236476854978,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014491640307275194,
+      "loss": 1.1681,
+      "step": 3707
+    },
+    {
+      "epoch": 1.1125365743866757,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014490134056333785,
+      "loss": 1.0228,
+      "step": 3708
+    },
+    {
+      "epoch": 1.1128366719183735,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001448862780539238,
+      "loss": 1.1148,
+      "step": 3709
+    },
+    {
+      "epoch": 1.1131367694500713,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014487121554450973,
+      "loss": 1.3013,
+      "step": 3710
+    },
+    {
+      "epoch": 1.1134368669817691,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014485615303509565,
+      "loss": 1.1411,
+      "step": 3711
+    },
+    {
+      "epoch": 1.113736964513467,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001448410905256816,
+      "loss": 1.0729,
+      "step": 3712
+    },
+    {
+      "epoch": 1.1140370620451647,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014482602801626753,
+      "loss": 0.9809,
+      "step": 3713
+    },
+    {
+      "epoch": 1.1143371595768625,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014481096550685344,
+      "loss": 1.0169,
+      "step": 3714
+    },
+    {
+      "epoch": 1.1146372571085603,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014479590299743938,
+      "loss": 1.0468,
+      "step": 3715
+    },
+    {
+      "epoch": 1.1149373546402581,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014478084048802532,
+      "loss": 1.0584,
+      "step": 3716
+    },
+    {
+      "epoch": 1.115237452171956,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014476577797861123,
+      "loss": 1.1091,
+      "step": 3717
+    },
+    {
+      "epoch": 1.1155375497036537,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00014475071546919717,
+      "loss": 1.0285,
+      "step": 3718
+    },
+    {
+      "epoch": 1.1158376472353515,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014473565295978311,
+      "loss": 1.0759,
+      "step": 3719
+    },
+    {
+      "epoch": 1.1161377447670493,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014472059045036903,
+      "loss": 1.1269,
+      "step": 3720
+    },
+    {
+      "epoch": 1.116437842298747,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00014470552794095497,
+      "loss": 1.3176,
+      "step": 3721
+    },
+    {
+      "epoch": 1.116737939830445,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001446904654315409,
+      "loss": 1.0723,
+      "step": 3722
+    },
+    {
+      "epoch": 1.1170380373621427,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014467540292212682,
+      "loss": 1.1318,
+      "step": 3723
+    },
+    {
+      "epoch": 1.1173381348938405,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014466034041271276,
+      "loss": 1.109,
+      "step": 3724
+    },
+    {
+      "epoch": 1.1176382324255383,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001446452779032987,
+      "loss": 1.0573,
+      "step": 3725
+    },
+    {
+      "epoch": 1.117938329957236,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014463021539388464,
+      "loss": 0.9602,
+      "step": 3726
+    },
+    {
+      "epoch": 1.1182384274889339,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014461515288447056,
+      "loss": 1.1069,
+      "step": 3727
+    },
+    {
+      "epoch": 1.1185385250206317,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001446000903750565,
+      "loss": 1.2682,
+      "step": 3728
+    },
+    {
+      "epoch": 1.1188386225523295,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014458502786564244,
+      "loss": 1.0924,
+      "step": 3729
+    },
+    {
+      "epoch": 1.1191387200840273,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014456996535622835,
+      "loss": 1.2534,
+      "step": 3730
+    },
+    {
+      "epoch": 1.119438817615725,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001445549028468143,
+      "loss": 1.068,
+      "step": 3731
+    },
+    {
+      "epoch": 1.1197389151474229,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014453984033740023,
+      "loss": 0.9726,
+      "step": 3732
+    },
+    {
+      "epoch": 1.1200390126791206,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014452477782798614,
+      "loss": 1.2314,
+      "step": 3733
+    },
+    {
+      "epoch": 1.1203391102108184,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014450971531857208,
+      "loss": 1.1111,
+      "step": 3734
+    },
+    {
+      "epoch": 1.1206392077425162,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014449465280915802,
+      "loss": 1.3168,
+      "step": 3735
+    },
+    {
+      "epoch": 1.120939305274214,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014447959029974394,
+      "loss": 1.058,
+      "step": 3736
+    },
+    {
+      "epoch": 1.1212394028059118,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014446452779032988,
+      "loss": 1.1347,
+      "step": 3737
+    },
+    {
+      "epoch": 1.1215395003376096,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014444946528091582,
+      "loss": 1.0479,
+      "step": 3738
+    },
+    {
+      "epoch": 1.1218395978693074,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00014443440277150173,
+      "loss": 1.1906,
+      "step": 3739
+    },
+    {
+      "epoch": 1.1221396954010054,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00014441934026208767,
+      "loss": 1.065,
+      "step": 3740
+    },
+    {
+      "epoch": 1.122439792932703,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001444042777526736,
+      "loss": 1.1315,
+      "step": 3741
+    },
+    {
+      "epoch": 1.122739890464401,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00014438921524325953,
+      "loss": 1.1157,
+      "step": 3742
+    },
+    {
+      "epoch": 1.1230399879960988,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014437415273384547,
+      "loss": 0.9715,
+      "step": 3743
+    },
+    {
+      "epoch": 1.1233400855277966,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001443590902244314,
+      "loss": 1.0997,
+      "step": 3744
+    },
+    {
+      "epoch": 1.1236401830594944,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014434402771501735,
+      "loss": 0.8815,
+      "step": 3745
+    },
+    {
+      "epoch": 1.1239402805911922,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014432896520560326,
+      "loss": 1.1341,
+      "step": 3746
+    },
+    {
+      "epoch": 1.12424037812289,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014431390269618917,
+      "loss": 1.0505,
+      "step": 3747
+    },
+    {
+      "epoch": 1.1245404756545878,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014429884018677514,
+      "loss": 0.9914,
+      "step": 3748
+    },
+    {
+      "epoch": 1.1248405731862856,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014428377767736105,
+      "loss": 1.0558,
+      "step": 3749
+    },
+    {
+      "epoch": 1.1251406707179834,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00014426871516794697,
+      "loss": 1.0019,
+      "step": 3750
+    },
+    {
+      "epoch": 1.1254407682496812,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014425365265853293,
+      "loss": 1.0625,
+      "step": 3751
+    },
+    {
+      "epoch": 1.125740865781379,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00014423859014911885,
+      "loss": 1.1532,
+      "step": 3752
+    },
+    {
+      "epoch": 1.1260409633130768,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001442235276397048,
+      "loss": 1.0837,
+      "step": 3753
+    },
+    {
+      "epoch": 1.1263410608447746,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014420846513029073,
+      "loss": 1.0919,
+      "step": 3754
+    },
+    {
+      "epoch": 1.1266411583764724,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014419340262087664,
+      "loss": 0.8997,
+      "step": 3755
+    },
+    {
+      "epoch": 1.1269412559081702,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014417834011146258,
+      "loss": 1.1182,
+      "step": 3756
+    },
+    {
+      "epoch": 1.127241353439868,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001441632776020485,
+      "loss": 1.0626,
+      "step": 3757
+    },
+    {
+      "epoch": 1.1275414509715658,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00014414821509263443,
+      "loss": 1.0487,
+      "step": 3758
+    },
+    {
+      "epoch": 1.1278415485032636,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014413315258322037,
+      "loss": 1.0575,
+      "step": 3759
+    },
+    {
+      "epoch": 1.1281416460349614,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001441180900738063,
+      "loss": 1.1667,
+      "step": 3760
+    },
+    {
+      "epoch": 1.1284417435666592,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014410302756439226,
+      "loss": 1.0557,
+      "step": 3761
+    },
+    {
+      "epoch": 1.128741841098357,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014408796505497817,
+      "loss": 0.9569,
+      "step": 3762
+    },
+    {
+      "epoch": 1.1290419386300548,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00014407290254556408,
+      "loss": 1.0383,
+      "step": 3763
+    },
+    {
+      "epoch": 1.1293420361617525,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014405784003615005,
+      "loss": 1.1795,
+      "step": 3764
+    },
+    {
+      "epoch": 1.1296421336934503,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014404277752673596,
+      "loss": 1.0922,
+      "step": 3765
+    },
+    {
+      "epoch": 1.1299422312251481,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00014402771501732188,
+      "loss": 1.1699,
+      "step": 3766
+    },
+    {
+      "epoch": 1.130242328756846,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014401265250790784,
+      "loss": 1.1704,
+      "step": 3767
+    },
+    {
+      "epoch": 1.1305424262885437,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014399758999849376,
+      "loss": 1.1268,
+      "step": 3768
+    },
+    {
+      "epoch": 1.1308425238202415,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014398252748907967,
+      "loss": 1.156,
+      "step": 3769
+    },
+    {
+      "epoch": 1.1311426213519393,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001439674649796656,
+      "loss": 0.9636,
+      "step": 3770
+    },
+    {
+      "epoch": 1.1314427188836371,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00014395240247025155,
+      "loss": 1.2994,
+      "step": 3771
+    },
+    {
+      "epoch": 1.131742816415335,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001439373399608375,
+      "loss": 0.9873,
+      "step": 3772
+    },
+    {
+      "epoch": 1.1320429139470327,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001439222774514234,
+      "loss": 1.0422,
+      "step": 3773
+    },
+    {
+      "epoch": 1.1323430114787305,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014390721494200934,
+      "loss": 1.0491,
+      "step": 3774
+    },
+    {
+      "epoch": 1.1326431090104283,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014389215243259528,
+      "loss": 0.9808,
+      "step": 3775
+    },
+    {
+      "epoch": 1.132943206542126,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001438770899231812,
+      "loss": 0.9784,
+      "step": 3776
+    },
+    {
+      "epoch": 1.133243304073824,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014386202741376714,
+      "loss": 1.012,
+      "step": 3777
+    },
+    {
+      "epoch": 1.1335434016055217,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014384696490435308,
+      "loss": 1.1254,
+      "step": 3778
+    },
+    {
+      "epoch": 1.1338434991372197,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.000143831902394939,
+      "loss": 1.07,
+      "step": 3779
+    },
+    {
+      "epoch": 1.1341435966689173,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014381683988552496,
+      "loss": 1.1312,
+      "step": 3780
+    },
+    {
+      "epoch": 1.1344436942006153,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00014380177737611087,
+      "loss": 1.058,
+      "step": 3781
+    },
+    {
+      "epoch": 1.1347437917323129,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00014378671486669679,
+      "loss": 1.1281,
+      "step": 3782
+    },
+    {
+      "epoch": 1.135043889264011,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014377165235728273,
+      "loss": 1.0642,
+      "step": 3783
+    },
+    {
+      "epoch": 1.1353439867957087,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014375658984786867,
+      "loss": 1.1909,
+      "step": 3784
+    },
+    {
+      "epoch": 1.1356440843274065,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014374152733845458,
+      "loss": 1.0169,
+      "step": 3785
+    },
+    {
+      "epoch": 1.1359441818591043,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014372646482904052,
+      "loss": 1.1055,
+      "step": 3786
+    },
+    {
+      "epoch": 1.136244279390802,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014371140231962646,
+      "loss": 1.0602,
+      "step": 3787
+    },
+    {
+      "epoch": 1.1365443769224999,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001436963398102124,
+      "loss": 1.0146,
+      "step": 3788
+    },
+    {
+      "epoch": 1.1368444744541977,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001436812773007983,
+      "loss": 1.1417,
+      "step": 3789
+    },
+    {
+      "epoch": 1.1371445719858955,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014366621479138425,
+      "loss": 1.0633,
+      "step": 3790
+    },
+    {
+      "epoch": 1.1374446695175933,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001436511522819702,
+      "loss": 1.11,
+      "step": 3791
+    },
+    {
+      "epoch": 1.137744767049291,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001436360897725561,
+      "loss": 0.9939,
+      "step": 3792
+    },
+    {
+      "epoch": 1.1380448645809889,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014362102726314205,
+      "loss": 1.0903,
+      "step": 3793
+    },
+    {
+      "epoch": 1.1383449621126867,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.000143605964753728,
+      "loss": 1.1556,
+      "step": 3794
+    },
+    {
+      "epoch": 1.1386450596443844,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001435909022443139,
+      "loss": 1.0903,
+      "step": 3795
+    },
+    {
+      "epoch": 1.1389451571760822,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014357583973489984,
+      "loss": 1.0031,
+      "step": 3796
+    },
+    {
+      "epoch": 1.13924525470778,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014356077722548578,
+      "loss": 1.1195,
+      "step": 3797
+    },
+    {
+      "epoch": 1.1395453522394778,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001435457147160717,
+      "loss": 1.0487,
+      "step": 3798
+    },
+    {
+      "epoch": 1.1398454497711756,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014353065220665763,
+      "loss": 0.9244,
+      "step": 3799
+    },
+    {
+      "epoch": 1.1401455473028734,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014351558969724357,
+      "loss": 1.057,
+      "step": 3800
+    },
+    {
+      "epoch": 1.1404456448345712,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001435005271878295,
+      "loss": 0.9652,
+      "step": 3801
+    },
+    {
+      "epoch": 1.140745742366269,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00014348546467841543,
+      "loss": 1.2478,
+      "step": 3802
+    },
+    {
+      "epoch": 1.1410458398979668,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014347040216900137,
+      "loss": 0.9983,
+      "step": 3803
+    },
+    {
+      "epoch": 1.1413459374296646,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00014345533965958728,
+      "loss": 0.9935,
+      "step": 3804
+    },
+    {
+      "epoch": 1.1416460349613624,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014344027715017322,
+      "loss": 0.9098,
+      "step": 3805
+    },
+    {
+      "epoch": 1.1419461324930602,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00014342521464075916,
+      "loss": 0.9486,
+      "step": 3806
+    },
+    {
+      "epoch": 1.142246230024758,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001434101521313451,
+      "loss": 1.0859,
+      "step": 3807
+    },
+    {
+      "epoch": 1.1425463275564558,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014339508962193102,
+      "loss": 1.0616,
+      "step": 3808
+    },
+    {
+      "epoch": 1.1428464250881536,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00014338002711251696,
+      "loss": 0.9465,
+      "step": 3809
+    },
+    {
+      "epoch": 1.1431465226198514,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001433649646031029,
+      "loss": 1.1768,
+      "step": 3810
+    },
+    {
+      "epoch": 1.1434466201515492,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001433499020936888,
+      "loss": 1.0287,
+      "step": 3811
+    },
+    {
+      "epoch": 1.143746717683247,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014333483958427475,
+      "loss": 1.1822,
+      "step": 3812
+    },
+    {
+      "epoch": 1.1440468152149448,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001433197770748607,
+      "loss": 0.9356,
+      "step": 3813
+    },
+    {
+      "epoch": 1.1443469127466426,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001433047145654466,
+      "loss": 1.0894,
+      "step": 3814
+    },
+    {
+      "epoch": 1.1446470102783404,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014328965205603254,
+      "loss": 1.0509,
+      "step": 3815
+    },
+    {
+      "epoch": 1.1449471078100382,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014327458954661848,
+      "loss": 1.0069,
+      "step": 3816
+    },
+    {
+      "epoch": 1.145247205341736,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001432595270372044,
+      "loss": 1.098,
+      "step": 3817
+    },
+    {
+      "epoch": 1.145547302873434,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014324446452779034,
+      "loss": 1.0392,
+      "step": 3818
+    },
+    {
+      "epoch": 1.1458474004051316,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014322940201837628,
+      "loss": 1.0768,
+      "step": 3819
+    },
+    {
+      "epoch": 1.1461474979368296,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001432143395089622,
+      "loss": 1.1294,
+      "step": 3820
+    },
+    {
+      "epoch": 1.1464475954685271,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00014319927699954813,
+      "loss": 1.0847,
+      "step": 3821
+    },
+    {
+      "epoch": 1.1467476930002252,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014318421449013407,
+      "loss": 1.2337,
+      "step": 3822
+    },
+    {
+      "epoch": 1.147047790531923,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014316915198071999,
+      "loss": 1.0332,
+      "step": 3823
+    },
+    {
+      "epoch": 1.1473478880636208,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00014315408947130593,
+      "loss": 1.1458,
+      "step": 3824
+    },
+    {
+      "epoch": 1.1476479855953186,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014313902696189187,
+      "loss": 1.0061,
+      "step": 3825
+    },
+    {
+      "epoch": 1.1479480831270163,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001431239644524778,
+      "loss": 0.9746,
+      "step": 3826
+    },
+    {
+      "epoch": 1.1482481806587141,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014310890194306372,
+      "loss": 0.9663,
+      "step": 3827
+    },
+    {
+      "epoch": 1.148548278190412,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014309383943364963,
+      "loss": 1.1192,
+      "step": 3828
+    },
+    {
+      "epoch": 1.1488483757221097,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001430787769242356,
+      "loss": 1.1417,
+      "step": 3829
+    },
+    {
+      "epoch": 1.1491484732538075,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001430637144148215,
+      "loss": 1.1511,
+      "step": 3830
+    },
+    {
+      "epoch": 1.1494485707855053,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00014304865190540743,
+      "loss": 1.242,
+      "step": 3831
+    },
+    {
+      "epoch": 1.1497486683172031,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001430335893959934,
+      "loss": 1.0068,
+      "step": 3832
+    },
+    {
+      "epoch": 1.150048765848901,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001430185268865793,
+      "loss": 1.1255,
+      "step": 3833
+    },
+    {
+      "epoch": 1.1503488633805987,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014300346437716525,
+      "loss": 0.9963,
+      "step": 3834
+    },
+    {
+      "epoch": 1.1506489609122965,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001429884018677512,
+      "loss": 1.2621,
+      "step": 3835
+    },
+    {
+      "epoch": 1.1509490584439943,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001429733393583371,
+      "loss": 1.0929,
+      "step": 3836
+    },
+    {
+      "epoch": 1.151249155975692,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014295827684892304,
+      "loss": 1.0621,
+      "step": 3837
+    },
+    {
+      "epoch": 1.15154925350739,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00014294321433950898,
+      "loss": 0.9112,
+      "step": 3838
+    },
+    {
+      "epoch": 1.1518493510390877,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001429281518300949,
+      "loss": 0.9958,
+      "step": 3839
+    },
+    {
+      "epoch": 1.1521494485707855,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014291308932068083,
+      "loss": 1.0247,
+      "step": 3840
+    },
+    {
+      "epoch": 1.1524495461024833,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014289802681126675,
+      "loss": 1.073,
+      "step": 3841
+    },
+    {
+      "epoch": 1.152749643634181,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014288296430185272,
+      "loss": 1.0943,
+      "step": 3842
+    },
+    {
+      "epoch": 1.1530497411658789,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014286790179243863,
+      "loss": 1.2023,
+      "step": 3843
+    },
+    {
+      "epoch": 1.1533498386975767,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014285283928302454,
+      "loss": 1.1856,
+      "step": 3844
+    },
+    {
+      "epoch": 1.1536499362292745,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001428377767736105,
+      "loss": 1.19,
+      "step": 3845
+    },
+    {
+      "epoch": 1.1539500337609723,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014282271426419642,
+      "loss": 1.0198,
+      "step": 3846
+    },
+    {
+      "epoch": 1.15425013129267,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00014280765175478234,
+      "loss": 1.0326,
+      "step": 3847
+    },
+    {
+      "epoch": 1.1545502288243679,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001427925892453683,
+      "loss": 1.0365,
+      "step": 3848
+    },
+    {
+      "epoch": 1.1548503263560657,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00014277752673595422,
+      "loss": 1.1578,
+      "step": 3849
+    },
+    {
+      "epoch": 1.1551504238877635,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014276246422654013,
+      "loss": 1.0897,
+      "step": 3850
+    },
+    {
+      "epoch": 1.1554505214194613,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001427474017171261,
+      "loss": 1.224,
+      "step": 3851
+    },
+    {
+      "epoch": 1.155750618951159,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.000142732339207712,
+      "loss": 1.2789,
+      "step": 3852
+    },
+    {
+      "epoch": 1.1560507164828568,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014271727669829795,
+      "loss": 1.1549,
+      "step": 3853
+    },
+    {
+      "epoch": 1.1563508140145546,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014270221418888386,
+      "loss": 1.1526,
+      "step": 3854
+    },
+    {
+      "epoch": 1.1566509115462524,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001426871516794698,
+      "loss": 1.1159,
+      "step": 3855
+    },
+    {
+      "epoch": 1.1569510090779502,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014267208917005574,
+      "loss": 1.0373,
+      "step": 3856
+    },
+    {
+      "epoch": 1.1572511066096483,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014265702666064166,
+      "loss": 1.0874,
+      "step": 3857
+    },
+    {
+      "epoch": 1.1575512041413458,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001426419641512276,
+      "loss": 1.1018,
+      "step": 3858
+    },
+    {
+      "epoch": 1.1578513016730438,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014262690164181354,
+      "loss": 1.0453,
+      "step": 3859
+    },
+    {
+      "epoch": 1.1581513992047414,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014261183913239945,
+      "loss": 1.0572,
+      "step": 3860
+    },
+    {
+      "epoch": 1.1584514967364394,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00014259677662298542,
+      "loss": 1.1511,
+      "step": 3861
+    },
+    {
+      "epoch": 1.1587515942681372,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014258171411357133,
+      "loss": 1.0758,
+      "step": 3862
+    },
+    {
+      "epoch": 1.159051691799835,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014256665160415725,
+      "loss": 1.1541,
+      "step": 3863
+    },
+    {
+      "epoch": 1.1593517893315328,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014255158909474319,
+      "loss": 1.2015,
+      "step": 3864
+    },
+    {
+      "epoch": 1.1596518868632306,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014253652658532913,
+      "loss": 1.1877,
+      "step": 3865
+    },
+    {
+      "epoch": 1.1599519843949284,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014252146407591504,
+      "loss": 1.0674,
+      "step": 3866
+    },
+    {
+      "epoch": 1.1602520819266262,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00014250640156650098,
+      "loss": 1.1999,
+      "step": 3867
+    },
+    {
+      "epoch": 1.160552179458324,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014249133905708692,
+      "loss": 1.1194,
+      "step": 3868
+    },
+    {
+      "epoch": 1.1608522769900218,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014247627654767286,
+      "loss": 1.0335,
+      "step": 3869
+    },
+    {
+      "epoch": 1.1611523745217196,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00014246121403825877,
+      "loss": 1.0615,
+      "step": 3870
+    },
+    {
+      "epoch": 1.1614524720534174,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001424461515288447,
+      "loss": 1.1807,
+      "step": 3871
+    },
+    {
+      "epoch": 1.1617525695851152,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014243108901943065,
+      "loss": 1.1587,
+      "step": 3872
+    },
+    {
+      "epoch": 1.162052667116813,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014241602651001657,
+      "loss": 1.1321,
+      "step": 3873
+    },
+    {
+      "epoch": 1.1623527646485108,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001424009640006025,
+      "loss": 0.9989,
+      "step": 3874
+    },
+    {
+      "epoch": 1.1626528621802086,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00014238590149118845,
+      "loss": 1.1396,
+      "step": 3875
+    },
+    {
+      "epoch": 1.1629529597119064,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014237083898177436,
+      "loss": 1.0403,
+      "step": 3876
+    },
+    {
+      "epoch": 1.1632530572436042,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001423557764723603,
+      "loss": 1.0861,
+      "step": 3877
+    },
+    {
+      "epoch": 1.163553154775302,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014234071396294624,
+      "loss": 1.1285,
+      "step": 3878
+    },
+    {
+      "epoch": 1.1638532523069998,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00014232565145353215,
+      "loss": 1.021,
+      "step": 3879
+    },
+    {
+      "epoch": 1.1641533498386976,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001423105889441181,
+      "loss": 1.0371,
+      "step": 3880
+    },
+    {
+      "epoch": 1.1644534473703954,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00014229552643470404,
+      "loss": 1.0264,
+      "step": 3881
+    },
+    {
+      "epoch": 1.1647535449020932,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014228046392528995,
+      "loss": 1.0713,
+      "step": 3882
+    },
+    {
+      "epoch": 1.165053642433791,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001422654014158759,
+      "loss": 1.0715,
+      "step": 3883
+    },
+    {
+      "epoch": 1.1653537399654887,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014225033890646183,
+      "loss": 1.0525,
+      "step": 3884
+    },
+    {
+      "epoch": 1.1656538374971865,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014223527639704774,
+      "loss": 1.0836,
+      "step": 3885
+    },
+    {
+      "epoch": 1.1659539350288843,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00014222021388763368,
+      "loss": 1.0116,
+      "step": 3886
+    },
+    {
+      "epoch": 1.1662540325605821,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014220515137821962,
+      "loss": 1.0559,
+      "step": 3887
+    },
+    {
+      "epoch": 1.16655413009228,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014219008886880556,
+      "loss": 1.1742,
+      "step": 3888
+    },
+    {
+      "epoch": 1.1668542276239777,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014217502635939148,
+      "loss": 1.1725,
+      "step": 3889
+    },
+    {
+      "epoch": 1.1671543251556755,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014215996384997742,
+      "loss": 1.1165,
+      "step": 3890
+    },
+    {
+      "epoch": 1.1674544226873733,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00014214490134056336,
+      "loss": 0.9855,
+      "step": 3891
+    },
+    {
+      "epoch": 1.1677545202190711,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014212983883114927,
+      "loss": 1.0058,
+      "step": 3892
+    },
+    {
+      "epoch": 1.168054617750769,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001421147763217352,
+      "loss": 1.1543,
+      "step": 3893
+    },
+    {
+      "epoch": 1.1683547152824667,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00014209971381232115,
+      "loss": 1.0445,
+      "step": 3894
+    },
+    {
+      "epoch": 1.1686548128141645,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00014208465130290706,
+      "loss": 1.055,
+      "step": 3895
+    },
+    {
+      "epoch": 1.1689549103458625,
+      "grad_norm": 1.25,
+      "learning_rate": 0.000142069588793493,
+      "loss": 1.1686,
+      "step": 3896
+    },
+    {
+      "epoch": 1.16925500787756,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00014205452628407894,
+      "loss": 1.1482,
+      "step": 3897
+    },
+    {
+      "epoch": 1.1695551054092581,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014203946377466486,
+      "loss": 1.2207,
+      "step": 3898
+    },
+    {
+      "epoch": 1.169855202940956,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001420244012652508,
+      "loss": 0.9693,
+      "step": 3899
+    },
+    {
+      "epoch": 1.1701553004726537,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014200933875583674,
+      "loss": 1.2164,
+      "step": 3900
+    },
+    {
+      "epoch": 1.1704553980043515,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014199427624642265,
+      "loss": 1.1945,
+      "step": 3901
+    },
+    {
+      "epoch": 1.1707554955360493,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001419792137370086,
+      "loss": 1.0521,
+      "step": 3902
+    },
+    {
+      "epoch": 1.171055593067747,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014196415122759453,
+      "loss": 1.1286,
+      "step": 3903
+    },
+    {
+      "epoch": 1.1713556905994449,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014194908871818045,
+      "loss": 0.9522,
+      "step": 3904
+    },
+    {
+      "epoch": 1.1716557881311427,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014193402620876639,
+      "loss": 1.0621,
+      "step": 3905
+    },
+    {
+      "epoch": 1.1719558856628405,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014191896369935233,
+      "loss": 1.2363,
+      "step": 3906
+    },
+    {
+      "epoch": 1.1722559831945383,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014190390118993827,
+      "loss": 0.9601,
+      "step": 3907
+    },
+    {
+      "epoch": 1.172556080726236,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014188883868052418,
+      "loss": 1.0339,
+      "step": 3908
+    },
+    {
+      "epoch": 1.1728561782579339,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00014187377617111012,
+      "loss": 1.0339,
+      "step": 3909
+    },
+    {
+      "epoch": 1.1731562757896317,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014185871366169606,
+      "loss": 1.0584,
+      "step": 3910
+    },
+    {
+      "epoch": 1.1734563733213295,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014184365115228197,
+      "loss": 0.9957,
+      "step": 3911
+    },
+    {
+      "epoch": 1.1737564708530273,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014182858864286789,
+      "loss": 1.0603,
+      "step": 3912
+    },
+    {
+      "epoch": 1.174056568384725,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014181352613345385,
+      "loss": 0.9498,
+      "step": 3913
+    },
+    {
+      "epoch": 1.1743566659164228,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014179846362403977,
+      "loss": 1.065,
+      "step": 3914
+    },
+    {
+      "epoch": 1.1746567634481206,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001417834011146257,
+      "loss": 1.0069,
+      "step": 3915
+    },
+    {
+      "epoch": 1.1749568609798184,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00014176833860521165,
+      "loss": 1.0055,
+      "step": 3916
+    },
+    {
+      "epoch": 1.1752569585115162,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014175327609579756,
+      "loss": 0.886,
+      "step": 3917
+    },
+    {
+      "epoch": 1.175557056043214,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001417382135863835,
+      "loss": 1.0336,
+      "step": 3918
+    },
+    {
+      "epoch": 1.1758571535749118,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00014172315107696944,
+      "loss": 1.0945,
+      "step": 3919
+    },
+    {
+      "epoch": 1.1761572511066096,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00014170808856755535,
+      "loss": 1.1452,
+      "step": 3920
+    },
+    {
+      "epoch": 1.1764573486383074,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001416930260581413,
+      "loss": 0.8759,
+      "step": 3921
+    },
+    {
+      "epoch": 1.1767574461700052,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014167796354872724,
+      "loss": 1.1405,
+      "step": 3922
+    },
+    {
+      "epoch": 1.177057543701703,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014166290103931318,
+      "loss": 1.1178,
+      "step": 3923
+    },
+    {
+      "epoch": 1.1773576412334008,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001416478385298991,
+      "loss": 1.0045,
+      "step": 3924
+    },
+    {
+      "epoch": 1.1776577387650986,
+      "grad_norm": 1.125,
+      "learning_rate": 0.000141632776020485,
+      "loss": 1.0567,
+      "step": 3925
+    },
+    {
+      "epoch": 1.1779578362967964,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014161771351107097,
+      "loss": 1.0583,
+      "step": 3926
+    },
+    {
+      "epoch": 1.1782579338284942,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014160265100165688,
+      "loss": 0.9839,
+      "step": 3927
+    },
+    {
+      "epoch": 1.178558031360192,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001415875884922428,
+      "loss": 1.0781,
+      "step": 3928
+    },
+    {
+      "epoch": 1.1788581288918898,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00014157252598282876,
+      "loss": 0.981,
+      "step": 3929
+    },
+    {
+      "epoch": 1.1791582264235876,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014155746347341468,
+      "loss": 1.0115,
+      "step": 3930
+    },
+    {
+      "epoch": 1.1794583239552854,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001415424009640006,
+      "loss": 1.1351,
+      "step": 3931
+    },
+    {
+      "epoch": 1.1797584214869832,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014152733845458656,
+      "loss": 1.037,
+      "step": 3932
+    },
+    {
+      "epoch": 1.1800585190186812,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014151227594517247,
+      "loss": 0.9697,
+      "step": 3933
+    },
+    {
+      "epoch": 1.1803586165503788,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001414972134357584,
+      "loss": 1.0363,
+      "step": 3934
+    },
+    {
+      "epoch": 1.1806587140820768,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00014148215092634432,
+      "loss": 1.1539,
+      "step": 3935
+    },
+    {
+      "epoch": 1.1809588116137744,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014146708841693026,
+      "loss": 0.9851,
+      "step": 3936
+    },
+    {
+      "epoch": 1.1812589091454724,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001414520259075162,
+      "loss": 1.1384,
+      "step": 3937
+    },
+    {
+      "epoch": 1.1815590066771702,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00014143696339810212,
+      "loss": 1.201,
+      "step": 3938
+    },
+    {
+      "epoch": 1.181859104208868,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014142190088868806,
+      "loss": 1.1508,
+      "step": 3939
+    },
+    {
+      "epoch": 1.1821592017405658,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.000141406838379274,
+      "loss": 1.0753,
+      "step": 3940
+    },
+    {
+      "epoch": 1.1824592992722636,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001413917758698599,
+      "loss": 1.075,
+      "step": 3941
+    },
+    {
+      "epoch": 1.1827593968039614,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014137671336044588,
+      "loss": 1.1966,
+      "step": 3942
+    },
+    {
+      "epoch": 1.1830594943356592,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001413616508510318,
+      "loss": 1.087,
+      "step": 3943
+    },
+    {
+      "epoch": 1.183359591867357,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001413465883416177,
+      "loss": 1.0609,
+      "step": 3944
+    },
+    {
+      "epoch": 1.1836596893990547,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014133152583220367,
+      "loss": 1.0775,
+      "step": 3945
+    },
+    {
+      "epoch": 1.1839597869307525,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014131646332278959,
+      "loss": 1.014,
+      "step": 3946
+    },
+    {
+      "epoch": 1.1842598844624503,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001413014008133755,
+      "loss": 1.0948,
+      "step": 3947
+    },
+    {
+      "epoch": 1.1845599819941481,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00014128633830396144,
+      "loss": 1.1061,
+      "step": 3948
+    },
+    {
+      "epoch": 1.184860079525846,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00014127127579454738,
+      "loss": 1.1023,
+      "step": 3949
+    },
+    {
+      "epoch": 1.1851601770575437,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00014125621328513332,
+      "loss": 0.973,
+      "step": 3950
+    },
+    {
+      "epoch": 1.1854602745892415,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014124115077571923,
+      "loss": 1.0235,
+      "step": 3951
+    },
+    {
+      "epoch": 1.1857603721209393,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014122608826630517,
+      "loss": 0.9941,
+      "step": 3952
+    },
+    {
+      "epoch": 1.1860604696526371,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014121102575689111,
+      "loss": 1.1284,
+      "step": 3953
+    },
+    {
+      "epoch": 1.186360567184335,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014119596324747703,
+      "loss": 1.0682,
+      "step": 3954
+    },
+    {
+      "epoch": 1.1866606647160327,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014118090073806297,
+      "loss": 1.0573,
+      "step": 3955
+    },
+    {
+      "epoch": 1.1869607622477305,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001411658382286489,
+      "loss": 1.2675,
+      "step": 3956
+    },
+    {
+      "epoch": 1.1872608597794283,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014115077571923482,
+      "loss": 1.0679,
+      "step": 3957
+    },
+    {
+      "epoch": 1.187560957311126,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014113571320982076,
+      "loss": 1.1597,
+      "step": 3958
+    },
+    {
+      "epoch": 1.187861054842824,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001411206507004067,
+      "loss": 1.2212,
+      "step": 3959
+    },
+    {
+      "epoch": 1.1881611523745217,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014110558819099261,
+      "loss": 1.0289,
+      "step": 3960
+    },
+    {
+      "epoch": 1.1884612499062195,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014109052568157855,
+      "loss": 0.9812,
+      "step": 3961
+    },
+    {
+      "epoch": 1.1887613474379173,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001410754631721645,
+      "loss": 1.0937,
+      "step": 3962
+    },
+    {
+      "epoch": 1.189061444969615,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001410604006627504,
+      "loss": 1.1736,
+      "step": 3963
+    },
+    {
+      "epoch": 1.1893615425013129,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014104533815333635,
+      "loss": 0.8703,
+      "step": 3964
+    },
+    {
+      "epoch": 1.1896616400330107,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001410302756439223,
+      "loss": 0.911,
+      "step": 3965
+    },
+    {
+      "epoch": 1.1899617375647085,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001410152131345082,
+      "loss": 1.0448,
+      "step": 3966
+    },
+    {
+      "epoch": 1.1902618350964063,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00014100015062509414,
+      "loss": 1.0946,
+      "step": 3967
+    },
+    {
+      "epoch": 1.190561932628104,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014098508811568008,
+      "loss": 0.8987,
+      "step": 3968
+    },
+    {
+      "epoch": 1.1908620301598019,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014097002560626602,
+      "loss": 1.0372,
+      "step": 3969
+    },
+    {
+      "epoch": 1.1911621276914997,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014095496309685194,
+      "loss": 1.1158,
+      "step": 3970
+    },
+    {
+      "epoch": 1.1914622252231974,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014093990058743788,
+      "loss": 1.1949,
+      "step": 3971
+    },
+    {
+      "epoch": 1.1917623227548955,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00014092483807802382,
+      "loss": 1.0735,
+      "step": 3972
+    },
+    {
+      "epoch": 1.192062420286593,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014090977556860973,
+      "loss": 1.0126,
+      "step": 3973
+    },
+    {
+      "epoch": 1.192362517818291,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00014089471305919567,
+      "loss": 1.0446,
+      "step": 3974
+    },
+    {
+      "epoch": 1.1926626153499886,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001408796505497816,
+      "loss": 1.0432,
+      "step": 3975
+    },
+    {
+      "epoch": 1.1929627128816866,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00014086458804036752,
+      "loss": 1.1812,
+      "step": 3976
+    },
+    {
+      "epoch": 1.1932628104133844,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014084952553095346,
+      "loss": 1.1601,
+      "step": 3977
+    },
+    {
+      "epoch": 1.1935629079450822,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001408344630215394,
+      "loss": 1.0271,
+      "step": 3978
+    },
+    {
+      "epoch": 1.19386300547678,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00014081940051212532,
+      "loss": 1.2395,
+      "step": 3979
+    },
+    {
+      "epoch": 1.1941631030084778,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014080433800271126,
+      "loss": 0.9473,
+      "step": 3980
+    },
+    {
+      "epoch": 1.1944632005401756,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001407892754932972,
+      "loss": 1.3639,
+      "step": 3981
+    },
+    {
+      "epoch": 1.1947632980718734,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001407742129838831,
+      "loss": 1.1013,
+      "step": 3982
+    },
+    {
+      "epoch": 1.1950633956035712,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00014075915047446905,
+      "loss": 1.1088,
+      "step": 3983
+    },
+    {
+      "epoch": 1.195363493135269,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.000140744087965055,
+      "loss": 1.1664,
+      "step": 3984
+    },
+    {
+      "epoch": 1.1956635906669668,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001407290254556409,
+      "loss": 1.0928,
+      "step": 3985
+    },
+    {
+      "epoch": 1.1959636881986646,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014071396294622685,
+      "loss": 1.2455,
+      "step": 3986
+    },
+    {
+      "epoch": 1.1962637857303624,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014069890043681279,
+      "loss": 1.1056,
+      "step": 3987
+    },
+    {
+      "epoch": 1.1965638832620602,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014068383792739873,
+      "loss": 1.127,
+      "step": 3988
+    },
+    {
+      "epoch": 1.196863980793758,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014066877541798464,
+      "loss": 1.0908,
+      "step": 3989
+    },
+    {
+      "epoch": 1.1971640783254558,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00014065371290857058,
+      "loss": 1.0094,
+      "step": 3990
+    },
+    {
+      "epoch": 1.1974641758571536,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00014063865039915652,
+      "loss": 1.0073,
+      "step": 3991
+    },
+    {
+      "epoch": 1.1977642733888514,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014062358788974243,
+      "loss": 1.0721,
+      "step": 3992
+    },
+    {
+      "epoch": 1.1980643709205492,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00014060852538032835,
+      "loss": 0.8422,
+      "step": 3993
+    },
+    {
+      "epoch": 1.198364468452247,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00014059346287091431,
+      "loss": 1.1921,
+      "step": 3994
+    },
+    {
+      "epoch": 1.1986645659839448,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00014057840036150023,
+      "loss": 1.1213,
+      "step": 3995
+    },
+    {
+      "epoch": 1.1989646635156426,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014056333785208617,
+      "loss": 1.0212,
+      "step": 3996
+    },
+    {
+      "epoch": 1.1992647610473404,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001405482753426721,
+      "loss": 1.1477,
+      "step": 3997
+    },
+    {
+      "epoch": 1.1995648585790382,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014053321283325802,
+      "loss": 1.0311,
+      "step": 3998
+    },
+    {
+      "epoch": 1.199864956110736,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014051815032384396,
+      "loss": 0.9763,
+      "step": 3999
+    },
+    {
+      "epoch": 1.2001650536424338,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001405030878144299,
+      "loss": 1.1422,
+      "step": 4000
+    },
+    {
+      "epoch": 1.2004651511741316,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014048802530501581,
+      "loss": 1.0358,
+      "step": 4001
+    },
+    {
+      "epoch": 1.2007652487058293,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014047296279560175,
+      "loss": 1.0296,
+      "step": 4002
+    },
+    {
+      "epoch": 1.2010653462375271,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001404579002861877,
+      "loss": 1.2324,
+      "step": 4003
+    },
+    {
+      "epoch": 1.201365443769225,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014044283777677364,
+      "loss": 1.1113,
+      "step": 4004
+    },
+    {
+      "epoch": 1.2016655413009227,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014042777526735955,
+      "loss": 1.0588,
+      "step": 4005
+    },
+    {
+      "epoch": 1.2019656388326205,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00014041271275794546,
+      "loss": 0.9934,
+      "step": 4006
+    },
+    {
+      "epoch": 1.2022657363643183,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00014039765024853143,
+      "loss": 1.0086,
+      "step": 4007
+    },
+    {
+      "epoch": 1.2025658338960161,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014038258773911734,
+      "loss": 1.1547,
+      "step": 4008
+    },
+    {
+      "epoch": 1.202865931427714,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014036752522970326,
+      "loss": 1.1366,
+      "step": 4009
+    },
+    {
+      "epoch": 1.2031660289594117,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00014035246272028922,
+      "loss": 1.0282,
+      "step": 4010
+    },
+    {
+      "epoch": 1.2034661264911097,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014033740021087514,
+      "loss": 1.003,
+      "step": 4011
+    },
+    {
+      "epoch": 1.2037662240228073,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00014032233770146105,
+      "loss": 1.0189,
+      "step": 4012
+    },
+    {
+      "epoch": 1.2040663215545053,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00014030727519204702,
+      "loss": 1.0303,
+      "step": 4013
+    },
+    {
+      "epoch": 1.204366419086203,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014029221268263293,
+      "loss": 1.0844,
+      "step": 4014
+    },
+    {
+      "epoch": 1.204666516617901,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014027715017321887,
+      "loss": 0.9968,
+      "step": 4015
+    },
+    {
+      "epoch": 1.2049666141495987,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001402620876638048,
+      "loss": 1.1583,
+      "step": 4016
+    },
+    {
+      "epoch": 1.2052667116812965,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00014024702515439072,
+      "loss": 1.1891,
+      "step": 4017
+    },
+    {
+      "epoch": 1.2055668092129943,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014023196264497666,
+      "loss": 1.1003,
+      "step": 4018
+    },
+    {
+      "epoch": 1.205866906744692,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014021690013556258,
+      "loss": 1.0483,
+      "step": 4019
+    },
+    {
+      "epoch": 1.20616700427639,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00014020183762614852,
+      "loss": 1.0233,
+      "step": 4020
+    },
+    {
+      "epoch": 1.2064671018080877,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014018677511673446,
+      "loss": 1.242,
+      "step": 4021
+    },
+    {
+      "epoch": 1.2067671993397855,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014017171260732037,
+      "loss": 1.0458,
+      "step": 4022
+    },
+    {
+      "epoch": 1.2070672968714833,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00014015665009790634,
+      "loss": 1.0472,
+      "step": 4023
+    },
+    {
+      "epoch": 1.207367394403181,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00014014158758849225,
+      "loss": 0.9151,
+      "step": 4024
+    },
+    {
+      "epoch": 1.2076674919348789,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00014012652507907817,
+      "loss": 1.0708,
+      "step": 4025
+    },
+    {
+      "epoch": 1.2079675894665767,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00014011146256966413,
+      "loss": 1.183,
+      "step": 4026
+    },
+    {
+      "epoch": 1.2082676869982745,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00014009640006025005,
+      "loss": 1.0084,
+      "step": 4027
+    },
+    {
+      "epoch": 1.2085677845299723,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00014008133755083596,
+      "loss": 1.0745,
+      "step": 4028
+    },
+    {
+      "epoch": 1.20886788206167,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001400662750414219,
+      "loss": 1.0562,
+      "step": 4029
+    },
+    {
+      "epoch": 1.2091679795933679,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00014005121253200784,
+      "loss": 1.3269,
+      "step": 4030
+    },
+    {
+      "epoch": 1.2094680771250657,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00014003615002259378,
+      "loss": 1.1288,
+      "step": 4031
+    },
+    {
+      "epoch": 1.2097681746567635,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001400210875131797,
+      "loss": 0.9805,
+      "step": 4032
+    },
+    {
+      "epoch": 1.2100682721884612,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00014000602500376563,
+      "loss": 1.1003,
+      "step": 4033
+    },
+    {
+      "epoch": 1.210368369720159,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013999096249435157,
+      "loss": 1.0722,
+      "step": 4034
+    },
+    {
+      "epoch": 1.2106684672518568,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001399758999849375,
+      "loss": 0.9454,
+      "step": 4035
+    },
+    {
+      "epoch": 1.2109685647835546,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013996083747552343,
+      "loss": 1.1972,
+      "step": 4036
+    },
+    {
+      "epoch": 1.2112686623152524,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013994577496610937,
+      "loss": 1.1495,
+      "step": 4037
+    },
+    {
+      "epoch": 1.2115687598469502,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013993071245669528,
+      "loss": 1.0486,
+      "step": 4038
+    },
+    {
+      "epoch": 1.211868857378648,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013991564994728122,
+      "loss": 1.0794,
+      "step": 4039
+    },
+    {
+      "epoch": 1.2121689549103458,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00013990058743786716,
+      "loss": 1.0449,
+      "step": 4040
+    },
+    {
+      "epoch": 1.2124690524420436,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013988552492845307,
+      "loss": 1.3785,
+      "step": 4041
+    },
+    {
+      "epoch": 1.2127691499737414,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013987046241903901,
+      "loss": 0.997,
+      "step": 4042
+    },
+    {
+      "epoch": 1.2130692475054392,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00013985539990962496,
+      "loss": 1.2011,
+      "step": 4043
+    },
+    {
+      "epoch": 1.213369345037137,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00013984033740021087,
+      "loss": 0.9537,
+      "step": 4044
+    },
+    {
+      "epoch": 1.2136694425688348,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001398252748907968,
+      "loss": 1.1246,
+      "step": 4045
+    },
+    {
+      "epoch": 1.2139695401005326,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013981021238138275,
+      "loss": 1.1152,
+      "step": 4046
+    },
+    {
+      "epoch": 1.2142696376322304,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013979514987196866,
+      "loss": 0.996,
+      "step": 4047
+    },
+    {
+      "epoch": 1.2145697351639282,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001397800873625546,
+      "loss": 0.9772,
+      "step": 4048
+    },
+    {
+      "epoch": 1.214869832695626,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013976502485314054,
+      "loss": 1.0265,
+      "step": 4049
+    },
+    {
+      "epoch": 1.215169930227324,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013974996234372648,
+      "loss": 0.9317,
+      "step": 4050
+    },
+    {
+      "epoch": 1.2154700277590216,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001397348998343124,
+      "loss": 1.0579,
+      "step": 4051
+    },
+    {
+      "epoch": 1.2157701252907196,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013971983732489834,
+      "loss": 1.0432,
+      "step": 4052
+    },
+    {
+      "epoch": 1.2160702228224172,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013970477481548428,
+      "loss": 1.2538,
+      "step": 4053
+    },
+    {
+      "epoch": 1.2163703203541152,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001396897123060702,
+      "loss": 1.0147,
+      "step": 4054
+    },
+    {
+      "epoch": 1.216670417885813,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013967464979665613,
+      "loss": 0.9689,
+      "step": 4055
+    },
+    {
+      "epoch": 1.2169705154175108,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013965958728724207,
+      "loss": 1.006,
+      "step": 4056
+    },
+    {
+      "epoch": 1.2172706129492086,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013964452477782798,
+      "loss": 1.1042,
+      "step": 4057
+    },
+    {
+      "epoch": 1.2175707104809064,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013962946226841392,
+      "loss": 1.0311,
+      "step": 4058
+    },
+    {
+      "epoch": 1.2178708080126042,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013961439975899986,
+      "loss": 1.0942,
+      "step": 4059
+    },
+    {
+      "epoch": 1.218170905544302,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013959933724958578,
+      "loss": 1.2513,
+      "step": 4060
+    },
+    {
+      "epoch": 1.2184710030759998,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013958427474017172,
+      "loss": 1.0003,
+      "step": 4061
+    },
+    {
+      "epoch": 1.2187711006076976,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00013956921223075766,
+      "loss": 1.1844,
+      "step": 4062
+    },
+    {
+      "epoch": 1.2190711981393954,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00013955414972134357,
+      "loss": 1.1738,
+      "step": 4063
+    },
+    {
+      "epoch": 1.2193712956710931,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001395390872119295,
+      "loss": 1.0029,
+      "step": 4064
+    },
+    {
+      "epoch": 1.219671393202791,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013952402470251545,
+      "loss": 0.9656,
+      "step": 4065
+    },
+    {
+      "epoch": 1.2199714907344887,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013950896219310137,
+      "loss": 1.0303,
+      "step": 4066
+    },
+    {
+      "epoch": 1.2202715882661865,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001394938996836873,
+      "loss": 1.0497,
+      "step": 4067
+    },
+    {
+      "epoch": 1.2205716857978843,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00013947883717427325,
+      "loss": 1.1172,
+      "step": 4068
+    },
+    {
+      "epoch": 1.2208717833295821,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00013946377466485919,
+      "loss": 1.1118,
+      "step": 4069
+    },
+    {
+      "epoch": 1.22117188086128,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001394487121554451,
+      "loss": 1.0887,
+      "step": 4070
+    },
+    {
+      "epoch": 1.2214719783929777,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013943364964603104,
+      "loss": 1.1623,
+      "step": 4071
+    },
+    {
+      "epoch": 1.2217720759246755,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013941858713661698,
+      "loss": 0.9744,
+      "step": 4072
+    },
+    {
+      "epoch": 1.2220721734563733,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001394035246272029,
+      "loss": 1.0952,
+      "step": 4073
+    },
+    {
+      "epoch": 1.222372270988071,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013938846211778883,
+      "loss": 1.1176,
+      "step": 4074
+    },
+    {
+      "epoch": 1.222672368519769,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013937339960837477,
+      "loss": 1.0073,
+      "step": 4075
+    },
+    {
+      "epoch": 1.2229724660514667,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001393583370989607,
+      "loss": 1.172,
+      "step": 4076
+    },
+    {
+      "epoch": 1.2232725635831645,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013934327458954663,
+      "loss": 1.0686,
+      "step": 4077
+    },
+    {
+      "epoch": 1.2235726611148623,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013932821208013257,
+      "loss": 1.0073,
+      "step": 4078
+    },
+    {
+      "epoch": 1.22387275864656,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013931314957071848,
+      "loss": 1.1208,
+      "step": 4079
+    },
+    {
+      "epoch": 1.2241728561782579,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013929808706130442,
+      "loss": 0.994,
+      "step": 4080
+    },
+    {
+      "epoch": 1.2244729537099557,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013928302455189036,
+      "loss": 1.1333,
+      "step": 4081
+    },
+    {
+      "epoch": 1.2247730512416535,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013926796204247627,
+      "loss": 1.1185,
+      "step": 4082
+    },
+    {
+      "epoch": 1.2250731487733513,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00013925289953306221,
+      "loss": 1.3046,
+      "step": 4083
+    },
+    {
+      "epoch": 1.225373246305049,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00013923783702364816,
+      "loss": 0.9325,
+      "step": 4084
+    },
+    {
+      "epoch": 1.2256733438367469,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001392227745142341,
+      "loss": 0.9492,
+      "step": 4085
+    },
+    {
+      "epoch": 1.2259734413684447,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00013920771200482,
+      "loss": 1.1451,
+      "step": 4086
+    },
+    {
+      "epoch": 1.2262735389001425,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013919264949540595,
+      "loss": 1.1969,
+      "step": 4087
+    },
+    {
+      "epoch": 1.2265736364318403,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001391775869859919,
+      "loss": 0.9897,
+      "step": 4088
+    },
+    {
+      "epoch": 1.2268737339635383,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001391625244765778,
+      "loss": 0.9852,
+      "step": 4089
+    },
+    {
+      "epoch": 1.2271738314952358,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013914746196716372,
+      "loss": 1.042,
+      "step": 4090
+    },
+    {
+      "epoch": 1.2274739290269339,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013913239945774968,
+      "loss": 0.9565,
+      "step": 4091
+    },
+    {
+      "epoch": 1.2277740265586314,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001391173369483356,
+      "loss": 1.0892,
+      "step": 4092
+    },
+    {
+      "epoch": 1.2280741240903295,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001391022744389215,
+      "loss": 1.1764,
+      "step": 4093
+    },
+    {
+      "epoch": 1.2283742216220273,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013908721192950748,
+      "loss": 1.0169,
+      "step": 4094
+    },
+    {
+      "epoch": 1.228674319153725,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001390721494200934,
+      "loss": 1.3301,
+      "step": 4095
+    },
+    {
+      "epoch": 1.2289744166854228,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00013905708691067933,
+      "loss": 1.1979,
+      "step": 4096
+    },
+    {
+      "epoch": 1.2292745142171206,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00013904202440126527,
+      "loss": 1.1296,
+      "step": 4097
+    },
+    {
+      "epoch": 1.2295746117488184,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013902696189185118,
+      "loss": 1.0619,
+      "step": 4098
+    },
+    {
+      "epoch": 1.2298747092805162,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00013901189938243712,
+      "loss": 1.1274,
+      "step": 4099
+    },
+    {
+      "epoch": 1.230174806812214,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013899683687302304,
+      "loss": 1.073,
+      "step": 4100
+    },
+    {
+      "epoch": 1.2304749043439118,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00013898177436360898,
+      "loss": 1.1049,
+      "step": 4101
+    },
+    {
+      "epoch": 1.2307750018756096,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00013896671185419492,
+      "loss": 0.9907,
+      "step": 4102
+    },
+    {
+      "epoch": 1.2310750994073074,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00013895164934478083,
+      "loss": 1.2332,
+      "step": 4103
+    },
+    {
+      "epoch": 1.2313751969390052,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001389365868353668,
+      "loss": 1.1063,
+      "step": 4104
+    },
+    {
+      "epoch": 1.231675294470703,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001389215243259527,
+      "loss": 1.1514,
+      "step": 4105
+    },
+    {
+      "epoch": 1.2319753920024008,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00013890646181653863,
+      "loss": 1.1004,
+      "step": 4106
+    },
+    {
+      "epoch": 1.2322754895340986,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001388913993071246,
+      "loss": 1.116,
+      "step": 4107
+    },
+    {
+      "epoch": 1.2325755870657964,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001388763367977105,
+      "loss": 1.0525,
+      "step": 4108
+    },
+    {
+      "epoch": 1.2328756845974942,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013886127428829642,
+      "loss": 1.235,
+      "step": 4109
+    },
+    {
+      "epoch": 1.233175782129192,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013884621177888239,
+      "loss": 1.063,
+      "step": 4110
+    },
+    {
+      "epoch": 1.2334758796608898,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001388311492694683,
+      "loss": 0.9827,
+      "step": 4111
+    },
+    {
+      "epoch": 1.2337759771925876,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013881608676005424,
+      "loss": 1.0584,
+      "step": 4112
+    },
+    {
+      "epoch": 1.2340760747242854,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013880102425064015,
+      "loss": 0.9898,
+      "step": 4113
+    },
+    {
+      "epoch": 1.2343761722559832,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001387859617412261,
+      "loss": 1.0277,
+      "step": 4114
+    },
+    {
+      "epoch": 1.234676269787681,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013877089923181203,
+      "loss": 1.1057,
+      "step": 4115
+    },
+    {
+      "epoch": 1.2349763673193788,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013875583672239795,
+      "loss": 0.9665,
+      "step": 4116
+    },
+    {
+      "epoch": 1.2352764648510766,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001387407742129839,
+      "loss": 0.9113,
+      "step": 4117
+    },
+    {
+      "epoch": 1.2355765623827744,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013872571170356983,
+      "loss": 1.182,
+      "step": 4118
+    },
+    {
+      "epoch": 1.2358766599144722,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013871064919415574,
+      "loss": 1.1186,
+      "step": 4119
+    },
+    {
+      "epoch": 1.23617675744617,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00013869558668474168,
+      "loss": 1.2172,
+      "step": 4120
+    },
+    {
+      "epoch": 1.2364768549778677,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013868052417532762,
+      "loss": 1.0447,
+      "step": 4121
+    },
+    {
+      "epoch": 1.2367769525095655,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013866546166591353,
+      "loss": 1.1996,
+      "step": 4122
+    },
+    {
+      "epoch": 1.2370770500412633,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001386503991564995,
+      "loss": 1.0238,
+      "step": 4123
+    },
+    {
+      "epoch": 1.2373771475729611,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00013863533664708542,
+      "loss": 1.0007,
+      "step": 4124
+    },
+    {
+      "epoch": 1.237677245104659,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013862027413767133,
+      "loss": 1.1042,
+      "step": 4125
+    },
+    {
+      "epoch": 1.2379773426363567,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013860521162825727,
+      "loss": 1.0681,
+      "step": 4126
+    },
+    {
+      "epoch": 1.2382774401680545,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001385901491188432,
+      "loss": 1.2642,
+      "step": 4127
+    },
+    {
+      "epoch": 1.2385775376997525,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013857508660942912,
+      "loss": 1.1651,
+      "step": 4128
+    },
+    {
+      "epoch": 1.2388776352314501,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013856002410001506,
+      "loss": 1.0,
+      "step": 4129
+    },
+    {
+      "epoch": 1.2391777327631481,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.000138544961590601,
+      "loss": 1.0636,
+      "step": 4130
+    },
+    {
+      "epoch": 1.2394778302948457,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013852989908118694,
+      "loss": 1.1262,
+      "step": 4131
+    },
+    {
+      "epoch": 1.2397779278265437,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013851483657177286,
+      "loss": 0.9991,
+      "step": 4132
+    },
+    {
+      "epoch": 1.2400780253582415,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001384997740623588,
+      "loss": 1.0627,
+      "step": 4133
+    },
+    {
+      "epoch": 1.2403781228899393,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013848471155294474,
+      "loss": 0.9991,
+      "step": 4134
+    },
+    {
+      "epoch": 1.2406782204216371,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013846964904353065,
+      "loss": 0.9844,
+      "step": 4135
+    },
+    {
+      "epoch": 1.240978317953335,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001384545865341166,
+      "loss": 1.0282,
+      "step": 4136
+    },
+    {
+      "epoch": 1.2412784154850327,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013843952402470253,
+      "loss": 1.0876,
+      "step": 4137
+    },
+    {
+      "epoch": 1.2415785130167305,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013842446151528844,
+      "loss": 1.0571,
+      "step": 4138
+    },
+    {
+      "epoch": 1.2418786105484283,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013840939900587438,
+      "loss": 1.1087,
+      "step": 4139
+    },
+    {
+      "epoch": 1.242178708080126,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013839433649646032,
+      "loss": 1.0198,
+      "step": 4140
+    },
+    {
+      "epoch": 1.2424788056118239,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013837927398704624,
+      "loss": 1.0282,
+      "step": 4141
+    },
+    {
+      "epoch": 1.2427789031435217,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013836421147763218,
+      "loss": 1.1918,
+      "step": 4142
+    },
+    {
+      "epoch": 1.2430790006752195,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013834914896821812,
+      "loss": 1.0737,
+      "step": 4143
+    },
+    {
+      "epoch": 1.2433790982069173,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013833408645880403,
+      "loss": 1.0788,
+      "step": 4144
+    },
+    {
+      "epoch": 1.243679195738615,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013831902394938997,
+      "loss": 1.1324,
+      "step": 4145
+    },
+    {
+      "epoch": 1.2439792932703129,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001383039614399759,
+      "loss": 0.9662,
+      "step": 4146
+    },
+    {
+      "epoch": 1.2442793908020107,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013828889893056183,
+      "loss": 1.1761,
+      "step": 4147
+    },
+    {
+      "epoch": 1.2445794883337085,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013827383642114777,
+      "loss": 1.0324,
+      "step": 4148
+    },
+    {
+      "epoch": 1.2448795858654063,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001382587739117337,
+      "loss": 1.1009,
+      "step": 4149
+    },
+    {
+      "epoch": 1.245179683397104,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013824371140231965,
+      "loss": 1.113,
+      "step": 4150
+    },
+    {
+      "epoch": 1.2454797809288018,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013822864889290556,
+      "loss": 1.148,
+      "step": 4151
+    },
+    {
+      "epoch": 1.2457798784604996,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001382135863834915,
+      "loss": 1.0119,
+      "step": 4152
+    },
+    {
+      "epoch": 1.2460799759921974,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013819852387407744,
+      "loss": 1.0546,
+      "step": 4153
+    },
+    {
+      "epoch": 1.2463800735238952,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013818346136466335,
+      "loss": 1.0128,
+      "step": 4154
+    },
+    {
+      "epoch": 1.246680171055593,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001381683988552493,
+      "loss": 1.0375,
+      "step": 4155
+    },
+    {
+      "epoch": 1.2469802685872908,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013815333634583523,
+      "loss": 1.0035,
+      "step": 4156
+    },
+    {
+      "epoch": 1.2472803661189886,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013813827383642115,
+      "loss": 1.1721,
+      "step": 4157
+    },
+    {
+      "epoch": 1.2475804636506864,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001381232113270071,
+      "loss": 1.1057,
+      "step": 4158
+    },
+    {
+      "epoch": 1.2478805611823842,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013810814881759303,
+      "loss": 1.0986,
+      "step": 4159
+    },
+    {
+      "epoch": 1.248180658714082,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00013809308630817894,
+      "loss": 1.0983,
+      "step": 4160
+    },
+    {
+      "epoch": 1.2484807562457798,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013807802379876488,
+      "loss": 0.9605,
+      "step": 4161
+    },
+    {
+      "epoch": 1.2487808537774776,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00013806296128935082,
+      "loss": 1.1229,
+      "step": 4162
+    },
+    {
+      "epoch": 1.2490809513091754,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00013804789877993673,
+      "loss": 1.1173,
+      "step": 4163
+    },
+    {
+      "epoch": 1.2493810488408732,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00013803283627052267,
+      "loss": 0.8382,
+      "step": 4164
+    },
+    {
+      "epoch": 1.249681146372571,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013801777376110862,
+      "loss": 1.0035,
+      "step": 4165
+    },
+    {
+      "epoch": 1.2499812439042688,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013800271125169456,
+      "loss": 1.0165,
+      "step": 4166
+    },
+    {
+      "epoch": 1.2502813414359668,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013798764874228047,
+      "loss": 1.1209,
+      "step": 4167
+    },
+    {
+      "epoch": 1.2505814389676644,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001379725862328664,
+      "loss": 1.0827,
+      "step": 4168
+    },
+    {
+      "epoch": 1.2508815364993624,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013795752372345235,
+      "loss": 1.0587,
+      "step": 4169
+    },
+    {
+      "epoch": 1.25118163403106,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013794246121403826,
+      "loss": 0.8974,
+      "step": 4170
+    },
+    {
+      "epoch": 1.251481731562758,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013792739870462418,
+      "loss": 1.2039,
+      "step": 4171
+    },
+    {
+      "epoch": 1.2517818290944556,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013791233619521014,
+      "loss": 1.0345,
+      "step": 4172
+    },
+    {
+      "epoch": 1.2520819266261536,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00013789727368579606,
+      "loss": 1.0223,
+      "step": 4173
+    },
+    {
+      "epoch": 1.2523820241578514,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013788221117638197,
+      "loss": 1.1769,
+      "step": 4174
+    },
+    {
+      "epoch": 1.2526821216895492,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013786714866696794,
+      "loss": 0.9653,
+      "step": 4175
+    },
+    {
+      "epoch": 1.252982219221247,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00013785208615755385,
+      "loss": 1.0096,
+      "step": 4176
+    },
+    {
+      "epoch": 1.2532823167529448,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001378370236481398,
+      "loss": 0.9926,
+      "step": 4177
+    },
+    {
+      "epoch": 1.2535824142846426,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013782196113872573,
+      "loss": 1.185,
+      "step": 4178
+    },
+    {
+      "epoch": 1.2538825118163404,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013780689862931164,
+      "loss": 1.0114,
+      "step": 4179
+    },
+    {
+      "epoch": 1.2541826093480382,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013779183611989758,
+      "loss": 0.9803,
+      "step": 4180
+    },
+    {
+      "epoch": 1.254482706879736,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013777677361048352,
+      "loss": 1.1185,
+      "step": 4181
+    },
+    {
+      "epoch": 1.2547828044114337,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00013776171110106944,
+      "loss": 1.1784,
+      "step": 4182
+    },
+    {
+      "epoch": 1.2550829019431315,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013774664859165538,
+      "loss": 1.1409,
+      "step": 4183
+    },
+    {
+      "epoch": 1.2553829994748293,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001377315860822413,
+      "loss": 1.1766,
+      "step": 4184
+    },
+    {
+      "epoch": 1.2556830970065271,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013771652357282726,
+      "loss": 1.0022,
+      "step": 4185
+    },
+    {
+      "epoch": 1.255983194538225,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013770146106341317,
+      "loss": 0.9008,
+      "step": 4186
+    },
+    {
+      "epoch": 1.2562832920699227,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013768639855399909,
+      "loss": 1.2298,
+      "step": 4187
+    },
+    {
+      "epoch": 1.2565833896016205,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013767133604458505,
+      "loss": 1.1006,
+      "step": 4188
+    },
+    {
+      "epoch": 1.2568834871333183,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013765627353517097,
+      "loss": 1.0589,
+      "step": 4189
+    },
+    {
+      "epoch": 1.2571835846650161,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013764121102575688,
+      "loss": 1.0201,
+      "step": 4190
+    },
+    {
+      "epoch": 1.257483682196714,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013762614851634285,
+      "loss": 1.0517,
+      "step": 4191
+    },
+    {
+      "epoch": 1.2577837797284117,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013761108600692876,
+      "loss": 1.0952,
+      "step": 4192
+    },
+    {
+      "epoch": 1.2580838772601095,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001375960234975147,
+      "loss": 1.1363,
+      "step": 4193
+    },
+    {
+      "epoch": 1.2583839747918073,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013758096098810064,
+      "loss": 0.9492,
+      "step": 4194
+    },
+    {
+      "epoch": 1.258684072323505,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013756589847868655,
+      "loss": 1.1322,
+      "step": 4195
+    },
+    {
+      "epoch": 1.258984169855203,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001375508359692725,
+      "loss": 1.0895,
+      "step": 4196
+    },
+    {
+      "epoch": 1.2592842673869007,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001375357734598584,
+      "loss": 1.2526,
+      "step": 4197
+    },
+    {
+      "epoch": 1.2595843649185985,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013752071095044435,
+      "loss": 1.0556,
+      "step": 4198
+    },
+    {
+      "epoch": 1.2598844624502963,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001375056484410303,
+      "loss": 1.1417,
+      "step": 4199
+    },
+    {
+      "epoch": 1.260184559981994,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001374905859316162,
+      "loss": 1.1541,
+      "step": 4200
+    },
+    {
+      "epoch": 1.2604846575136919,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013747552342220214,
+      "loss": 1.004,
+      "step": 4201
+    },
+    {
+      "epoch": 1.2607847550453897,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013746046091278808,
+      "loss": 1.0028,
+      "step": 4202
+    },
+    {
+      "epoch": 1.2610848525770875,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.000137445398403374,
+      "loss": 1.0785,
+      "step": 4203
+    },
+    {
+      "epoch": 1.2613849501087855,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013743033589395996,
+      "loss": 1.141,
+      "step": 4204
+    },
+    {
+      "epoch": 1.261685047640483,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013741527338454588,
+      "loss": 0.9867,
+      "step": 4205
+    },
+    {
+      "epoch": 1.261985145172181,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001374002108751318,
+      "loss": 1.0994,
+      "step": 4206
+    },
+    {
+      "epoch": 1.2622852427038787,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013738514836571773,
+      "loss": 1.144,
+      "step": 4207
+    },
+    {
+      "epoch": 1.2625853402355767,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013737008585630367,
+      "loss": 1.0597,
+      "step": 4208
+    },
+    {
+      "epoch": 1.2628854377672742,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013735502334688958,
+      "loss": 1.0119,
+      "step": 4209
+    },
+    {
+      "epoch": 1.2631855352989723,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013733996083747552,
+      "loss": 1.0234,
+      "step": 4210
+    },
+    {
+      "epoch": 1.2634856328306698,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013732489832806146,
+      "loss": 1.1177,
+      "step": 4211
+    },
+    {
+      "epoch": 1.2637857303623679,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001373098358186474,
+      "loss": 0.9775,
+      "step": 4212
+    },
+    {
+      "epoch": 1.2640858278940657,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00013729477330923332,
+      "loss": 1.2485,
+      "step": 4213
+    },
+    {
+      "epoch": 1.2643859254257634,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013727971079981926,
+      "loss": 1.1858,
+      "step": 4214
+    },
+    {
+      "epoch": 1.2646860229574612,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001372646482904052,
+      "loss": 1.1105,
+      "step": 4215
+    },
+    {
+      "epoch": 1.264986120489159,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001372495857809911,
+      "loss": 1.0018,
+      "step": 4216
+    },
+    {
+      "epoch": 1.2652862180208568,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00013723452327157705,
+      "loss": 1.3059,
+      "step": 4217
+    },
+    {
+      "epoch": 1.2655863155525546,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.000137219460762163,
+      "loss": 1.1305,
+      "step": 4218
+    },
+    {
+      "epoch": 1.2658864130842524,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001372043982527489,
+      "loss": 1.015,
+      "step": 4219
+    },
+    {
+      "epoch": 1.2661865106159502,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013718933574333484,
+      "loss": 1.0393,
+      "step": 4220
+    },
+    {
+      "epoch": 1.266486608147648,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00013717427323392078,
+      "loss": 1.0849,
+      "step": 4221
+    },
+    {
+      "epoch": 1.2667867056793458,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001371592107245067,
+      "loss": 1.1183,
+      "step": 4222
+    },
+    {
+      "epoch": 1.2670868032110436,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00013714414821509264,
+      "loss": 0.8868,
+      "step": 4223
+    },
+    {
+      "epoch": 1.2673869007427414,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013712908570567858,
+      "loss": 1.0639,
+      "step": 4224
+    },
+    {
+      "epoch": 1.2676869982744392,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001371140231962645,
+      "loss": 1.1056,
+      "step": 4225
+    },
+    {
+      "epoch": 1.267987095806137,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00013709896068685043,
+      "loss": 1.2317,
+      "step": 4226
+    },
+    {
+      "epoch": 1.2682871933378348,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013708389817743637,
+      "loss": 1.1127,
+      "step": 4227
+    },
+    {
+      "epoch": 1.2685872908695326,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013706883566802229,
+      "loss": 1.0751,
+      "step": 4228
+    },
+    {
+      "epoch": 1.2688873884012304,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013705377315860823,
+      "loss": 1.0605,
+      "step": 4229
+    },
+    {
+      "epoch": 1.2691874859329282,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013703871064919417,
+      "loss": 1.0222,
+      "step": 4230
+    },
+    {
+      "epoch": 1.269487583464626,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001370236481397801,
+      "loss": 1.2494,
+      "step": 4231
+    },
+    {
+      "epoch": 1.2697876809963238,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013700858563036602,
+      "loss": 1.0829,
+      "step": 4232
+    },
+    {
+      "epoch": 1.2700877785280216,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013699352312095196,
+      "loss": 1.1911,
+      "step": 4233
+    },
+    {
+      "epoch": 1.2703878760597194,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001369784606115379,
+      "loss": 0.9955,
+      "step": 4234
+    },
+    {
+      "epoch": 1.2706879735914172,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001369633981021238,
+      "loss": 0.9612,
+      "step": 4235
+    },
+    {
+      "epoch": 1.270988071123115,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00013694833559270975,
+      "loss": 1.0999,
+      "step": 4236
+    },
+    {
+      "epoch": 1.2712881686548128,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001369332730832957,
+      "loss": 1.0468,
+      "step": 4237
+    },
+    {
+      "epoch": 1.2715882661865106,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001369182105738816,
+      "loss": 1.0075,
+      "step": 4238
+    },
+    {
+      "epoch": 1.2718883637182083,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013690314806446755,
+      "loss": 0.9667,
+      "step": 4239
+    },
+    {
+      "epoch": 1.2721884612499061,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001368880855550535,
+      "loss": 0.9695,
+      "step": 4240
+    },
+    {
+      "epoch": 1.272488558781604,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001368730230456394,
+      "loss": 1.0413,
+      "step": 4241
+    },
+    {
+      "epoch": 1.2727886563133017,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013685796053622534,
+      "loss": 1.1303,
+      "step": 4242
+    },
+    {
+      "epoch": 1.2730887538449998,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013684289802681128,
+      "loss": 1.057,
+      "step": 4243
+    },
+    {
+      "epoch": 1.2733888513766973,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001368278355173972,
+      "loss": 0.9531,
+      "step": 4244
+    },
+    {
+      "epoch": 1.2736889489083953,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00013681277300798313,
+      "loss": 1.0521,
+      "step": 4245
+    },
+    {
+      "epoch": 1.273989046440093,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013679771049856908,
+      "loss": 1.1745,
+      "step": 4246
+    },
+    {
+      "epoch": 1.274289143971791,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013678264798915502,
+      "loss": 1.0241,
+      "step": 4247
+    },
+    {
+      "epoch": 1.2745892415034885,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013676758547974093,
+      "loss": 1.031,
+      "step": 4248
+    },
+    {
+      "epoch": 1.2748893390351865,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013675252297032687,
+      "loss": 1.0589,
+      "step": 4249
+    },
+    {
+      "epoch": 1.275189436566884,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001367374604609128,
+      "loss": 0.979,
+      "step": 4250
+    },
+    {
+      "epoch": 1.2754895340985821,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013672239795149872,
+      "loss": 1.073,
+      "step": 4251
+    },
+    {
+      "epoch": 1.27578963163028,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013670733544208466,
+      "loss": 1.0899,
+      "step": 4252
+    },
+    {
+      "epoch": 1.2760897291619777,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001366922729326706,
+      "loss": 1.0968,
+      "step": 4253
+    },
+    {
+      "epoch": 1.2763898266936755,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013667721042325652,
+      "loss": 1.0766,
+      "step": 4254
+    },
+    {
+      "epoch": 1.2766899242253733,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00013666214791384243,
+      "loss": 1.0139,
+      "step": 4255
+    },
+    {
+      "epoch": 1.276990021757071,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001366470854044284,
+      "loss": 1.2624,
+      "step": 4256
+    },
+    {
+      "epoch": 1.277290119288769,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001366320228950143,
+      "loss": 0.951,
+      "step": 4257
+    },
+    {
+      "epoch": 1.2775902168204667,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013661696038560025,
+      "loss": 1.0934,
+      "step": 4258
+    },
+    {
+      "epoch": 1.2778903143521645,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001366018978761862,
+      "loss": 1.0777,
+      "step": 4259
+    },
+    {
+      "epoch": 1.2781904118838623,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001365868353667721,
+      "loss": 1.0174,
+      "step": 4260
+    },
+    {
+      "epoch": 1.27849050941556,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00013657177285735804,
+      "loss": 1.0127,
+      "step": 4261
+    },
+    {
+      "epoch": 1.2787906069472579,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00013655671034794398,
+      "loss": 1.0585,
+      "step": 4262
+    },
+    {
+      "epoch": 1.2790907044789557,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001365416478385299,
+      "loss": 1.1264,
+      "step": 4263
+    },
+    {
+      "epoch": 1.2793908020106535,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00013652658532911584,
+      "loss": 1.2203,
+      "step": 4264
+    },
+    {
+      "epoch": 1.2796908995423513,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013651152281970175,
+      "loss": 1.0925,
+      "step": 4265
+    },
+    {
+      "epoch": 1.279990997074049,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013649646031028772,
+      "loss": 1.0924,
+      "step": 4266
+    },
+    {
+      "epoch": 1.2802910946057469,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013648139780087363,
+      "loss": 1.135,
+      "step": 4267
+    },
+    {
+      "epoch": 1.2805911921374447,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013646633529145955,
+      "loss": 1.1553,
+      "step": 4268
+    },
+    {
+      "epoch": 1.2808912896691425,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001364512727820455,
+      "loss": 0.9041,
+      "step": 4269
+    },
+    {
+      "epoch": 1.2811913872008402,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013643621027263143,
+      "loss": 1.1199,
+      "step": 4270
+    },
+    {
+      "epoch": 1.281491484732538,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013642114776321734,
+      "loss": 1.1923,
+      "step": 4271
+    },
+    {
+      "epoch": 1.2817915822642358,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001364060852538033,
+      "loss": 1.0629,
+      "step": 4272
+    },
+    {
+      "epoch": 1.2820916797959336,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00013639102274438922,
+      "loss": 1.2119,
+      "step": 4273
+    },
+    {
+      "epoch": 1.2823917773276314,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00013637596023497516,
+      "loss": 1.0964,
+      "step": 4274
+    },
+    {
+      "epoch": 1.2826918748593292,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001363608977255611,
+      "loss": 1.0864,
+      "step": 4275
+    },
+    {
+      "epoch": 1.282991972391027,
+      "grad_norm": 1.125,
+      "learning_rate": 0.000136345835216147,
+      "loss": 1.0173,
+      "step": 4276
+    },
+    {
+      "epoch": 1.2832920699227248,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013633077270673295,
+      "loss": 0.9633,
+      "step": 4277
+    },
+    {
+      "epoch": 1.2835921674544226,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00013631571019731887,
+      "loss": 0.9929,
+      "step": 4278
+    },
+    {
+      "epoch": 1.2838922649861204,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001363006476879048,
+      "loss": 1.2773,
+      "step": 4279
+    },
+    {
+      "epoch": 1.2841923625178182,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013628558517849075,
+      "loss": 1.07,
+      "step": 4280
+    },
+    {
+      "epoch": 1.284492460049516,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013627052266907666,
+      "loss": 1.1098,
+      "step": 4281
+    },
+    {
+      "epoch": 1.284792557581214,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001362554601596626,
+      "loss": 1.0514,
+      "step": 4282
+    },
+    {
+      "epoch": 1.2850926551129116,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013624039765024854,
+      "loss": 1.0717,
+      "step": 4283
+    },
+    {
+      "epoch": 1.2853927526446096,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00013622533514083445,
+      "loss": 1.0506,
+      "step": 4284
+    },
+    {
+      "epoch": 1.2856928501763072,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00013621027263142042,
+      "loss": 1.2502,
+      "step": 4285
+    },
+    {
+      "epoch": 1.2859929477080052,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013619521012200634,
+      "loss": 1.0847,
+      "step": 4286
+    },
+    {
+      "epoch": 1.2862930452397028,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013618014761259225,
+      "loss": 1.0668,
+      "step": 4287
+    },
+    {
+      "epoch": 1.2865931427714008,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013616508510317822,
+      "loss": 1.0624,
+      "step": 4288
+    },
+    {
+      "epoch": 1.2868932403030984,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013615002259376413,
+      "loss": 0.842,
+      "step": 4289
+    },
+    {
+      "epoch": 1.2871933378347964,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013613496008435004,
+      "loss": 1.1126,
+      "step": 4290
+    },
+    {
+      "epoch": 1.2874934353664942,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00013611989757493598,
+      "loss": 1.0577,
+      "step": 4291
+    },
+    {
+      "epoch": 1.287793532898192,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013610483506552192,
+      "loss": 1.0279,
+      "step": 4292
+    },
+    {
+      "epoch": 1.2880936304298898,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013608977255610786,
+      "loss": 1.1787,
+      "step": 4293
+    },
+    {
+      "epoch": 1.2883937279615876,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013607471004669378,
+      "loss": 0.9958,
+      "step": 4294
+    },
+    {
+      "epoch": 1.2886938254932854,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013605964753727972,
+      "loss": 0.9342,
+      "step": 4295
+    },
+    {
+      "epoch": 1.2889939230249832,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013604458502786566,
+      "loss": 1.14,
+      "step": 4296
+    },
+    {
+      "epoch": 1.289294020556681,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013602952251845157,
+      "loss": 1.0167,
+      "step": 4297
+    },
+    {
+      "epoch": 1.2895941180883788,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001360144600090375,
+      "loss": 1.1466,
+      "step": 4298
+    },
+    {
+      "epoch": 1.2898942156200766,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00013599939749962345,
+      "loss": 1.1212,
+      "step": 4299
+    },
+    {
+      "epoch": 1.2901943131517744,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00013598433499020936,
+      "loss": 1.0865,
+      "step": 4300
+    },
+    {
+      "epoch": 1.2904944106834721,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013596927248079533,
+      "loss": 0.9391,
+      "step": 4301
+    },
+    {
+      "epoch": 1.29079450821517,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013595420997138124,
+      "loss": 1.0269,
+      "step": 4302
+    },
+    {
+      "epoch": 1.2910946057468677,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013593914746196716,
+      "loss": 1.1162,
+      "step": 4303
+    },
+    {
+      "epoch": 1.2913947032785655,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001359240849525531,
+      "loss": 0.9281,
+      "step": 4304
+    },
+    {
+      "epoch": 1.2916948008102633,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013590902244313904,
+      "loss": 1.1514,
+      "step": 4305
+    },
+    {
+      "epoch": 1.2919948983419611,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013589395993372495,
+      "loss": 1.1008,
+      "step": 4306
+    },
+    {
+      "epoch": 1.292294995873659,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001358788974243109,
+      "loss": 1.0663,
+      "step": 4307
+    },
+    {
+      "epoch": 1.2925950934053567,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013586383491489683,
+      "loss": 0.996,
+      "step": 4308
+    },
+    {
+      "epoch": 1.2928951909370545,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013584877240548275,
+      "loss": 0.7994,
+      "step": 4309
+    },
+    {
+      "epoch": 1.2931952884687523,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013583370989606869,
+      "loss": 1.0517,
+      "step": 4310
+    },
+    {
+      "epoch": 1.29349538600045,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013581864738665463,
+      "loss": 1.0839,
+      "step": 4311
+    },
+    {
+      "epoch": 1.293795483532148,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013580358487724057,
+      "loss": 1.0388,
+      "step": 4312
+    },
+    {
+      "epoch": 1.2940955810638457,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013578852236782648,
+      "loss": 1.102,
+      "step": 4313
+    },
+    {
+      "epoch": 1.2943956785955435,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013577345985841242,
+      "loss": 1.1384,
+      "step": 4314
+    },
+    {
+      "epoch": 1.2946957761272413,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013575839734899836,
+      "loss": 0.9855,
+      "step": 4315
+    },
+    {
+      "epoch": 1.294995873658939,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013574333483958427,
+      "loss": 1.0868,
+      "step": 4316
+    },
+    {
+      "epoch": 1.2952959711906369,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001357282723301702,
+      "loss": 1.0424,
+      "step": 4317
+    },
+    {
+      "epoch": 1.2955960687223347,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013571320982075615,
+      "loss": 1.1383,
+      "step": 4318
+    },
+    {
+      "epoch": 1.2958961662540325,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013569814731134207,
+      "loss": 1.1998,
+      "step": 4319
+    },
+    {
+      "epoch": 1.2961962637857303,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.000135683084801928,
+      "loss": 0.8856,
+      "step": 4320
+    },
+    {
+      "epoch": 1.2964963613174283,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013566802229251395,
+      "loss": 1.0496,
+      "step": 4321
+    },
+    {
+      "epoch": 1.2967964588491259,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00013565295978309986,
+      "loss": 0.996,
+      "step": 4322
+    },
+    {
+      "epoch": 1.2970965563808239,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001356378972736858,
+      "loss": 1.1242,
+      "step": 4323
+    },
+    {
+      "epoch": 1.2973966539125215,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013562283476427174,
+      "loss": 1.0692,
+      "step": 4324
+    },
+    {
+      "epoch": 1.2976967514442195,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013560777225485765,
+      "loss": 0.9355,
+      "step": 4325
+    },
+    {
+      "epoch": 1.297996848975917,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001355927097454436,
+      "loss": 1.0633,
+      "step": 4326
+    },
+    {
+      "epoch": 1.298296946507615,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013557764723602954,
+      "loss": 0.9317,
+      "step": 4327
+    },
+    {
+      "epoch": 1.2985970440393126,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00013556258472661548,
+      "loss": 1.0567,
+      "step": 4328
+    },
+    {
+      "epoch": 1.2988971415710107,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001355475222172014,
+      "loss": 1.0931,
+      "step": 4329
+    },
+    {
+      "epoch": 1.2991972391027085,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013553245970778733,
+      "loss": 1.1843,
+      "step": 4330
+    },
+    {
+      "epoch": 1.2994973366344063,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00013551739719837327,
+      "loss": 1.0635,
+      "step": 4331
+    },
+    {
+      "epoch": 1.299797434166104,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00013550233468895918,
+      "loss": 1.1147,
+      "step": 4332
+    },
+    {
+      "epoch": 1.3000975316978018,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013548727217954512,
+      "loss": 1.1229,
+      "step": 4333
+    },
+    {
+      "epoch": 1.3003976292294996,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013547220967013106,
+      "loss": 1.0055,
+      "step": 4334
+    },
+    {
+      "epoch": 1.3006977267611974,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013545714716071698,
+      "loss": 0.9817,
+      "step": 4335
+    },
+    {
+      "epoch": 1.3009978242928952,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001354420846513029,
+      "loss": 1.0382,
+      "step": 4336
+    },
+    {
+      "epoch": 1.301297921824593,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013542702214188886,
+      "loss": 1.0653,
+      "step": 4337
+    },
+    {
+      "epoch": 1.3015980193562908,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00013541195963247477,
+      "loss": 1.0234,
+      "step": 4338
+    },
+    {
+      "epoch": 1.3018981168879886,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001353968971230607,
+      "loss": 1.0935,
+      "step": 4339
+    },
+    {
+      "epoch": 1.3021982144196864,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00013538183461364665,
+      "loss": 1.0663,
+      "step": 4340
+    },
+    {
+      "epoch": 1.3024983119513842,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00013536677210423256,
+      "loss": 1.2203,
+      "step": 4341
+    },
+    {
+      "epoch": 1.302798409483082,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001353517095948185,
+      "loss": 0.9314,
+      "step": 4342
+    },
+    {
+      "epoch": 1.3030985070147798,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013533664708540444,
+      "loss": 1.0696,
+      "step": 4343
+    },
+    {
+      "epoch": 1.3033986045464776,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013532158457599036,
+      "loss": 1.0461,
+      "step": 4344
+    },
+    {
+      "epoch": 1.3036987020781754,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001353065220665763,
+      "loss": 1.1419,
+      "step": 4345
+    },
+    {
+      "epoch": 1.3039987996098732,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013529145955716224,
+      "loss": 1.0115,
+      "step": 4346
+    },
+    {
+      "epoch": 1.304298897141571,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013527639704774818,
+      "loss": 1.0018,
+      "step": 4347
+    },
+    {
+      "epoch": 1.3045989946732688,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001352613345383341,
+      "loss": 1.195,
+      "step": 4348
+    },
+    {
+      "epoch": 1.3048990922049666,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013524627202892,
+      "loss": 1.0232,
+      "step": 4349
+    },
+    {
+      "epoch": 1.3051991897366644,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013523120951950597,
+      "loss": 0.9414,
+      "step": 4350
+    },
+    {
+      "epoch": 1.3054992872683622,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013521614701009189,
+      "loss": 1.0741,
+      "step": 4351
+    },
+    {
+      "epoch": 1.30579938480006,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001352010845006778,
+      "loss": 0.9345,
+      "step": 4352
+    },
+    {
+      "epoch": 1.3060994823317578,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013518602199126377,
+      "loss": 1.0203,
+      "step": 4353
+    },
+    {
+      "epoch": 1.3063995798634556,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013517095948184968,
+      "loss": 1.197,
+      "step": 4354
+    },
+    {
+      "epoch": 1.3066996773951534,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00013515589697243562,
+      "loss": 0.9921,
+      "step": 4355
+    },
+    {
+      "epoch": 1.3069997749268512,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013514083446302156,
+      "loss": 1.0223,
+      "step": 4356
+    },
+    {
+      "epoch": 1.307299872458549,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00013512577195360747,
+      "loss": 1.1222,
+      "step": 4357
+    },
+    {
+      "epoch": 1.307599969990247,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00013511070944419341,
+      "loss": 1.1657,
+      "step": 4358
+    },
+    {
+      "epoch": 1.3079000675219445,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00013509564693477935,
+      "loss": 1.1949,
+      "step": 4359
+    },
+    {
+      "epoch": 1.3082001650536426,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013508058442536527,
+      "loss": 1.0607,
+      "step": 4360
+    },
+    {
+      "epoch": 1.3085002625853401,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001350655219159512,
+      "loss": 1.047,
+      "step": 4361
+    },
+    {
+      "epoch": 1.3088003601170382,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00013505045940653712,
+      "loss": 0.9689,
+      "step": 4362
+    },
+    {
+      "epoch": 1.3091004576487357,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013503539689712306,
+      "loss": 1.1638,
+      "step": 4363
+    },
+    {
+      "epoch": 1.3094005551804337,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.000135020334387709,
+      "loss": 1.1378,
+      "step": 4364
+    },
+    {
+      "epoch": 1.3097006527121313,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00013500527187829491,
+      "loss": 1.0681,
+      "step": 4365
+    },
+    {
+      "epoch": 1.3100007502438293,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013499020936888088,
+      "loss": 0.9606,
+      "step": 4366
+    },
+    {
+      "epoch": 1.310300847775527,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001349751468594668,
+      "loss": 1.1223,
+      "step": 4367
+    },
+    {
+      "epoch": 1.310600945307225,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001349600843500527,
+      "loss": 1.2152,
+      "step": 4368
+    },
+    {
+      "epoch": 1.3109010428389227,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00013494502184063868,
+      "loss": 0.9654,
+      "step": 4369
+    },
+    {
+      "epoch": 1.3112011403706205,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001349299593312246,
+      "loss": 1.0574,
+      "step": 4370
+    },
+    {
+      "epoch": 1.3115012379023183,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001349148968218105,
+      "loss": 0.9554,
+      "step": 4371
+    },
+    {
+      "epoch": 1.3118013354340161,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00013489983431239644,
+      "loss": 1.1023,
+      "step": 4372
+    },
+    {
+      "epoch": 1.312101432965714,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013488477180298238,
+      "loss": 0.954,
+      "step": 4373
+    },
+    {
+      "epoch": 1.3124015304974117,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00013486970929356832,
+      "loss": 0.9818,
+      "step": 4374
+    },
+    {
+      "epoch": 1.3127016280291095,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013485464678415424,
+      "loss": 1.1043,
+      "step": 4375
+    },
+    {
+      "epoch": 1.3130017255608073,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013483958427474018,
+      "loss": 0.9958,
+      "step": 4376
+    },
+    {
+      "epoch": 1.313301823092505,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013482452176532612,
+      "loss": 1.0208,
+      "step": 4377
+    },
+    {
+      "epoch": 1.313601920624203,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00013480945925591203,
+      "loss": 1.1493,
+      "step": 4378
+    },
+    {
+      "epoch": 1.3139020181559007,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00013479439674649797,
+      "loss": 0.9543,
+      "step": 4379
+    },
+    {
+      "epoch": 1.3142021156875985,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001347793342370839,
+      "loss": 1.0931,
+      "step": 4380
+    },
+    {
+      "epoch": 1.3145022132192963,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013476427172766982,
+      "loss": 1.0752,
+      "step": 4381
+    },
+    {
+      "epoch": 1.314802310750994,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001347492092182558,
+      "loss": 0.9531,
+      "step": 4382
+    },
+    {
+      "epoch": 1.3151024082826919,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001347341467088417,
+      "loss": 0.9976,
+      "step": 4383
+    },
+    {
+      "epoch": 1.3154025058143897,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013471908419942762,
+      "loss": 1.1097,
+      "step": 4384
+    },
+    {
+      "epoch": 1.3157026033460875,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013470402169001356,
+      "loss": 1.2131,
+      "step": 4385
+    },
+    {
+      "epoch": 1.3160027008777853,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001346889591805995,
+      "loss": 1.0121,
+      "step": 4386
+    },
+    {
+      "epoch": 1.316302798409483,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001346738966711854,
+      "loss": 0.9917,
+      "step": 4387
+    },
+    {
+      "epoch": 1.3166028959411809,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013465883416177135,
+      "loss": 1.0175,
+      "step": 4388
+    },
+    {
+      "epoch": 1.3169029934728786,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001346437716523573,
+      "loss": 1.049,
+      "step": 4389
+    },
+    {
+      "epoch": 1.3172030910045764,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001346287091429432,
+      "loss": 0.9685,
+      "step": 4390
+    },
+    {
+      "epoch": 1.3175031885362742,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013461364663352915,
+      "loss": 1.0954,
+      "step": 4391
+    },
+    {
+      "epoch": 1.317803286067972,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013459858412411509,
+      "loss": 1.1732,
+      "step": 4392
+    },
+    {
+      "epoch": 1.3181033835996698,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013458352161470103,
+      "loss": 1.0414,
+      "step": 4393
+    },
+    {
+      "epoch": 1.3184034811313676,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013456845910528694,
+      "loss": 1.0199,
+      "step": 4394
+    },
+    {
+      "epoch": 1.3187035786630654,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013455339659587288,
+      "loss": 1.1306,
+      "step": 4395
+    },
+    {
+      "epoch": 1.3190036761947632,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00013453833408645882,
+      "loss": 0.8602,
+      "step": 4396
+    },
+    {
+      "epoch": 1.3193037737264612,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013452327157704473,
+      "loss": 1.0578,
+      "step": 4397
+    },
+    {
+      "epoch": 1.3196038712581588,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00013450820906763067,
+      "loss": 1.1383,
+      "step": 4398
+    },
+    {
+      "epoch": 1.3199039687898568,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013449314655821661,
+      "loss": 1.0663,
+      "step": 4399
+    },
+    {
+      "epoch": 1.3202040663215544,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013447808404880253,
+      "loss": 1.0714,
+      "step": 4400
+    },
+    {
+      "epoch": 1.3205041638532524,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013446302153938847,
+      "loss": 1.1457,
+      "step": 4401
+    },
+    {
+      "epoch": 1.32080426138495,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001344479590299744,
+      "loss": 1.1852,
+      "step": 4402
+    },
+    {
+      "epoch": 1.321104358916648,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013443289652056032,
+      "loss": 1.0798,
+      "step": 4403
+    },
+    {
+      "epoch": 1.3214044564483456,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013441783401114626,
+      "loss": 0.9883,
+      "step": 4404
+    },
+    {
+      "epoch": 1.3217045539800436,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001344027715017322,
+      "loss": 0.9951,
+      "step": 4405
+    },
+    {
+      "epoch": 1.3220046515117412,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013438770899231811,
+      "loss": 1.1201,
+      "step": 4406
+    },
+    {
+      "epoch": 1.3223047490434392,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013437264648290405,
+      "loss": 1.0766,
+      "step": 4407
+    },
+    {
+      "epoch": 1.322604846575137,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013435758397349,
+      "loss": 0.9174,
+      "step": 4408
+    },
+    {
+      "epoch": 1.3229049441068348,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013434252146407594,
+      "loss": 1.0834,
+      "step": 4409
+    },
+    {
+      "epoch": 1.3232050416385326,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00013432745895466185,
+      "loss": 1.0886,
+      "step": 4410
+    },
+    {
+      "epoch": 1.3235051391702304,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001343123964452478,
+      "loss": 1.0664,
+      "step": 4411
+    },
+    {
+      "epoch": 1.3238052367019282,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00013429733393583373,
+      "loss": 1.0453,
+      "step": 4412
+    },
+    {
+      "epoch": 1.324105334233626,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013428227142641964,
+      "loss": 1.0138,
+      "step": 4413
+    },
+    {
+      "epoch": 1.3244054317653238,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013426720891700558,
+      "loss": 1.1292,
+      "step": 4414
+    },
+    {
+      "epoch": 1.3247055292970216,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013425214640759152,
+      "loss": 1.0862,
+      "step": 4415
+    },
+    {
+      "epoch": 1.3250056268287194,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00013423708389817744,
+      "loss": 1.1122,
+      "step": 4416
+    },
+    {
+      "epoch": 1.3253057243604172,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013422202138876338,
+      "loss": 1.0381,
+      "step": 4417
+    },
+    {
+      "epoch": 1.325605821892115,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00013420695887934932,
+      "loss": 1.0053,
+      "step": 4418
+    },
+    {
+      "epoch": 1.3259059194238128,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013419189636993523,
+      "loss": 1.0176,
+      "step": 4419
+    },
+    {
+      "epoch": 1.3262060169555105,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013417683386052117,
+      "loss": 0.9117,
+      "step": 4420
+    },
+    {
+      "epoch": 1.3265061144872083,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001341617713511071,
+      "loss": 1.0731,
+      "step": 4421
+    },
+    {
+      "epoch": 1.3268062120189061,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013414670884169302,
+      "loss": 1.0625,
+      "step": 4422
+    },
+    {
+      "epoch": 1.327106309550604,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013413164633227896,
+      "loss": 1.0377,
+      "step": 4423
+    },
+    {
+      "epoch": 1.3274064070823017,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001341165838228649,
+      "loss": 1.0249,
+      "step": 4424
+    },
+    {
+      "epoch": 1.3277065046139995,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013410152131345082,
+      "loss": 1.02,
+      "step": 4425
+    },
+    {
+      "epoch": 1.3280066021456973,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013408645880403676,
+      "loss": 0.938,
+      "step": 4426
+    },
+    {
+      "epoch": 1.3283066996773951,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.0001340713962946227,
+      "loss": 1.1122,
+      "step": 4427
+    },
+    {
+      "epoch": 1.328606797209093,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013405633378520864,
+      "loss": 1.1096,
+      "step": 4428
+    },
+    {
+      "epoch": 1.3289068947407907,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013404127127579455,
+      "loss": 1.1165,
+      "step": 4429
+    },
+    {
+      "epoch": 1.3292069922724885,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001340262087663805,
+      "loss": 1.0483,
+      "step": 4430
+    },
+    {
+      "epoch": 1.3295070898041863,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013401114625696643,
+      "loss": 1.064,
+      "step": 4431
+    },
+    {
+      "epoch": 1.329807187335884,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013399608374755235,
+      "loss": 1.0466,
+      "step": 4432
+    },
+    {
+      "epoch": 1.330107284867582,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013398102123813826,
+      "loss": 1.0747,
+      "step": 4433
+    },
+    {
+      "epoch": 1.3304073823992797,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013396595872872423,
+      "loss": 1.1408,
+      "step": 4434
+    },
+    {
+      "epoch": 1.3307074799309775,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013395089621931014,
+      "loss": 0.8977,
+      "step": 4435
+    },
+    {
+      "epoch": 1.3310075774626755,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013393583370989608,
+      "loss": 1.0603,
+      "step": 4436
+    },
+    {
+      "epoch": 1.331307674994373,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00013392077120048202,
+      "loss": 0.9998,
+      "step": 4437
+    },
+    {
+      "epoch": 1.331607772526071,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013390570869106793,
+      "loss": 0.944,
+      "step": 4438
+    },
+    {
+      "epoch": 1.3319078700577687,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013389064618165387,
+      "loss": 1.1702,
+      "step": 4439
+    },
+    {
+      "epoch": 1.3322079675894667,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013387558367223981,
+      "loss": 1.093,
+      "step": 4440
+    },
+    {
+      "epoch": 1.3325080651211643,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013386052116282573,
+      "loss": 1.0192,
+      "step": 4441
+    },
+    {
+      "epoch": 1.3328081626528623,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013384545865341167,
+      "loss": 1.0663,
+      "step": 4442
+    },
+    {
+      "epoch": 1.3331082601845599,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00013383039614399758,
+      "loss": 1.2187,
+      "step": 4443
+    },
+    {
+      "epoch": 1.3334083577162579,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013381533363458352,
+      "loss": 0.947,
+      "step": 4444
+    },
+    {
+      "epoch": 1.3337084552479554,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013380027112516946,
+      "loss": 1.1034,
+      "step": 4445
+    },
+    {
+      "epoch": 1.3340085527796535,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013378520861575537,
+      "loss": 1.1,
+      "step": 4446
+    },
+    {
+      "epoch": 1.3343086503113513,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013377014610634134,
+      "loss": 1.0342,
+      "step": 4447
+    },
+    {
+      "epoch": 1.334608747843049,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00013375508359692726,
+      "loss": 0.9696,
+      "step": 4448
+    },
+    {
+      "epoch": 1.3349088453747469,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00013374002108751317,
+      "loss": 1.0003,
+      "step": 4449
+    },
+    {
+      "epoch": 1.3352089429064447,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013372495857809914,
+      "loss": 1.0192,
+      "step": 4450
+    },
+    {
+      "epoch": 1.3355090404381424,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00013370989606868505,
+      "loss": 0.9321,
+      "step": 4451
+    },
+    {
+      "epoch": 1.3358091379698402,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013369483355927096,
+      "loss": 1.067,
+      "step": 4452
+    },
+    {
+      "epoch": 1.336109235501538,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013367977104985693,
+      "loss": 1.078,
+      "step": 4453
+    },
+    {
+      "epoch": 1.3364093330332358,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013366470854044284,
+      "loss": 1.0324,
+      "step": 4454
+    },
+    {
+      "epoch": 1.3367094305649336,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013364964603102878,
+      "loss": 1.1242,
+      "step": 4455
+    },
+    {
+      "epoch": 1.3370095280966314,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001336345835216147,
+      "loss": 0.9675,
+      "step": 4456
+    },
+    {
+      "epoch": 1.3373096256283292,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013361952101220064,
+      "loss": 1.0639,
+      "step": 4457
+    },
+    {
+      "epoch": 1.337609723160027,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013360445850278658,
+      "loss": 1.0594,
+      "step": 4458
+    },
+    {
+      "epoch": 1.3379098206917248,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.0001335893959933725,
+      "loss": 0.9583,
+      "step": 4459
+    },
+    {
+      "epoch": 1.3382099182234226,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013357433348395843,
+      "loss": 1.1603,
+      "step": 4460
+    },
+    {
+      "epoch": 1.3385100157551204,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013355927097454437,
+      "loss": 0.9947,
+      "step": 4461
+    },
+    {
+      "epoch": 1.3388101132868182,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013354420846513028,
+      "loss": 1.0349,
+      "step": 4462
+    },
+    {
+      "epoch": 1.339110210818516,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00013352914595571625,
+      "loss": 1.1147,
+      "step": 4463
+    },
+    {
+      "epoch": 1.3394103083502138,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013351408344630216,
+      "loss": 1.0779,
+      "step": 4464
+    },
+    {
+      "epoch": 1.3397104058819116,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013349902093688808,
+      "loss": 1.0741,
+      "step": 4465
+    },
+    {
+      "epoch": 1.3400105034136094,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013348395842747405,
+      "loss": 0.9846,
+      "step": 4466
+    },
+    {
+      "epoch": 1.3403106009453072,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00013346889591805996,
+      "loss": 1.1124,
+      "step": 4467
+    },
+    {
+      "epoch": 1.340610698477005,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00013345383340864587,
+      "loss": 1.022,
+      "step": 4468
+    },
+    {
+      "epoch": 1.3409107960087028,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001334387708992318,
+      "loss": 1.1902,
+      "step": 4469
+    },
+    {
+      "epoch": 1.3412108935404006,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013342370838981775,
+      "loss": 1.021,
+      "step": 4470
+    },
+    {
+      "epoch": 1.3415109910720984,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013340864588040367,
+      "loss": 1.0205,
+      "step": 4471
+    },
+    {
+      "epoch": 1.3418110886037962,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001333935833709896,
+      "loss": 1.063,
+      "step": 4472
+    },
+    {
+      "epoch": 1.342111186135494,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013337852086157555,
+      "loss": 1.2062,
+      "step": 4473
+    },
+    {
+      "epoch": 1.3424112836671918,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013336345835216149,
+      "loss": 1.0432,
+      "step": 4474
+    },
+    {
+      "epoch": 1.3427113811988898,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001333483958427474,
+      "loss": 1.0454,
+      "step": 4475
+    },
+    {
+      "epoch": 1.3430114787305873,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.9856,
+      "step": 4476
+    },
+    {
+      "epoch": 1.3433115762622854,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00013331827082391928,
+      "loss": 1.0705,
+      "step": 4477
+    },
+    {
+      "epoch": 1.343611673793983,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001333032083145052,
+      "loss": 1.136,
+      "step": 4478
+    },
+    {
+      "epoch": 1.343911771325681,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013328814580509113,
+      "loss": 1.1222,
+      "step": 4479
+    },
+    {
+      "epoch": 1.3442118688573785,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013327308329567707,
+      "loss": 1.0946,
+      "step": 4480
+    },
+    {
+      "epoch": 1.3445119663890766,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.000133258020786263,
+      "loss": 1.0196,
+      "step": 4481
+    },
+    {
+      "epoch": 1.3448120639207741,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013324295827684893,
+      "loss": 1.0529,
+      "step": 4482
+    },
+    {
+      "epoch": 1.3451121614524721,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013322789576743487,
+      "loss": 1.1046,
+      "step": 4483
+    },
+    {
+      "epoch": 1.3454122589841697,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013321283325802078,
+      "loss": 0.9687,
+      "step": 4484
+    },
+    {
+      "epoch": 1.3457123565158677,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013319777074860672,
+      "loss": 1.0138,
+      "step": 4485
+    },
+    {
+      "epoch": 1.3460124540475655,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013318270823919266,
+      "loss": 1.0769,
+      "step": 4486
+    },
+    {
+      "epoch": 1.3463125515792633,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013316764572977857,
+      "loss": 1.0934,
+      "step": 4487
+    },
+    {
+      "epoch": 1.3466126491109611,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00013315258322036451,
+      "loss": 1.1739,
+      "step": 4488
+    },
+    {
+      "epoch": 1.346912746642659,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00013313752071095046,
+      "loss": 1.1123,
+      "step": 4489
+    },
+    {
+      "epoch": 1.3472128441743567,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001331224582015364,
+      "loss": 1.0409,
+      "step": 4490
+    },
+    {
+      "epoch": 1.3475129417060545,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001331073956921223,
+      "loss": 1.12,
+      "step": 4491
+    },
+    {
+      "epoch": 1.3478130392377523,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013309233318270825,
+      "loss": 1.0624,
+      "step": 4492
+    },
+    {
+      "epoch": 1.34811313676945,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001330772706732942,
+      "loss": 1.0673,
+      "step": 4493
+    },
+    {
+      "epoch": 1.348413234301148,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001330622081638801,
+      "loss": 1.1025,
+      "step": 4494
+    },
+    {
+      "epoch": 1.3487133318328457,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013304714565446604,
+      "loss": 1.0731,
+      "step": 4495
+    },
+    {
+      "epoch": 1.3490134293645435,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00013303208314505198,
+      "loss": 1.0285,
+      "step": 4496
+    },
+    {
+      "epoch": 1.3493135268962413,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001330170206356379,
+      "loss": 1.1613,
+      "step": 4497
+    },
+    {
+      "epoch": 1.349613624427939,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00013300195812622384,
+      "loss": 1.0022,
+      "step": 4498
+    },
+    {
+      "epoch": 1.3499137219596369,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013298689561680978,
+      "loss": 1.1069,
+      "step": 4499
+    },
+    {
+      "epoch": 1.3502138194913347,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001329718331073957,
+      "loss": 1.0095,
+      "step": 4500
+    },
+    {
+      "epoch": 1.3505139170230325,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013295677059798163,
+      "loss": 1.082,
+      "step": 4501
+    },
+    {
+      "epoch": 1.3508140145547303,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00013294170808856757,
+      "loss": 0.9633,
+      "step": 4502
+    },
+    {
+      "epoch": 1.351114112086428,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013292664557915348,
+      "loss": 1.0861,
+      "step": 4503
+    },
+    {
+      "epoch": 1.3514142096181259,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013291158306973942,
+      "loss": 1.111,
+      "step": 4504
+    },
+    {
+      "epoch": 1.3517143071498237,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013289652056032536,
+      "loss": 0.9767,
+      "step": 4505
+    },
+    {
+      "epoch": 1.3520144046815215,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013288145805091128,
+      "loss": 1.136,
+      "step": 4506
+    },
+    {
+      "epoch": 1.3523145022132192,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013286639554149722,
+      "loss": 1.2192,
+      "step": 4507
+    },
+    {
+      "epoch": 1.352614599744917,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013285133303208316,
+      "loss": 1.0533,
+      "step": 4508
+    },
+    {
+      "epoch": 1.3529146972766148,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001328362705226691,
+      "loss": 1.1523,
+      "step": 4509
+    },
+    {
+      "epoch": 1.3532147948083126,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.000132821208013255,
+      "loss": 0.9285,
+      "step": 4510
+    },
+    {
+      "epoch": 1.3535148923400104,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013280614550384095,
+      "loss": 0.9292,
+      "step": 4511
+    },
+    {
+      "epoch": 1.3538149898717082,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001327910829944269,
+      "loss": 1.0148,
+      "step": 4512
+    },
+    {
+      "epoch": 1.354115087403406,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001327760204850128,
+      "loss": 1.0713,
+      "step": 4513
+    },
+    {
+      "epoch": 1.354415184935104,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013276095797559872,
+      "loss": 1.1863,
+      "step": 4514
+    },
+    {
+      "epoch": 1.3547152824668016,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013274589546618469,
+      "loss": 1.0024,
+      "step": 4515
+    },
+    {
+      "epoch": 1.3550153799984996,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001327308329567706,
+      "loss": 1.0782,
+      "step": 4516
+    },
+    {
+      "epoch": 1.3553154775301972,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00013271577044735654,
+      "loss": 1.1514,
+      "step": 4517
+    },
+    {
+      "epoch": 1.3556155750618952,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00013270070793794248,
+      "loss": 1.132,
+      "step": 4518
+    },
+    {
+      "epoch": 1.3559156725935928,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001326856454285284,
+      "loss": 1.3489,
+      "step": 4519
+    },
+    {
+      "epoch": 1.3562157701252908,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013267058291911433,
+      "loss": 1.0391,
+      "step": 4520
+    },
+    {
+      "epoch": 1.3565158676569884,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013265552040970027,
+      "loss": 1.222,
+      "step": 4521
+    },
+    {
+      "epoch": 1.3568159651886864,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001326404579002862,
+      "loss": 1.0699,
+      "step": 4522
+    },
+    {
+      "epoch": 1.357116062720384,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013262539539087213,
+      "loss": 1.0708,
+      "step": 4523
+    },
+    {
+      "epoch": 1.357416160252082,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00013261033288145807,
+      "loss": 1.072,
+      "step": 4524
+    },
+    {
+      "epoch": 1.3577162577837798,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00013259527037204398,
+      "loss": 1.0238,
+      "step": 4525
+    },
+    {
+      "epoch": 1.3580163553154776,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013258020786262992,
+      "loss": 1.0147,
+      "step": 4526
+    },
+    {
+      "epoch": 1.3583164528471754,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00013256514535321583,
+      "loss": 1.1491,
+      "step": 4527
+    },
+    {
+      "epoch": 1.3586165503788732,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001325500828438018,
+      "loss": 1.0011,
+      "step": 4528
+    },
+    {
+      "epoch": 1.358916647910571,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013253502033438772,
+      "loss": 1.0,
+      "step": 4529
+    },
+    {
+      "epoch": 1.3592167454422688,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013251995782497363,
+      "loss": 1.1299,
+      "step": 4530
+    },
+    {
+      "epoch": 1.3595168429739666,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001325048953155596,
+      "loss": 1.1441,
+      "step": 4531
+    },
+    {
+      "epoch": 1.3598169405056644,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001324898328061455,
+      "loss": 1.15,
+      "step": 4532
+    },
+    {
+      "epoch": 1.3601170380373622,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013247477029673142,
+      "loss": 1.0727,
+      "step": 4533
+    },
+    {
+      "epoch": 1.36041713556906,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001324597077873174,
+      "loss": 0.9612,
+      "step": 4534
+    },
+    {
+      "epoch": 1.3607172331007578,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001324446452779033,
+      "loss": 1.1174,
+      "step": 4535
+    },
+    {
+      "epoch": 1.3610173306324556,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00013242958276848924,
+      "loss": 1.1498,
+      "step": 4536
+    },
+    {
+      "epoch": 1.3613174281641534,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013241452025907518,
+      "loss": 0.9752,
+      "step": 4537
+    },
+    {
+      "epoch": 1.3616175256958511,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001323994577496611,
+      "loss": 1.0656,
+      "step": 4538
+    },
+    {
+      "epoch": 1.361917623227549,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013238439524024704,
+      "loss": 1.0419,
+      "step": 4539
+    },
+    {
+      "epoch": 1.3622177207592467,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013236933273083295,
+      "loss": 1.1516,
+      "step": 4540
+    },
+    {
+      "epoch": 1.3625178182909445,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001323542702214189,
+      "loss": 1.1309,
+      "step": 4541
+    },
+    {
+      "epoch": 1.3628179158226423,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013233920771200483,
+      "loss": 1.0599,
+      "step": 4542
+    },
+    {
+      "epoch": 1.3631180133543401,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013232414520259074,
+      "loss": 1.0069,
+      "step": 4543
+    },
+    {
+      "epoch": 1.363418110886038,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001323090826931767,
+      "loss": 1.0817,
+      "step": 4544
+    },
+    {
+      "epoch": 1.3637182084177357,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013229402018376262,
+      "loss": 1.0477,
+      "step": 4545
+    },
+    {
+      "epoch": 1.3640183059494335,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013227895767434854,
+      "loss": 0.9822,
+      "step": 4546
+    },
+    {
+      "epoch": 1.3643184034811313,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001322638951649345,
+      "loss": 1.0979,
+      "step": 4547
+    },
+    {
+      "epoch": 1.364618501012829,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013224883265552042,
+      "loss": 1.1834,
+      "step": 4548
+    },
+    {
+      "epoch": 1.364918598544527,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013223377014610633,
+      "loss": 1.127,
+      "step": 4549
+    },
+    {
+      "epoch": 1.3652186960762247,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013221870763669227,
+      "loss": 1.0754,
+      "step": 4550
+    },
+    {
+      "epoch": 1.3655187936079225,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001322036451272782,
+      "loss": 1.0019,
+      "step": 4551
+    },
+    {
+      "epoch": 1.3658188911396203,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013218858261786413,
+      "loss": 1.0058,
+      "step": 4552
+    },
+    {
+      "epoch": 1.3661189886713183,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013217352010845007,
+      "loss": 1.0963,
+      "step": 4553
+    },
+    {
+      "epoch": 1.3664190862030159,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.000132158457599036,
+      "loss": 1.0504,
+      "step": 4554
+    },
+    {
+      "epoch": 1.366719183734714,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013214339508962195,
+      "loss": 1.0595,
+      "step": 4555
+    },
+    {
+      "epoch": 1.3670192812664115,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00013212833258020786,
+      "loss": 1.0049,
+      "step": 4556
+    },
+    {
+      "epoch": 1.3673193787981095,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001321132700707938,
+      "loss": 1.0706,
+      "step": 4557
+    },
+    {
+      "epoch": 1.367619476329807,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00013209820756137974,
+      "loss": 1.1316,
+      "step": 4558
+    },
+    {
+      "epoch": 1.367919573861505,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013208314505196565,
+      "loss": 1.0071,
+      "step": 4559
+    },
+    {
+      "epoch": 1.3682196713932027,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001320680825425516,
+      "loss": 1.0266,
+      "step": 4560
+    },
+    {
+      "epoch": 1.3685197689249007,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013205302003313753,
+      "loss": 1.0441,
+      "step": 4561
+    },
+    {
+      "epoch": 1.3688198664565985,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013203795752372345,
+      "loss": 0.9665,
+      "step": 4562
+    },
+    {
+      "epoch": 1.3691199639882963,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001320228950143094,
+      "loss": 1.0459,
+      "step": 4563
+    },
+    {
+      "epoch": 1.369420061519994,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00013200783250489533,
+      "loss": 0.9548,
+      "step": 4564
+    },
+    {
+      "epoch": 1.3697201590516919,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013199276999548124,
+      "loss": 1.0351,
+      "step": 4565
+    },
+    {
+      "epoch": 1.3700202565833897,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013197770748606718,
+      "loss": 0.9204,
+      "step": 4566
+    },
+    {
+      "epoch": 1.3703203541150875,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013196264497665312,
+      "loss": 1.0772,
+      "step": 4567
+    },
+    {
+      "epoch": 1.3706204516467853,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013194758246723903,
+      "loss": 1.0786,
+      "step": 4568
+    },
+    {
+      "epoch": 1.370920549178483,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013193251995782497,
+      "loss": 1.1091,
+      "step": 4569
+    },
+    {
+      "epoch": 1.3712206467101808,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013191745744841092,
+      "loss": 0.9929,
+      "step": 4570
+    },
+    {
+      "epoch": 1.3715207442418786,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013190239493899686,
+      "loss": 1.0353,
+      "step": 4571
+    },
+    {
+      "epoch": 1.3718208417735764,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013188733242958277,
+      "loss": 1.0812,
+      "step": 4572
+    },
+    {
+      "epoch": 1.3721209393052742,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001318722699201687,
+      "loss": 1.0319,
+      "step": 4573
+    },
+    {
+      "epoch": 1.372421036836972,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00013185720741075465,
+      "loss": 0.9107,
+      "step": 4574
+    },
+    {
+      "epoch": 1.3727211343686698,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013184214490134056,
+      "loss": 1.1097,
+      "step": 4575
+    },
+    {
+      "epoch": 1.3730212319003676,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001318270823919265,
+      "loss": 1.0779,
+      "step": 4576
+    },
+    {
+      "epoch": 1.3733213294320654,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00013181201988251244,
+      "loss": 0.914,
+      "step": 4577
+    },
+    {
+      "epoch": 1.3736214269637632,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013179695737309836,
+      "loss": 1.0447,
+      "step": 4578
+    },
+    {
+      "epoch": 1.373921524495461,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001317818948636843,
+      "loss": 1.1249,
+      "step": 4579
+    },
+    {
+      "epoch": 1.3742216220271588,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013176683235427024,
+      "loss": 1.0045,
+      "step": 4580
+    },
+    {
+      "epoch": 1.3745217195588566,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013175176984485615,
+      "loss": 1.0123,
+      "step": 4581
+    },
+    {
+      "epoch": 1.3748218170905544,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001317367073354421,
+      "loss": 1.0604,
+      "step": 4582
+    },
+    {
+      "epoch": 1.3751219146222522,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00013172164482602803,
+      "loss": 1.182,
+      "step": 4583
+    },
+    {
+      "epoch": 1.37542201215395,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013170658231661394,
+      "loss": 1.0501,
+      "step": 4584
+    },
+    {
+      "epoch": 1.3757221096856478,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00013169151980719988,
+      "loss": 1.0724,
+      "step": 4585
+    },
+    {
+      "epoch": 1.3760222072173456,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00013167645729778582,
+      "loss": 1.1365,
+      "step": 4586
+    },
+    {
+      "epoch": 1.3763223047490434,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013166139478837174,
+      "loss": 1.1214,
+      "step": 4587
+    },
+    {
+      "epoch": 1.3766224022807412,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00013164633227895768,
+      "loss": 1.1982,
+      "step": 4588
+    },
+    {
+      "epoch": 1.376922499812439,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013163126976954362,
+      "loss": 1.1243,
+      "step": 4589
+    },
+    {
+      "epoch": 1.3772225973441368,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00013161620726012956,
+      "loss": 1.2044,
+      "step": 4590
+    },
+    {
+      "epoch": 1.3775226948758346,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013160114475071547,
+      "loss": 1.0264,
+      "step": 4591
+    },
+    {
+      "epoch": 1.3778227924075326,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001315860822413014,
+      "loss": 1.1803,
+      "step": 4592
+    },
+    {
+      "epoch": 1.3781228899392302,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013157101973188735,
+      "loss": 0.9627,
+      "step": 4593
+    },
+    {
+      "epoch": 1.3784229874709282,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013155595722247327,
+      "loss": 0.9777,
+      "step": 4594
+    },
+    {
+      "epoch": 1.3787230850026257,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001315408947130592,
+      "loss": 1.0108,
+      "step": 4595
+    },
+    {
+      "epoch": 1.3790231825343238,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00013152583220364515,
+      "loss": 1.0917,
+      "step": 4596
+    },
+    {
+      "epoch": 1.3793232800660213,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00013151076969423106,
+      "loss": 0.9837,
+      "step": 4597
+    },
+    {
+      "epoch": 1.3796233775977194,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.000131495707184817,
+      "loss": 1.0276,
+      "step": 4598
+    },
+    {
+      "epoch": 1.379923475129417,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013148064467540294,
+      "loss": 0.9553,
+      "step": 4599
+    },
+    {
+      "epoch": 1.380223572661115,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00013146558216598885,
+      "loss": 1.1035,
+      "step": 4600
+    },
+    {
+      "epoch": 1.3805236701928127,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001314505196565748,
+      "loss": 1.0622,
+      "step": 4601
+    },
+    {
+      "epoch": 1.3808237677245105,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013143545714716073,
+      "loss": 1.121,
+      "step": 4602
+    },
+    {
+      "epoch": 1.3811238652562083,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013142039463774665,
+      "loss": 1.0112,
+      "step": 4603
+    },
+    {
+      "epoch": 1.3814239627879061,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001314053321283326,
+      "loss": 1.1097,
+      "step": 4604
+    },
+    {
+      "epoch": 1.381724060319604,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013139026961891853,
+      "loss": 1.158,
+      "step": 4605
+    },
+    {
+      "epoch": 1.3820241578513017,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013137520710950444,
+      "loss": 0.9527,
+      "step": 4606
+    },
+    {
+      "epoch": 1.3823242553829995,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00013136014460009038,
+      "loss": 0.9791,
+      "step": 4607
+    },
+    {
+      "epoch": 1.3826243529146973,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001313450820906763,
+      "loss": 1.1968,
+      "step": 4608
+    },
+    {
+      "epoch": 1.3829244504463951,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013133001958126226,
+      "loss": 0.934,
+      "step": 4609
+    },
+    {
+      "epoch": 1.383224547978093,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00013131495707184818,
+      "loss": 1.0723,
+      "step": 4610
+    },
+    {
+      "epoch": 1.3835246455097907,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001312998945624341,
+      "loss": 1.1689,
+      "step": 4611
+    },
+    {
+      "epoch": 1.3838247430414885,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013128483205302006,
+      "loss": 0.8383,
+      "step": 4612
+    },
+    {
+      "epoch": 1.3841248405731863,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013126976954360597,
+      "loss": 1.2099,
+      "step": 4613
+    },
+    {
+      "epoch": 1.384424938104884,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013125470703419188,
+      "loss": 1.0681,
+      "step": 4614
+    },
+    {
+      "epoch": 1.384725035636582,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013123964452477785,
+      "loss": 1.2548,
+      "step": 4615
+    },
+    {
+      "epoch": 1.3850251331682797,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013122458201536376,
+      "loss": 1.1083,
+      "step": 4616
+    },
+    {
+      "epoch": 1.3853252306999775,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001312095195059497,
+      "loss": 0.9583,
+      "step": 4617
+    },
+    {
+      "epoch": 1.3856253282316753,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013119445699653564,
+      "loss": 1.1243,
+      "step": 4618
+    },
+    {
+      "epoch": 1.385925425763373,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013117939448712156,
+      "loss": 1.0585,
+      "step": 4619
+    },
+    {
+      "epoch": 1.3862255232950709,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001311643319777075,
+      "loss": 1.0571,
+      "step": 4620
+    },
+    {
+      "epoch": 1.3865256208267687,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001311492694682934,
+      "loss": 0.9754,
+      "step": 4621
+    },
+    {
+      "epoch": 1.3868257183584665,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00013113420695887935,
+      "loss": 0.9878,
+      "step": 4622
+    },
+    {
+      "epoch": 1.3871258158901643,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001311191444494653,
+      "loss": 1.0213,
+      "step": 4623
+    },
+    {
+      "epoch": 1.387425913421862,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001311040819400512,
+      "loss": 1.0848,
+      "step": 4624
+    },
+    {
+      "epoch": 1.3877260109535599,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00013108901943063717,
+      "loss": 1.0824,
+      "step": 4625
+    },
+    {
+      "epoch": 1.3880261084852576,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00013107395692122308,
+      "loss": 1.0377,
+      "step": 4626
+    },
+    {
+      "epoch": 1.3883262060169554,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.000131058894411809,
+      "loss": 1.1167,
+      "step": 4627
+    },
+    {
+      "epoch": 1.3886263035486532,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00013104383190239497,
+      "loss": 0.9588,
+      "step": 4628
+    },
+    {
+      "epoch": 1.388926401080351,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013102876939298088,
+      "loss": 1.0085,
+      "step": 4629
+    },
+    {
+      "epoch": 1.3892264986120488,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001310137068835668,
+      "loss": 1.1343,
+      "step": 4630
+    },
+    {
+      "epoch": 1.3895265961437469,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013099864437415276,
+      "loss": 1.0506,
+      "step": 4631
+    },
+    {
+      "epoch": 1.3898266936754444,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00013098358186473867,
+      "loss": 1.0486,
+      "step": 4632
+    },
+    {
+      "epoch": 1.3901267912071424,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00013096851935532459,
+      "loss": 1.1116,
+      "step": 4633
+    },
+    {
+      "epoch": 1.39042688873884,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00013095345684591053,
+      "loss": 1.1291,
+      "step": 4634
+    },
+    {
+      "epoch": 1.390726986270538,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013093839433649647,
+      "loss": 1.0818,
+      "step": 4635
+    },
+    {
+      "epoch": 1.3910270838022356,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001309233318270824,
+      "loss": 0.9634,
+      "step": 4636
+    },
+    {
+      "epoch": 1.3913271813339336,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013090826931766832,
+      "loss": 1.0393,
+      "step": 4637
+    },
+    {
+      "epoch": 1.3916272788656312,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00013089320680825426,
+      "loss": 1.0675,
+      "step": 4638
+    },
+    {
+      "epoch": 1.3919273763973292,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001308781442988402,
+      "loss": 1.1023,
+      "step": 4639
+    },
+    {
+      "epoch": 1.392227473929027,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001308630817894261,
+      "loss": 0.9321,
+      "step": 4640
+    },
+    {
+      "epoch": 1.3925275714607248,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013084801928001205,
+      "loss": 1.0013,
+      "step": 4641
+    },
+    {
+      "epoch": 1.3928276689924226,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.000130832956770598,
+      "loss": 1.2212,
+      "step": 4642
+    },
+    {
+      "epoch": 1.3931277665241204,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001308178942611839,
+      "loss": 0.9246,
+      "step": 4643
+    },
+    {
+      "epoch": 1.3934278640558182,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00013080283175176987,
+      "loss": 0.937,
+      "step": 4644
+    },
+    {
+      "epoch": 1.393727961587516,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001307877692423558,
+      "loss": 1.0144,
+      "step": 4645
+    },
+    {
+      "epoch": 1.3940280591192138,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001307727067329417,
+      "loss": 1.1508,
+      "step": 4646
+    },
+    {
+      "epoch": 1.3943281566509116,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013075764422352764,
+      "loss": 1.0881,
+      "step": 4647
+    },
+    {
+      "epoch": 1.3946282541826094,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013074258171411358,
+      "loss": 0.9665,
+      "step": 4648
+    },
+    {
+      "epoch": 1.3949283517143072,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001307275192046995,
+      "loss": 1.023,
+      "step": 4649
+    },
+    {
+      "epoch": 1.395228449246005,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013071245669528543,
+      "loss": 0.979,
+      "step": 4650
+    },
+    {
+      "epoch": 1.3955285467777028,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013069739418587138,
+      "loss": 1.1186,
+      "step": 4651
+    },
+    {
+      "epoch": 1.3958286443094006,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00013068233167645732,
+      "loss": 0.922,
+      "step": 4652
+    },
+    {
+      "epoch": 1.3961287418410984,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013066726916704323,
+      "loss": 0.976,
+      "step": 4653
+    },
+    {
+      "epoch": 1.3964288393727962,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00013065220665762917,
+      "loss": 0.9391,
+      "step": 4654
+    },
+    {
+      "epoch": 1.396728936904494,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001306371441482151,
+      "loss": 1.1129,
+      "step": 4655
+    },
+    {
+      "epoch": 1.3970290344361918,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00013062208163880102,
+      "loss": 0.9702,
+      "step": 4656
+    },
+    {
+      "epoch": 1.3973291319678895,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00013060701912938696,
+      "loss": 1.0597,
+      "step": 4657
+    },
+    {
+      "epoch": 1.3976292294995873,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001305919566199729,
+      "loss": 1.1618,
+      "step": 4658
+    },
+    {
+      "epoch": 1.3979293270312851,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013057689411055882,
+      "loss": 1.1795,
+      "step": 4659
+    },
+    {
+      "epoch": 1.398229424562983,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013056183160114476,
+      "loss": 1.1573,
+      "step": 4660
+    },
+    {
+      "epoch": 1.3985295220946807,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001305467690917307,
+      "loss": 1.0102,
+      "step": 4661
+    },
+    {
+      "epoch": 1.3988296196263785,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001305317065823166,
+      "loss": 1.0814,
+      "step": 4662
+    },
+    {
+      "epoch": 1.3991297171580763,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013051664407290255,
+      "loss": 1.0811,
+      "step": 4663
+    },
+    {
+      "epoch": 1.3994298146897741,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001305015815634885,
+      "loss": 1.0231,
+      "step": 4664
+    },
+    {
+      "epoch": 1.399729912221472,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001304865190540744,
+      "loss": 0.9982,
+      "step": 4665
+    },
+    {
+      "epoch": 1.4000300097531697,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00013047145654466034,
+      "loss": 1.2512,
+      "step": 4666
+    },
+    {
+      "epoch": 1.4003301072848675,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00013045639403524628,
+      "loss": 0.9481,
+      "step": 4667
+    },
+    {
+      "epoch": 1.4006302048165653,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001304413315258322,
+      "loss": 1.0033,
+      "step": 4668
+    },
+    {
+      "epoch": 1.400930302348263,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013042626901641814,
+      "loss": 1.0136,
+      "step": 4669
+    },
+    {
+      "epoch": 1.4012303998799611,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013041120650700408,
+      "loss": 1.0154,
+      "step": 4670
+    },
+    {
+      "epoch": 1.4015304974116587,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013039614399759002,
+      "loss": 1.1146,
+      "step": 4671
+    },
+    {
+      "epoch": 1.4018305949433567,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00013038108148817593,
+      "loss": 1.1219,
+      "step": 4672
+    },
+    {
+      "epoch": 1.4021306924750543,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00013036601897876187,
+      "loss": 0.9844,
+      "step": 4673
+    },
+    {
+      "epoch": 1.4024307900067523,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001303509564693478,
+      "loss": 1.091,
+      "step": 4674
+    },
+    {
+      "epoch": 1.4027308875384499,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00013033589395993373,
+      "loss": 1.0001,
+      "step": 4675
+    },
+    {
+      "epoch": 1.403030985070148,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00013032083145051967,
+      "loss": 1.0177,
+      "step": 4676
+    },
+    {
+      "epoch": 1.4033310826018455,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001303057689411056,
+      "loss": 1.0401,
+      "step": 4677
+    },
+    {
+      "epoch": 1.4036311801335435,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013029070643169152,
+      "loss": 1.1725,
+      "step": 4678
+    },
+    {
+      "epoch": 1.4039312776652413,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00013027564392227746,
+      "loss": 0.9513,
+      "step": 4679
+    },
+    {
+      "epoch": 1.404231375196939,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001302605814128634,
+      "loss": 1.1868,
+      "step": 4680
+    },
+    {
+      "epoch": 1.4045314727286369,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001302455189034493,
+      "loss": 0.9697,
+      "step": 4681
+    },
+    {
+      "epoch": 1.4048315702603347,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013023045639403525,
+      "loss": 1.0155,
+      "step": 4682
+    },
+    {
+      "epoch": 1.4051316677920325,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001302153938846212,
+      "loss": 1.1275,
+      "step": 4683
+    },
+    {
+      "epoch": 1.4054317653237303,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001302003313752071,
+      "loss": 1.0793,
+      "step": 4684
+    },
+    {
+      "epoch": 1.405731862855428,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00013018526886579305,
+      "loss": 1.0792,
+      "step": 4685
+    },
+    {
+      "epoch": 1.4060319603871259,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.000130170206356379,
+      "loss": 1.0047,
+      "step": 4686
+    },
+    {
+      "epoch": 1.4063320579188237,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001301551438469649,
+      "loss": 1.016,
+      "step": 4687
+    },
+    {
+      "epoch": 1.4066321554505214,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00013014008133755084,
+      "loss": 0.96,
+      "step": 4688
+    },
+    {
+      "epoch": 1.4069322529822192,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00013012501882813678,
+      "loss": 0.9951,
+      "step": 4689
+    },
+    {
+      "epoch": 1.407232350513917,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00013010995631872272,
+      "loss": 1.0692,
+      "step": 4690
+    },
+    {
+      "epoch": 1.4075324480456148,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00013009489380930864,
+      "loss": 1.1511,
+      "step": 4691
+    },
+    {
+      "epoch": 1.4078325455773126,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00013007983129989455,
+      "loss": 1.0671,
+      "step": 4692
+    },
+    {
+      "epoch": 1.4081326431090104,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00013006476879048052,
+      "loss": 1.061,
+      "step": 4693
+    },
+    {
+      "epoch": 1.4084327406407082,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00013004970628106643,
+      "loss": 0.9427,
+      "step": 4694
+    },
+    {
+      "epoch": 1.408732838172406,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00013003464377165234,
+      "loss": 1.0135,
+      "step": 4695
+    },
+    {
+      "epoch": 1.4090329357041038,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001300195812622383,
+      "loss": 1.2837,
+      "step": 4696
+    },
+    {
+      "epoch": 1.4093330332358016,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00013000451875282422,
+      "loss": 1.147,
+      "step": 4697
+    },
+    {
+      "epoch": 1.4096331307674994,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012998945624341016,
+      "loss": 1.1334,
+      "step": 4698
+    },
+    {
+      "epoch": 1.4099332282991972,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001299743937339961,
+      "loss": 1.1499,
+      "step": 4699
+    },
+    {
+      "epoch": 1.410233325830895,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012995933122458202,
+      "loss": 1.0929,
+      "step": 4700
+    },
+    {
+      "epoch": 1.4105334233625928,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00012994426871516796,
+      "loss": 1.0493,
+      "step": 4701
+    },
+    {
+      "epoch": 1.4108335208942906,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001299292062057539,
+      "loss": 1.0895,
+      "step": 4702
+    },
+    {
+      "epoch": 1.4111336184259884,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001299141436963398,
+      "loss": 1.0328,
+      "step": 4703
+    },
+    {
+      "epoch": 1.4114337159576862,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012989908118692575,
+      "loss": 1.2135,
+      "step": 4704
+    },
+    {
+      "epoch": 1.411733813489384,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012988401867751166,
+      "loss": 1.2431,
+      "step": 4705
+    },
+    {
+      "epoch": 1.4120339110210818,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00012986895616809763,
+      "loss": 1.1143,
+      "step": 4706
+    },
+    {
+      "epoch": 1.4123340085527796,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012985389365868354,
+      "loss": 0.9382,
+      "step": 4707
+    },
+    {
+      "epoch": 1.4126341060844774,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012983883114926946,
+      "loss": 1.1121,
+      "step": 4708
+    },
+    {
+      "epoch": 1.4129342036161754,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012982376863985543,
+      "loss": 1.0857,
+      "step": 4709
+    },
+    {
+      "epoch": 1.413234301147873,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012980870613044134,
+      "loss": 1.2003,
+      "step": 4710
+    },
+    {
+      "epoch": 1.413534398679571,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00012979364362102725,
+      "loss": 0.896,
+      "step": 4711
+    },
+    {
+      "epoch": 1.4138344962112686,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012977858111161322,
+      "loss": 1.0786,
+      "step": 4712
+    },
+    {
+      "epoch": 1.4141345937429666,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00012976351860219913,
+      "loss": 1.0872,
+      "step": 4713
+    },
+    {
+      "epoch": 1.4144346912746641,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00012974845609278505,
+      "loss": 1.0185,
+      "step": 4714
+    },
+    {
+      "epoch": 1.4147347888063622,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012973339358337099,
+      "loss": 0.8887,
+      "step": 4715
+    },
+    {
+      "epoch": 1.4150348863380597,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00012971833107395693,
+      "loss": 1.0276,
+      "step": 4716
+    },
+    {
+      "epoch": 1.4153349838697578,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012970326856454287,
+      "loss": 1.1015,
+      "step": 4717
+    },
+    {
+      "epoch": 1.4156350814014556,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012968820605512878,
+      "loss": 0.9451,
+      "step": 4718
+    },
+    {
+      "epoch": 1.4159351789331533,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00012967314354571472,
+      "loss": 1.1249,
+      "step": 4719
+    },
+    {
+      "epoch": 1.4162352764648511,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012965808103630066,
+      "loss": 0.957,
+      "step": 4720
+    },
+    {
+      "epoch": 1.416535373996549,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012964301852688657,
+      "loss": 0.8764,
+      "step": 4721
+    },
+    {
+      "epoch": 1.4168354715282467,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012962795601747251,
+      "loss": 1.0829,
+      "step": 4722
+    },
+    {
+      "epoch": 1.4171355690599445,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012961289350805845,
+      "loss": 0.9838,
+      "step": 4723
+    },
+    {
+      "epoch": 1.4174356665916423,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00012959783099864437,
+      "loss": 1.0646,
+      "step": 4724
+    },
+    {
+      "epoch": 1.4177357641233401,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00012958276848923033,
+      "loss": 1.0027,
+      "step": 4725
+    },
+    {
+      "epoch": 1.418035861655038,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012956770597981625,
+      "loss": 0.8898,
+      "step": 4726
+    },
+    {
+      "epoch": 1.4183359591867357,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00012955264347040216,
+      "loss": 1.0632,
+      "step": 4727
+    },
+    {
+      "epoch": 1.4186360567184335,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001295375809609881,
+      "loss": 1.1825,
+      "step": 4728
+    },
+    {
+      "epoch": 1.4189361542501313,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012952251845157404,
+      "loss": 1.1143,
+      "step": 4729
+    },
+    {
+      "epoch": 1.419236251781829,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012950745594215995,
+      "loss": 1.1205,
+      "step": 4730
+    },
+    {
+      "epoch": 1.419536349313527,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001294923934327459,
+      "loss": 1.1704,
+      "step": 4731
+    },
+    {
+      "epoch": 1.4198364468452247,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012947733092333184,
+      "loss": 1.1923,
+      "step": 4732
+    },
+    {
+      "epoch": 1.4201365443769225,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012946226841391778,
+      "loss": 1.2363,
+      "step": 4733
+    },
+    {
+      "epoch": 1.4204366419086203,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001294472059045037,
+      "loss": 1.0229,
+      "step": 4734
+    },
+    {
+      "epoch": 1.420736739440318,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012943214339508963,
+      "loss": 1.1452,
+      "step": 4735
+    },
+    {
+      "epoch": 1.4210368369720159,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012941708088567557,
+      "loss": 0.992,
+      "step": 4736
+    },
+    {
+      "epoch": 1.4213369345037137,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012940201837626148,
+      "loss": 1.1299,
+      "step": 4737
+    },
+    {
+      "epoch": 1.4216370320354115,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012938695586684742,
+      "loss": 1.0242,
+      "step": 4738
+    },
+    {
+      "epoch": 1.4219371295671093,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012937189335743336,
+      "loss": 1.0758,
+      "step": 4739
+    },
+    {
+      "epoch": 1.422237227098807,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012935683084801928,
+      "loss": 1.0757,
+      "step": 4740
+    },
+    {
+      "epoch": 1.4225373246305049,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012934176833860522,
+      "loss": 0.9545,
+      "step": 4741
+    },
+    {
+      "epoch": 1.4228374221622027,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012932670582919116,
+      "loss": 1.1231,
+      "step": 4742
+    },
+    {
+      "epoch": 1.4231375196939005,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00012931164331977707,
+      "loss": 1.2553,
+      "step": 4743
+    },
+    {
+      "epoch": 1.4234376172255983,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.000129296580810363,
+      "loss": 0.9276,
+      "step": 4744
+    },
+    {
+      "epoch": 1.423737714757296,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012928151830094895,
+      "loss": 0.9672,
+      "step": 4745
+    },
+    {
+      "epoch": 1.4240378122889938,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012926645579153486,
+      "loss": 1.0394,
+      "step": 4746
+    },
+    {
+      "epoch": 1.4243379098206916,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001292513932821208,
+      "loss": 1.0387,
+      "step": 4747
+    },
+    {
+      "epoch": 1.4246380073523897,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012923633077270674,
+      "loss": 0.9782,
+      "step": 4748
+    },
+    {
+      "epoch": 1.4249381048840872,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012922126826329266,
+      "loss": 1.0761,
+      "step": 4749
+    },
+    {
+      "epoch": 1.4252382024157852,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001292062057538786,
+      "loss": 0.9548,
+      "step": 4750
+    },
+    {
+      "epoch": 1.4255382999474828,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012919114324446454,
+      "loss": 1.1849,
+      "step": 4751
+    },
+    {
+      "epoch": 1.4258383974791808,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00012917608073505048,
+      "loss": 1.0897,
+      "step": 4752
+    },
+    {
+      "epoch": 1.4261384950108784,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001291610182256364,
+      "loss": 1.0804,
+      "step": 4753
+    },
+    {
+      "epoch": 1.4264385925425764,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012914595571622233,
+      "loss": 0.9889,
+      "step": 4754
+    },
+    {
+      "epoch": 1.426738690074274,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012913089320680827,
+      "loss": 1.1448,
+      "step": 4755
+    },
+    {
+      "epoch": 1.427038787605972,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012911583069739419,
+      "loss": 0.995,
+      "step": 4756
+    },
+    {
+      "epoch": 1.4273388851376698,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012910076818798013,
+      "loss": 1.0583,
+      "step": 4757
+    },
+    {
+      "epoch": 1.4276389826693676,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012908570567856607,
+      "loss": 1.1186,
+      "step": 4758
+    },
+    {
+      "epoch": 1.4279390802010654,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012907064316915198,
+      "loss": 1.0702,
+      "step": 4759
+    },
+    {
+      "epoch": 1.4282391777327632,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012905558065973792,
+      "loss": 1.1475,
+      "step": 4760
+    },
+    {
+      "epoch": 1.428539275264461,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012904051815032386,
+      "loss": 1.0882,
+      "step": 4761
+    },
+    {
+      "epoch": 1.4288393727961588,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00012902545564090977,
+      "loss": 1.0338,
+      "step": 4762
+    },
+    {
+      "epoch": 1.4291394703278566,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012901039313149571,
+      "loss": 0.9604,
+      "step": 4763
+    },
+    {
+      "epoch": 1.4294395678595544,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00012899533062208165,
+      "loss": 0.9225,
+      "step": 4764
+    },
+    {
+      "epoch": 1.4297396653912522,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012898026811266757,
+      "loss": 1.1883,
+      "step": 4765
+    },
+    {
+      "epoch": 1.43003976292295,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001289652056032535,
+      "loss": 1.0559,
+      "step": 4766
+    },
+    {
+      "epoch": 1.4303398604546478,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00012895014309383945,
+      "loss": 0.9145,
+      "step": 4767
+    },
+    {
+      "epoch": 1.4306399579863456,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012893508058442536,
+      "loss": 1.102,
+      "step": 4768
+    },
+    {
+      "epoch": 1.4309400555180434,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001289200180750113,
+      "loss": 1.0958,
+      "step": 4769
+    },
+    {
+      "epoch": 1.4312401530497412,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012890495556559724,
+      "loss": 1.0507,
+      "step": 4770
+    },
+    {
+      "epoch": 1.431540250581439,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012888989305618318,
+      "loss": 1.0072,
+      "step": 4771
+    },
+    {
+      "epoch": 1.4318403481131368,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001288748305467691,
+      "loss": 0.9899,
+      "step": 4772
+    },
+    {
+      "epoch": 1.4321404456448346,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012885976803735504,
+      "loss": 1.0703,
+      "step": 4773
+    },
+    {
+      "epoch": 1.4324405431765324,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012884470552794098,
+      "loss": 1.2467,
+      "step": 4774
+    },
+    {
+      "epoch": 1.4327406407082302,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001288296430185269,
+      "loss": 1.0012,
+      "step": 4775
+    },
+    {
+      "epoch": 1.433040738239928,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001288145805091128,
+      "loss": 1.2078,
+      "step": 4776
+    },
+    {
+      "epoch": 1.4333408357716257,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012879951799969877,
+      "loss": 0.9564,
+      "step": 4777
+    },
+    {
+      "epoch": 1.4336409333033235,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012878445549028468,
+      "loss": 1.03,
+      "step": 4778
+    },
+    {
+      "epoch": 1.4339410308350213,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012876939298087062,
+      "loss": 0.9414,
+      "step": 4779
+    },
+    {
+      "epoch": 1.4342411283667191,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012875433047145656,
+      "loss": 1.0321,
+      "step": 4780
+    },
+    {
+      "epoch": 1.434541225898417,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012873926796204248,
+      "loss": 1.0151,
+      "step": 4781
+    },
+    {
+      "epoch": 1.4348413234301147,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012872420545262842,
+      "loss": 1.1683,
+      "step": 4782
+    },
+    {
+      "epoch": 1.4351414209618125,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012870914294321436,
+      "loss": 1.0099,
+      "step": 4783
+    },
+    {
+      "epoch": 1.4354415184935103,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012869408043380027,
+      "loss": 1.0007,
+      "step": 4784
+    },
+    {
+      "epoch": 1.4357416160252081,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001286790179243862,
+      "loss": 1.1418,
+      "step": 4785
+    },
+    {
+      "epoch": 1.436041713556906,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012866395541497212,
+      "loss": 1.0451,
+      "step": 4786
+    },
+    {
+      "epoch": 1.436341811088604,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001286488929055581,
+      "loss": 0.9543,
+      "step": 4787
+    },
+    {
+      "epoch": 1.4366419086203015,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.000128633830396144,
+      "loss": 1.0521,
+      "step": 4788
+    },
+    {
+      "epoch": 1.4369420061519995,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012861876788672992,
+      "loss": 1.051,
+      "step": 4789
+    },
+    {
+      "epoch": 1.437242103683697,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012860370537731589,
+      "loss": 1.0533,
+      "step": 4790
+    },
+    {
+      "epoch": 1.437542201215395,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001285886428679018,
+      "loss": 1.0966,
+      "step": 4791
+    },
+    {
+      "epoch": 1.4378422987470927,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001285735803584877,
+      "loss": 1.0493,
+      "step": 4792
+    },
+    {
+      "epoch": 1.4381423962787907,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012855851784907368,
+      "loss": 1.1329,
+      "step": 4793
+    },
+    {
+      "epoch": 1.4384424938104883,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001285434553396596,
+      "loss": 0.988,
+      "step": 4794
+    },
+    {
+      "epoch": 1.4387425913421863,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001285283928302455,
+      "loss": 1.0874,
+      "step": 4795
+    },
+    {
+      "epoch": 1.439042688873884,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00012851333032083147,
+      "loss": 0.9597,
+      "step": 4796
+    },
+    {
+      "epoch": 1.4393427864055819,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012849826781141739,
+      "loss": 1.0835,
+      "step": 4797
+    },
+    {
+      "epoch": 1.4396428839372797,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012848320530200333,
+      "loss": 1.1477,
+      "step": 4798
+    },
+    {
+      "epoch": 1.4399429814689775,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012846814279258924,
+      "loss": 1.0477,
+      "step": 4799
+    },
+    {
+      "epoch": 1.4402430790006753,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00012845308028317518,
+      "loss": 0.9965,
+      "step": 4800
+    },
+    {
+      "epoch": 1.440543176532373,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012843801777376112,
+      "loss": 1.0345,
+      "step": 4801
+    },
+    {
+      "epoch": 1.4408432740640709,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012842295526434703,
+      "loss": 1.1117,
+      "step": 4802
+    },
+    {
+      "epoch": 1.4411433715957687,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012840789275493297,
+      "loss": 0.9895,
+      "step": 4803
+    },
+    {
+      "epoch": 1.4414434691274665,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012839283024551891,
+      "loss": 1.0881,
+      "step": 4804
+    },
+    {
+      "epoch": 1.4417435666591643,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00012837776773610483,
+      "loss": 0.9772,
+      "step": 4805
+    },
+    {
+      "epoch": 1.442043664190862,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001283627052266908,
+      "loss": 1.0033,
+      "step": 4806
+    },
+    {
+      "epoch": 1.4423437617225598,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001283476427172767,
+      "loss": 0.9727,
+      "step": 4807
+    },
+    {
+      "epoch": 1.4426438592542576,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012833258020786262,
+      "loss": 1.0893,
+      "step": 4808
+    },
+    {
+      "epoch": 1.4429439567859554,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001283175176984486,
+      "loss": 1.0976,
+      "step": 4809
+    },
+    {
+      "epoch": 1.4432440543176532,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001283024551890345,
+      "loss": 0.9643,
+      "step": 4810
+    },
+    {
+      "epoch": 1.443544151849351,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012828739267962041,
+      "loss": 1.0497,
+      "step": 4811
+    },
+    {
+      "epoch": 1.4438442493810488,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012827233017020636,
+      "loss": 1.055,
+      "step": 4812
+    },
+    {
+      "epoch": 1.4441443469127466,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001282572676607923,
+      "loss": 1.1497,
+      "step": 4813
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012824220515137824,
+      "loss": 1.4558,
+      "step": 4814
+    },
+    {
+      "epoch": 1.4447445419761422,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00012822714264196415,
+      "loss": 0.8834,
+      "step": 4815
+    },
+    {
+      "epoch": 1.44504463950784,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001282120801325501,
+      "loss": 1.032,
+      "step": 4816
+    },
+    {
+      "epoch": 1.4453447370395378,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012819701762313603,
+      "loss": 1.1129,
+      "step": 4817
+    },
+    {
+      "epoch": 1.4456448345712356,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012818195511372194,
+      "loss": 1.0245,
+      "step": 4818
+    },
+    {
+      "epoch": 1.4459449321029334,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012816689260430788,
+      "loss": 1.0402,
+      "step": 4819
+    },
+    {
+      "epoch": 1.4462450296346312,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012815183009489382,
+      "loss": 1.035,
+      "step": 4820
+    },
+    {
+      "epoch": 1.446545127166329,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012813676758547974,
+      "loss": 1.1569,
+      "step": 4821
+    },
+    {
+      "epoch": 1.4468452246980268,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012812170507606568,
+      "loss": 1.0215,
+      "step": 4822
+    },
+    {
+      "epoch": 1.4471453222297246,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012810664256665162,
+      "loss": 1.1554,
+      "step": 4823
+    },
+    {
+      "epoch": 1.4474454197614224,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012809158005723753,
+      "loss": 1.0141,
+      "step": 4824
+    },
+    {
+      "epoch": 1.4477455172931202,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00012807651754782347,
+      "loss": 1.0882,
+      "step": 4825
+    },
+    {
+      "epoch": 1.4480456148248182,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001280614550384094,
+      "loss": 1.0868,
+      "step": 4826
+    },
+    {
+      "epoch": 1.4483457123565158,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012804639252899532,
+      "loss": 0.9684,
+      "step": 4827
+    },
+    {
+      "epoch": 1.4486458098882138,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012803133001958126,
+      "loss": 1.1946,
+      "step": 4828
+    },
+    {
+      "epoch": 1.4489459074199114,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001280162675101672,
+      "loss": 1.0585,
+      "step": 4829
+    },
+    {
+      "epoch": 1.4492460049516094,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012800120500075312,
+      "loss": 1.1085,
+      "step": 4830
+    },
+    {
+      "epoch": 1.449546102483307,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012798614249133906,
+      "loss": 0.9802,
+      "step": 4831
+    },
+    {
+      "epoch": 1.449846200015005,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.000127971079981925,
+      "loss": 0.9858,
+      "step": 4832
+    },
+    {
+      "epoch": 1.4501462975467025,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012795601747251094,
+      "loss": 0.9986,
+      "step": 4833
+    },
+    {
+      "epoch": 1.4504463950784006,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012794095496309685,
+      "loss": 1.071,
+      "step": 4834
+    },
+    {
+      "epoch": 1.4507464926100984,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001279258924536828,
+      "loss": 1.158,
+      "step": 4835
+    },
+    {
+      "epoch": 1.4510465901417962,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012791082994426873,
+      "loss": 1.0905,
+      "step": 4836
+    },
+    {
+      "epoch": 1.451346687673494,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012789576743485465,
+      "loss": 1.0519,
+      "step": 4837
+    },
+    {
+      "epoch": 1.4516467852051917,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012788070492544059,
+      "loss": 1.0108,
+      "step": 4838
+    },
+    {
+      "epoch": 1.4519468827368895,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012786564241602653,
+      "loss": 0.8861,
+      "step": 4839
+    },
+    {
+      "epoch": 1.4522469802685873,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00012785057990661244,
+      "loss": 1.1633,
+      "step": 4840
+    },
+    {
+      "epoch": 1.4525470778002851,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012783551739719838,
+      "loss": 1.0956,
+      "step": 4841
+    },
+    {
+      "epoch": 1.452847175331983,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00012782045488778432,
+      "loss": 1.0065,
+      "step": 4842
+    },
+    {
+      "epoch": 1.4531472728636807,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012780539237837023,
+      "loss": 1.1177,
+      "step": 4843
+    },
+    {
+      "epoch": 1.4534473703953785,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00012779032986895617,
+      "loss": 0.9104,
+      "step": 4844
+    },
+    {
+      "epoch": 1.4537474679270763,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012777526735954211,
+      "loss": 1.1985,
+      "step": 4845
+    },
+    {
+      "epoch": 1.4540475654587741,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00012776020485012803,
+      "loss": 1.1305,
+      "step": 4846
+    },
+    {
+      "epoch": 1.454347662990472,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00012774514234071397,
+      "loss": 1.1795,
+      "step": 4847
+    },
+    {
+      "epoch": 1.4546477605221697,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001277300798312999,
+      "loss": 1.0585,
+      "step": 4848
+    },
+    {
+      "epoch": 1.4549478580538675,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012771501732188582,
+      "loss": 1.0034,
+      "step": 4849
+    },
+    {
+      "epoch": 1.4552479555855653,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012769995481247176,
+      "loss": 1.0853,
+      "step": 4850
+    },
+    {
+      "epoch": 1.455548053117263,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001276848923030577,
+      "loss": 0.9505,
+      "step": 4851
+    },
+    {
+      "epoch": 1.455848150648961,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00012766982979364364,
+      "loss": 1.1235,
+      "step": 4852
+    },
+    {
+      "epoch": 1.4561482481806587,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00012765476728422956,
+      "loss": 1.1898,
+      "step": 4853
+    },
+    {
+      "epoch": 1.4564483457123565,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.0001276397047748155,
+      "loss": 1.035,
+      "step": 4854
+    },
+    {
+      "epoch": 1.4567484432440543,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012762464226540144,
+      "loss": 1.0166,
+      "step": 4855
+    },
+    {
+      "epoch": 1.457048540775752,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012760957975598735,
+      "loss": 0.9837,
+      "step": 4856
+    },
+    {
+      "epoch": 1.4573486383074499,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00012759451724657326,
+      "loss": 1.1529,
+      "step": 4857
+    },
+    {
+      "epoch": 1.4576487358391477,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00012757945473715923,
+      "loss": 1.1063,
+      "step": 4858
+    },
+    {
+      "epoch": 1.4579488333708455,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00012756439222774514,
+      "loss": 1.3431,
+      "step": 4859
+    },
+    {
+      "epoch": 1.4582489309025433,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012754932971833108,
+      "loss": 0.8592,
+      "step": 4860
+    },
+    {
+      "epoch": 1.458549028434241,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012753426720891702,
+      "loss": 1.1412,
+      "step": 4861
+    },
+    {
+      "epoch": 1.4588491259659389,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012751920469950294,
+      "loss": 1.0968,
+      "step": 4862
+    },
+    {
+      "epoch": 1.4591492234976366,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012750414219008888,
+      "loss": 1.0817,
+      "step": 4863
+    },
+    {
+      "epoch": 1.4594493210293344,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012748907968067482,
+      "loss": 1.067,
+      "step": 4864
+    },
+    {
+      "epoch": 1.4597494185610325,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012747401717126073,
+      "loss": 1.1113,
+      "step": 4865
+    },
+    {
+      "epoch": 1.46004951609273,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012745895466184667,
+      "loss": 1.227,
+      "step": 4866
+    },
+    {
+      "epoch": 1.460349613624428,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001274438921524326,
+      "loss": 1.0601,
+      "step": 4867
+    },
+    {
+      "epoch": 1.4606497111561256,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012742882964301855,
+      "loss": 1.0922,
+      "step": 4868
+    },
+    {
+      "epoch": 1.4609498086878236,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012741376713360446,
+      "loss": 1.1111,
+      "step": 4869
+    },
+    {
+      "epoch": 1.4612499062195212,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012739870462419038,
+      "loss": 1.0425,
+      "step": 4870
+    },
+    {
+      "epoch": 1.4615500037512192,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00012738364211477635,
+      "loss": 1.018,
+      "step": 4871
+    },
+    {
+      "epoch": 1.4618501012829168,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012736857960536226,
+      "loss": 1.0842,
+      "step": 4872
+    },
+    {
+      "epoch": 1.4621501988146148,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00012735351709594817,
+      "loss": 1.1904,
+      "step": 4873
+    },
+    {
+      "epoch": 1.4624502963463126,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012733845458653414,
+      "loss": 1.0787,
+      "step": 4874
+    },
+    {
+      "epoch": 1.4627503938780104,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012732339207712005,
+      "loss": 0.8874,
+      "step": 4875
+    },
+    {
+      "epoch": 1.4630504914097082,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012730832956770597,
+      "loss": 0.9435,
+      "step": 4876
+    },
+    {
+      "epoch": 1.463350588941406,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012729326705829193,
+      "loss": 0.9393,
+      "step": 4877
+    },
+    {
+      "epoch": 1.4636506864731038,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012727820454887785,
+      "loss": 1.043,
+      "step": 4878
+    },
+    {
+      "epoch": 1.4639507840048016,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012726314203946379,
+      "loss": 1.1226,
+      "step": 4879
+    },
+    {
+      "epoch": 1.4642508815364994,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012724807953004973,
+      "loss": 1.032,
+      "step": 4880
+    },
+    {
+      "epoch": 1.4645509790681972,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00012723301702063564,
+      "loss": 1.1096,
+      "step": 4881
+    },
+    {
+      "epoch": 1.464851076599895,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012721795451122158,
+      "loss": 0.9778,
+      "step": 4882
+    },
+    {
+      "epoch": 1.4651511741315928,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001272028920018075,
+      "loss": 0.9586,
+      "step": 4883
+    },
+    {
+      "epoch": 1.4654512716632906,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012718782949239343,
+      "loss": 0.9724,
+      "step": 4884
+    },
+    {
+      "epoch": 1.4657513691949884,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012717276698297937,
+      "loss": 1.0517,
+      "step": 4885
+    },
+    {
+      "epoch": 1.4660514667266862,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001271577044735653,
+      "loss": 0.9741,
+      "step": 4886
+    },
+    {
+      "epoch": 1.466351564258384,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00012714264196415125,
+      "loss": 1.2258,
+      "step": 4887
+    },
+    {
+      "epoch": 1.4666516617900818,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012712757945473717,
+      "loss": 1.0788,
+      "step": 4888
+    },
+    {
+      "epoch": 1.4669517593217796,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012711251694532308,
+      "loss": 1.0053,
+      "step": 4889
+    },
+    {
+      "epoch": 1.4672518568534774,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012709745443590905,
+      "loss": 1.0093,
+      "step": 4890
+    },
+    {
+      "epoch": 1.4675519543851752,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012708239192649496,
+      "loss": 1.0473,
+      "step": 4891
+    },
+    {
+      "epoch": 1.467852051916873,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012706732941708087,
+      "loss": 1.035,
+      "step": 4892
+    },
+    {
+      "epoch": 1.4681521494485708,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012705226690766682,
+      "loss": 0.9784,
+      "step": 4893
+    },
+    {
+      "epoch": 1.4684522469802685,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012703720439825276,
+      "loss": 0.9437,
+      "step": 4894
+    },
+    {
+      "epoch": 1.4687523445119663,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001270221418888387,
+      "loss": 1.0546,
+      "step": 4895
+    },
+    {
+      "epoch": 1.4690524420436641,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001270070793794246,
+      "loss": 1.0823,
+      "step": 4896
+    },
+    {
+      "epoch": 1.469352539575362,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012699201687001055,
+      "loss": 1.1576,
+      "step": 4897
+    },
+    {
+      "epoch": 1.4696526371070597,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001269769543605965,
+      "loss": 1.0468,
+      "step": 4898
+    },
+    {
+      "epoch": 1.4699527346387575,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.0001269618918511824,
+      "loss": 1.1154,
+      "step": 4899
+    },
+    {
+      "epoch": 1.4702528321704553,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012694682934176834,
+      "loss": 0.9908,
+      "step": 4900
+    },
+    {
+      "epoch": 1.4705529297021531,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012693176683235428,
+      "loss": 1.0428,
+      "step": 4901
+    },
+    {
+      "epoch": 1.4708530272338511,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001269167043229402,
+      "loss": 1.0525,
+      "step": 4902
+    },
+    {
+      "epoch": 1.4711531247655487,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012690164181352614,
+      "loss": 0.9356,
+      "step": 4903
+    },
+    {
+      "epoch": 1.4714532222972467,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00012688657930411208,
+      "loss": 1.0937,
+      "step": 4904
+    },
+    {
+      "epoch": 1.4717533198289443,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.000126871516794698,
+      "loss": 0.9487,
+      "step": 4905
+    },
+    {
+      "epoch": 1.4720534173606423,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012685645428528393,
+      "loss": 0.9705,
+      "step": 4906
+    },
+    {
+      "epoch": 1.47235351489234,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012684139177586987,
+      "loss": 1.0668,
+      "step": 4907
+    },
+    {
+      "epoch": 1.472653612424038,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012682632926645578,
+      "loss": 1.1948,
+      "step": 4908
+    },
+    {
+      "epoch": 1.4729537099557355,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00012681126675704172,
+      "loss": 1.0736,
+      "step": 4909
+    },
+    {
+      "epoch": 1.4732538074874335,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012679620424762766,
+      "loss": 0.9928,
+      "step": 4910
+    },
+    {
+      "epoch": 1.473553905019131,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012678114173821358,
+      "loss": 1.0483,
+      "step": 4911
+    },
+    {
+      "epoch": 1.473854002550829,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012676607922879952,
+      "loss": 0.9029,
+      "step": 4912
+    },
+    {
+      "epoch": 1.474154100082527,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00012675101671938546,
+      "loss": 0.9703,
+      "step": 4913
+    },
+    {
+      "epoch": 1.4744541976142247,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001267359542099714,
+      "loss": 1.0779,
+      "step": 4914
+    },
+    {
+      "epoch": 1.4747542951459225,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001267208917005573,
+      "loss": 1.0548,
+      "step": 4915
+    },
+    {
+      "epoch": 1.4750543926776203,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012670582919114325,
+      "loss": 0.9619,
+      "step": 4916
+    },
+    {
+      "epoch": 1.475354490209318,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001266907666817292,
+      "loss": 1.1002,
+      "step": 4917
+    },
+    {
+      "epoch": 1.4756545877410159,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001266757041723151,
+      "loss": 1.0355,
+      "step": 4918
+    },
+    {
+      "epoch": 1.4759546852727137,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012666064166290105,
+      "loss": 1.0401,
+      "step": 4919
+    },
+    {
+      "epoch": 1.4762547828044115,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.000126645579153487,
+      "loss": 1.2436,
+      "step": 4920
+    },
+    {
+      "epoch": 1.4765548803361093,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.0001266305166440729,
+      "loss": 0.8793,
+      "step": 4921
+    },
+    {
+      "epoch": 1.476854977867807,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012661545413465884,
+      "loss": 1.0276,
+      "step": 4922
+    },
+    {
+      "epoch": 1.4771550753995049,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012660039162524478,
+      "loss": 1.1489,
+      "step": 4923
+    },
+    {
+      "epoch": 1.4774551729312027,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001265853291158307,
+      "loss": 1.1946,
+      "step": 4924
+    },
+    {
+      "epoch": 1.4777552704629004,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012657026660641663,
+      "loss": 1.0135,
+      "step": 4925
+    },
+    {
+      "epoch": 1.4780553679945982,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00012655520409700257,
+      "loss": 1.0851,
+      "step": 4926
+    },
+    {
+      "epoch": 1.478355465526296,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001265401415875885,
+      "loss": 0.904,
+      "step": 4927
+    },
+    {
+      "epoch": 1.4786555630579938,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00012652507907817443,
+      "loss": 1.0335,
+      "step": 4928
+    },
+    {
+      "epoch": 1.4789556605896916,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00012651001656876037,
+      "loss": 1.0583,
+      "step": 4929
+    },
+    {
+      "epoch": 1.4792557581213894,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012649495405934628,
+      "loss": 0.9736,
+      "step": 4930
+    },
+    {
+      "epoch": 1.4795558556530872,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00012647989154993222,
+      "loss": 1.3502,
+      "step": 4931
+    },
+    {
+      "epoch": 1.479855953184785,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012646482904051816,
+      "loss": 1.0769,
+      "step": 4932
+    },
+    {
+      "epoch": 1.4801560507164828,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001264497665311041,
+      "loss": 0.9719,
+      "step": 4933
+    },
+    {
+      "epoch": 1.4804561482481806,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00012643470402169002,
+      "loss": 1.085,
+      "step": 4934
+    },
+    {
+      "epoch": 1.4807562457798784,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00012641964151227596,
+      "loss": 1.1521,
+      "step": 4935
+    },
+    {
+      "epoch": 1.4810563433115762,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001264045790028619,
+      "loss": 0.9829,
+      "step": 4936
+    },
+    {
+      "epoch": 1.481356440843274,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001263895164934478,
+      "loss": 0.9387,
+      "step": 4937
+    },
+    {
+      "epoch": 1.4816565383749718,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00012637445398403375,
+      "loss": 0.9913,
+      "step": 4938
+    },
+    {
+      "epoch": 1.4819566359066696,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001263593914746197,
+      "loss": 1.1676,
+      "step": 4939
+    },
+    {
+      "epoch": 1.4822567334383674,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001263443289652056,
+      "loss": 0.9969,
+      "step": 4940
+    },
+    {
+      "epoch": 1.4825568309700654,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012632926645579154,
+      "loss": 1.0051,
+      "step": 4941
+    },
+    {
+      "epoch": 1.482856928501763,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012631420394637748,
+      "loss": 1.0977,
+      "step": 4942
+    },
+    {
+      "epoch": 1.483157026033461,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001262991414369634,
+      "loss": 0.9976,
+      "step": 4943
+    },
+    {
+      "epoch": 1.4834571235651586,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012628407892754934,
+      "loss": 0.9668,
+      "step": 4944
+    },
+    {
+      "epoch": 1.4837572210968566,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00012626901641813528,
+      "loss": 0.9141,
+      "step": 4945
+    },
+    {
+      "epoch": 1.4840573186285542,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.0001262539539087212,
+      "loss": 0.994,
+      "step": 4946
+    },
+    {
+      "epoch": 1.4843574161602522,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00012623889139930713,
+      "loss": 1.0207,
+      "step": 4947
+    },
+    {
+      "epoch": 1.4846575136919498,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012622382888989307,
+      "loss": 1.1028,
+      "step": 4948
+    },
+    {
+      "epoch": 1.4849576112236478,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.000126208766380479,
+      "loss": 1.0933,
+      "step": 4949
+    },
+    {
+      "epoch": 1.4852577087553454,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012619370387106492,
+      "loss": 1.0449,
+      "step": 4950
+    },
+    {
+      "epoch": 1.4855578062870434,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012617864136165084,
+      "loss": 1.0592,
+      "step": 4951
+    },
+    {
+      "epoch": 1.4858579038187412,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001261635788522368,
+      "loss": 1.0116,
+      "step": 4952
+    },
+    {
+      "epoch": 1.486158001350439,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00012614851634282272,
+      "loss": 1.1032,
+      "step": 4953
+    },
+    {
+      "epoch": 1.4864580988821368,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012613345383340863,
+      "loss": 1.0542,
+      "step": 4954
+    },
+    {
+      "epoch": 1.4867581964138346,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001261183913239946,
+      "loss": 1.0278,
+      "step": 4955
+    },
+    {
+      "epoch": 1.4870582939455324,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001261033288145805,
+      "loss": 1.0841,
+      "step": 4956
+    },
+    {
+      "epoch": 1.4873583914772301,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012608826630516643,
+      "loss": 1.1501,
+      "step": 4957
+    },
+    {
+      "epoch": 1.487658489008928,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001260732037957524,
+      "loss": 0.9568,
+      "step": 4958
+    },
+    {
+      "epoch": 1.4879585865406257,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001260581412863383,
+      "loss": 1.002,
+      "step": 4959
+    },
+    {
+      "epoch": 1.4882586840723235,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00012604307877692425,
+      "loss": 0.9392,
+      "step": 4960
+    },
+    {
+      "epoch": 1.4885587816040213,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001260280162675102,
+      "loss": 1.0982,
+      "step": 4961
+    },
+    {
+      "epoch": 1.4888588791357191,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001260129537580961,
+      "loss": 1.1464,
+      "step": 4962
+    },
+    {
+      "epoch": 1.489158976667417,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012599789124868204,
+      "loss": 1.0122,
+      "step": 4963
+    },
+    {
+      "epoch": 1.4894590741991147,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012598282873926795,
+      "loss": 1.1906,
+      "step": 4964
+    },
+    {
+      "epoch": 1.4897591717308125,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001259677662298539,
+      "loss": 0.91,
+      "step": 4965
+    },
+    {
+      "epoch": 1.4900592692625103,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012595270372043983,
+      "loss": 1.133,
+      "step": 4966
+    },
+    {
+      "epoch": 1.490359366794208,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00012593764121102575,
+      "loss": 0.8585,
+      "step": 4967
+    },
+    {
+      "epoch": 1.490659464325906,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012592257870161171,
+      "loss": 1.0061,
+      "step": 4968
+    },
+    {
+      "epoch": 1.4909595618576037,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012590751619219763,
+      "loss": 1.1638,
+      "step": 4969
+    },
+    {
+      "epoch": 1.4912596593893015,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012589245368278354,
+      "loss": 1.1258,
+      "step": 4970
+    },
+    {
+      "epoch": 1.4915597569209993,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001258773911733695,
+      "loss": 1.0205,
+      "step": 4971
+    },
+    {
+      "epoch": 1.491859854452697,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012586232866395542,
+      "loss": 1.0167,
+      "step": 4972
+    },
+    {
+      "epoch": 1.4921599519843949,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00012584726615454133,
+      "loss": 0.9083,
+      "step": 4973
+    },
+    {
+      "epoch": 1.4924600495160927,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001258322036451273,
+      "loss": 1.0307,
+      "step": 4974
+    },
+    {
+      "epoch": 1.4927601470477905,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012581714113571322,
+      "loss": 1.0001,
+      "step": 4975
+    },
+    {
+      "epoch": 1.4930602445794883,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012580207862629916,
+      "loss": 1.0376,
+      "step": 4976
+    },
+    {
+      "epoch": 1.493360342111186,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012578701611688507,
+      "loss": 1.1409,
+      "step": 4977
+    },
+    {
+      "epoch": 1.4936604396428839,
+      "grad_norm": 1.0,
+      "learning_rate": 0.000125771953607471,
+      "loss": 1.0459,
+      "step": 4978
+    },
+    {
+      "epoch": 1.4939605371745817,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012575689109805695,
+      "loss": 0.976,
+      "step": 4979
+    },
+    {
+      "epoch": 1.4942606347062797,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00012574182858864286,
+      "loss": 0.9414,
+      "step": 4980
+    },
+    {
+      "epoch": 1.4945607322379773,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001257267660792288,
+      "loss": 1.186,
+      "step": 4981
+    },
+    {
+      "epoch": 1.4948608297696753,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012571170356981474,
+      "loss": 0.9915,
+      "step": 4982
+    },
+    {
+      "epoch": 1.4951609273013728,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00012569664106040066,
+      "loss": 0.8764,
+      "step": 4983
+    },
+    {
+      "epoch": 1.4954610248330709,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001256815785509866,
+      "loss": 0.996,
+      "step": 4984
+    },
+    {
+      "epoch": 1.4957611223647684,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012566651604157254,
+      "loss": 1.069,
+      "step": 4985
+    },
+    {
+      "epoch": 1.4960612198964665,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012565145353215845,
+      "loss": 1.0565,
+      "step": 4986
+    },
+    {
+      "epoch": 1.496361317428164,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001256363910227444,
+      "loss": 1.1376,
+      "step": 4987
+    },
+    {
+      "epoch": 1.496661414959862,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012562132851333033,
+      "loss": 1.0323,
+      "step": 4988
+    },
+    {
+      "epoch": 1.4969615124915596,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012560626600391624,
+      "loss": 0.9719,
+      "step": 4989
+    },
+    {
+      "epoch": 1.4972616100232576,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012559120349450218,
+      "loss": 0.9821,
+      "step": 4990
+    },
+    {
+      "epoch": 1.4975617075549554,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012557614098508812,
+      "loss": 1.0425,
+      "step": 4991
+    },
+    {
+      "epoch": 1.4978618050866532,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012556107847567404,
+      "loss": 1.0035,
+      "step": 4992
+    },
+    {
+      "epoch": 1.498161902618351,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012554601596625998,
+      "loss": 0.9452,
+      "step": 4993
+    },
+    {
+      "epoch": 1.4984620001500488,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00012553095345684592,
+      "loss": 1.1393,
+      "step": 4994
+    },
+    {
+      "epoch": 1.4987620976817466,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012551589094743186,
+      "loss": 1.0544,
+      "step": 4995
+    },
+    {
+      "epoch": 1.4990621952134444,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00012550082843801777,
+      "loss": 1.0265,
+      "step": 4996
+    },
+    {
+      "epoch": 1.4993622927451422,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001254857659286037,
+      "loss": 1.2215,
+      "step": 4997
+    },
+    {
+      "epoch": 1.49966239027684,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012547070341918965,
+      "loss": 1.0711,
+      "step": 4998
+    },
+    {
+      "epoch": 1.4999624878085378,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012545564090977557,
+      "loss": 1.0076,
+      "step": 4999
+    },
+    {
+      "epoch": 1.5002625853402356,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001254405784003615,
+      "loss": 0.9993,
+      "step": 5000
+    },
+    {
+      "epoch": 1.5005626828719334,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012542551589094745,
+      "loss": 1.1032,
+      "step": 5001
+    },
+    {
+      "epoch": 1.5008627804036312,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012541045338153336,
+      "loss": 1.0624,
+      "step": 5002
+    },
+    {
+      "epoch": 1.501162877935329,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001253953908721193,
+      "loss": 1.167,
+      "step": 5003
+    },
+    {
+      "epoch": 1.5014629754670268,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00012538032836270524,
+      "loss": 0.8643,
+      "step": 5004
+    },
+    {
+      "epoch": 1.5017630729987246,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012536526585329115,
+      "loss": 0.9996,
+      "step": 5005
+    },
+    {
+      "epoch": 1.5020631705304224,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001253502033438771,
+      "loss": 1.1792,
+      "step": 5006
+    },
+    {
+      "epoch": 1.5023632680621202,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012533514083446303,
+      "loss": 1.0315,
+      "step": 5007
+    },
+    {
+      "epoch": 1.502663365593818,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012532007832504895,
+      "loss": 1.0014,
+      "step": 5008
+    },
+    {
+      "epoch": 1.5029634631255158,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001253050158156349,
+      "loss": 1.0028,
+      "step": 5009
+    },
+    {
+      "epoch": 1.5032635606572136,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012528995330622083,
+      "loss": 1.1904,
+      "step": 5010
+    },
+    {
+      "epoch": 1.5035636581889114,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00012527489079680674,
+      "loss": 0.9956,
+      "step": 5011
+    },
+    {
+      "epoch": 1.5038637557206092,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012525982828739268,
+      "loss": 1.0837,
+      "step": 5012
+    },
+    {
+      "epoch": 1.504163853252307,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012524476577797862,
+      "loss": 1.0464,
+      "step": 5013
+    },
+    {
+      "epoch": 1.5044639507840047,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012522970326856456,
+      "loss": 1.0015,
+      "step": 5014
+    },
+    {
+      "epoch": 1.5047640483157028,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012521464075915048,
+      "loss": 1.0971,
+      "step": 5015
+    },
+    {
+      "epoch": 1.5050641458474003,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012519957824973642,
+      "loss": 0.892,
+      "step": 5016
+    },
+    {
+      "epoch": 1.5053642433790984,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012518451574032236,
+      "loss": 1.0328,
+      "step": 5017
+    },
+    {
+      "epoch": 1.505664340910796,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012516945323090827,
+      "loss": 1.0981,
+      "step": 5018
+    },
+    {
+      "epoch": 1.505964438442494,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001251543907214942,
+      "loss": 1.0101,
+      "step": 5019
+    },
+    {
+      "epoch": 1.5062645359741915,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012513932821208015,
+      "loss": 1.1252,
+      "step": 5020
+    },
+    {
+      "epoch": 1.5065646335058895,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012512426570266606,
+      "loss": 1.1108,
+      "step": 5021
+    },
+    {
+      "epoch": 1.5068647310375871,
+      "grad_norm": 1.25,
+      "learning_rate": 0.000125109203193252,
+      "loss": 1.0287,
+      "step": 5022
+    },
+    {
+      "epoch": 1.5071648285692851,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012509414068383794,
+      "loss": 1.1225,
+      "step": 5023
+    },
+    {
+      "epoch": 1.5074649261009827,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00012507907817442386,
+      "loss": 1.105,
+      "step": 5024
+    },
+    {
+      "epoch": 1.5077650236326807,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001250640156650098,
+      "loss": 1.1168,
+      "step": 5025
+    },
+    {
+      "epoch": 1.5080651211643783,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012504895315559574,
+      "loss": 1.0107,
+      "step": 5026
+    },
+    {
+      "epoch": 1.5083652186960763,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00012503389064618165,
+      "loss": 0.9115,
+      "step": 5027
+    },
+    {
+      "epoch": 1.508665316227774,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001250188281367676,
+      "loss": 1.2458,
+      "step": 5028
+    },
+    {
+      "epoch": 1.508965413759472,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012500376562735353,
+      "loss": 0.8876,
+      "step": 5029
+    },
+    {
+      "epoch": 1.5092655112911695,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012498870311793947,
+      "loss": 1.0663,
+      "step": 5030
+    },
+    {
+      "epoch": 1.5095656088228675,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00012497364060852538,
+      "loss": 0.9957,
+      "step": 5031
+    },
+    {
+      "epoch": 1.509865706354565,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00012495857809911132,
+      "loss": 0.9948,
+      "step": 5032
+    },
+    {
+      "epoch": 1.510165803886263,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012494351558969727,
+      "loss": 1.0322,
+      "step": 5033
+    },
+    {
+      "epoch": 1.5104659014179609,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00012492845308028318,
+      "loss": 1.0154,
+      "step": 5034
+    },
+    {
+      "epoch": 1.5107659989496587,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001249133905708691,
+      "loss": 1.0747,
+      "step": 5035
+    },
+    {
+      "epoch": 1.5110660964813565,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012489832806145506,
+      "loss": 1.0061,
+      "step": 5036
+    },
+    {
+      "epoch": 1.5113661940130543,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012488326555204097,
+      "loss": 0.9763,
+      "step": 5037
+    },
+    {
+      "epoch": 1.511666291544752,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012486820304262689,
+      "loss": 1.1218,
+      "step": 5038
+    },
+    {
+      "epoch": 1.5119663890764499,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012485314053321285,
+      "loss": 1.1147,
+      "step": 5039
+    },
+    {
+      "epoch": 1.5122664866081477,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012483807802379877,
+      "loss": 1.1731,
+      "step": 5040
+    },
+    {
+      "epoch": 1.5125665841398455,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001248230155143847,
+      "loss": 1.1086,
+      "step": 5041
+    },
+    {
+      "epoch": 1.5128666816715433,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012480795300497065,
+      "loss": 0.9954,
+      "step": 5042
+    },
+    {
+      "epoch": 1.513166779203241,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00012479289049555656,
+      "loss": 1.0946,
+      "step": 5043
+    },
+    {
+      "epoch": 1.5134668767349388,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001247778279861425,
+      "loss": 1.3248,
+      "step": 5044
+    },
+    {
+      "epoch": 1.5137669742666366,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012476276547672844,
+      "loss": 1.0467,
+      "step": 5045
+    },
+    {
+      "epoch": 1.5140670717983344,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00012474770296731435,
+      "loss": 0.9471,
+      "step": 5046
+    },
+    {
+      "epoch": 1.5143671693300322,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001247326404579003,
+      "loss": 1.015,
+      "step": 5047
+    },
+    {
+      "epoch": 1.51466726686173,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001247175779484862,
+      "loss": 1.0711,
+      "step": 5048
+    },
+    {
+      "epoch": 1.5149673643934278,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012470251543907217,
+      "loss": 1.1418,
+      "step": 5049
+    },
+    {
+      "epoch": 1.5152674619251256,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001246874529296581,
+      "loss": 0.9841,
+      "step": 5050
+    },
+    {
+      "epoch": 1.5155675594568234,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.000124672390420244,
+      "loss": 1.1658,
+      "step": 5051
+    },
+    {
+      "epoch": 1.5158676569885212,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012465732791082997,
+      "loss": 0.9869,
+      "step": 5052
+    },
+    {
+      "epoch": 1.516167754520219,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00012464226540141588,
+      "loss": 1.0856,
+      "step": 5053
+    },
+    {
+      "epoch": 1.516467852051917,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001246272028920018,
+      "loss": 1.057,
+      "step": 5054
+    },
+    {
+      "epoch": 1.5167679495836146,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012461214038258776,
+      "loss": 1.1091,
+      "step": 5055
+    },
+    {
+      "epoch": 1.5170680471153126,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012459707787317368,
+      "loss": 1.0528,
+      "step": 5056
+    },
+    {
+      "epoch": 1.5173681446470102,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012458201536375962,
+      "loss": 0.8778,
+      "step": 5057
+    },
+    {
+      "epoch": 1.5176682421787082,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012456695285434553,
+      "loss": 1.0562,
+      "step": 5058
+    },
+    {
+      "epoch": 1.5179683397104058,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012455189034493147,
+      "loss": 0.9595,
+      "step": 5059
+    },
+    {
+      "epoch": 1.5182684372421038,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001245368278355174,
+      "loss": 1.0376,
+      "step": 5060
+    },
+    {
+      "epoch": 1.5185685347738014,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00012452176532610332,
+      "loss": 1.082,
+      "step": 5061
+    },
+    {
+      "epoch": 1.5188686323054994,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012450670281668926,
+      "loss": 1.0593,
+      "step": 5062
+    },
+    {
+      "epoch": 1.519168729837197,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001244916403072752,
+      "loss": 1.0403,
+      "step": 5063
+    },
+    {
+      "epoch": 1.519468827368895,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012447657779786112,
+      "loss": 1.0821,
+      "step": 5064
+    },
+    {
+      "epoch": 1.5197689249005926,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012446151528844706,
+      "loss": 0.9593,
+      "step": 5065
+    },
+    {
+      "epoch": 1.5200690224322906,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.000124446452779033,
+      "loss": 1.1147,
+      "step": 5066
+    },
+    {
+      "epoch": 1.5203691199639882,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001244313902696189,
+      "loss": 0.985,
+      "step": 5067
+    },
+    {
+      "epoch": 1.5206692174956862,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012441632776020488,
+      "loss": 0.8792,
+      "step": 5068
+    },
+    {
+      "epoch": 1.5209693150273838,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.0001244012652507908,
+      "loss": 1.1097,
+      "step": 5069
+    },
+    {
+      "epoch": 1.5212694125590818,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001243862027413767,
+      "loss": 1.0498,
+      "step": 5070
+    },
+    {
+      "epoch": 1.5215695100907793,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012437114023196264,
+      "loss": 1.0244,
+      "step": 5071
+    },
+    {
+      "epoch": 1.5218696076224774,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012435607772254858,
+      "loss": 0.995,
+      "step": 5072
+    },
+    {
+      "epoch": 1.5221697051541752,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001243410152131345,
+      "loss": 1.1414,
+      "step": 5073
+    },
+    {
+      "epoch": 1.522469802685873,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012432595270372044,
+      "loss": 1.0495,
+      "step": 5074
+    },
+    {
+      "epoch": 1.5227699002175707,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012431089019430638,
+      "loss": 1.0383,
+      "step": 5075
+    },
+    {
+      "epoch": 1.5230699977492685,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012429582768489232,
+      "loss": 1.0703,
+      "step": 5076
+    },
+    {
+      "epoch": 1.5233700952809663,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012428076517547823,
+      "loss": 1.0393,
+      "step": 5077
+    },
+    {
+      "epoch": 1.5236701928126641,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012426570266606417,
+      "loss": 1.086,
+      "step": 5078
+    },
+    {
+      "epoch": 1.523970290344362,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001242506401566501,
+      "loss": 0.9824,
+      "step": 5079
+    },
+    {
+      "epoch": 1.5242703878760597,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012423557764723603,
+      "loss": 1.0824,
+      "step": 5080
+    },
+    {
+      "epoch": 1.5245704854077575,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012422051513782197,
+      "loss": 0.9764,
+      "step": 5081
+    },
+    {
+      "epoch": 1.5248705829394553,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001242054526284079,
+      "loss": 0.9952,
+      "step": 5082
+    },
+    {
+      "epoch": 1.5251706804711531,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012419039011899382,
+      "loss": 1.1464,
+      "step": 5083
+    },
+    {
+      "epoch": 1.525470778002851,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012417532760957976,
+      "loss": 1.0132,
+      "step": 5084
+    },
+    {
+      "epoch": 1.5257708755345487,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001241602651001657,
+      "loss": 0.9163,
+      "step": 5085
+    },
+    {
+      "epoch": 1.5260709730662465,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001241452025907516,
+      "loss": 1.0709,
+      "step": 5086
+    },
+    {
+      "epoch": 1.5263710705979443,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012413014008133755,
+      "loss": 1.0971,
+      "step": 5087
+    },
+    {
+      "epoch": 1.526671168129642,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001241150775719235,
+      "loss": 1.0961,
+      "step": 5088
+    },
+    {
+      "epoch": 1.52697126566134,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001241000150625094,
+      "loss": 1.0215,
+      "step": 5089
+    },
+    {
+      "epoch": 1.5272713631930377,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012408495255309535,
+      "loss": 0.9396,
+      "step": 5090
+    },
+    {
+      "epoch": 1.5275714607247355,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001240698900436813,
+      "loss": 1.1683,
+      "step": 5091
+    },
+    {
+      "epoch": 1.5278715582564333,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001240548275342672,
+      "loss": 1.0307,
+      "step": 5092
+    },
+    {
+      "epoch": 1.5281716557881313,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012403976502485314,
+      "loss": 1.1189,
+      "step": 5093
+    },
+    {
+      "epoch": 1.5284717533198289,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00012402470251543908,
+      "loss": 1.1077,
+      "step": 5094
+    },
+    {
+      "epoch": 1.528771850851527,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012400964000602502,
+      "loss": 1.0542,
+      "step": 5095
+    },
+    {
+      "epoch": 1.5290719483832245,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00012399457749661094,
+      "loss": 0.9872,
+      "step": 5096
+    },
+    {
+      "epoch": 1.5293720459149225,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00012397951498719688,
+      "loss": 1.1125,
+      "step": 5097
+    },
+    {
+      "epoch": 1.52967214344662,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012396445247778282,
+      "loss": 1.017,
+      "step": 5098
+    },
+    {
+      "epoch": 1.529972240978318,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00012394938996836873,
+      "loss": 1.1249,
+      "step": 5099
+    },
+    {
+      "epoch": 1.5302723385100157,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012393432745895467,
+      "loss": 1.0318,
+      "step": 5100
+    },
+    {
+      "epoch": 1.5305724360417137,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001239192649495406,
+      "loss": 1.0781,
+      "step": 5101
+    },
+    {
+      "epoch": 1.5308725335734112,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012390420244012652,
+      "loss": 1.0366,
+      "step": 5102
+    },
+    {
+      "epoch": 1.5311726311051093,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012388913993071246,
+      "loss": 1.0656,
+      "step": 5103
+    },
+    {
+      "epoch": 1.5314727286368068,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001238740774212984,
+      "loss": 1.0406,
+      "step": 5104
+    },
+    {
+      "epoch": 1.5317728261685049,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012385901491188432,
+      "loss": 1.1039,
+      "step": 5105
+    },
+    {
+      "epoch": 1.5320729237002024,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012384395240247026,
+      "loss": 1.212,
+      "step": 5106
+    },
+    {
+      "epoch": 1.5323730212319004,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001238288898930562,
+      "loss": 1.0711,
+      "step": 5107
+    },
+    {
+      "epoch": 1.532673118763598,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001238138273836421,
+      "loss": 0.9831,
+      "step": 5108
+    },
+    {
+      "epoch": 1.532973216295296,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012379876487422805,
+      "loss": 1.005,
+      "step": 5109
+    },
+    {
+      "epoch": 1.5332733138269936,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.000123783702364814,
+      "loss": 1.1342,
+      "step": 5110
+    },
+    {
+      "epoch": 1.5335734113586916,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012376863985539993,
+      "loss": 1.0686,
+      "step": 5111
+    },
+    {
+      "epoch": 1.5338735088903894,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012375357734598584,
+      "loss": 1.0282,
+      "step": 5112
+    },
+    {
+      "epoch": 1.5341736064220872,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012373851483657178,
+      "loss": 1.0602,
+      "step": 5113
+    },
+    {
+      "epoch": 1.534473703953785,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00012372345232715773,
+      "loss": 1.1392,
+      "step": 5114
+    },
+    {
+      "epoch": 1.5347738014854828,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012370838981774364,
+      "loss": 0.9483,
+      "step": 5115
+    },
+    {
+      "epoch": 1.5350738990171806,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012369332730832958,
+      "loss": 1.0342,
+      "step": 5116
+    },
+    {
+      "epoch": 1.5353739965488784,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012367826479891552,
+      "loss": 0.9458,
+      "step": 5117
+    },
+    {
+      "epoch": 1.5356740940805762,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012366320228950143,
+      "loss": 1.1116,
+      "step": 5118
+    },
+    {
+      "epoch": 1.535974191612274,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00012364813978008735,
+      "loss": 1.123,
+      "step": 5119
+    },
+    {
+      "epoch": 1.5362742891439718,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001236330772706733,
+      "loss": 0.8916,
+      "step": 5120
+    },
+    {
+      "epoch": 1.5365743866756696,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00012361801476125923,
+      "loss": 1.1133,
+      "step": 5121
+    },
+    {
+      "epoch": 1.5368744842073674,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012360295225184517,
+      "loss": 1.1128,
+      "step": 5122
+    },
+    {
+      "epoch": 1.5371745817390652,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001235878897424311,
+      "loss": 1.0557,
+      "step": 5123
+    },
+    {
+      "epoch": 1.537474679270763,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012357282723301702,
+      "loss": 0.9653,
+      "step": 5124
+    },
+    {
+      "epoch": 1.5377747768024608,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012355776472360296,
+      "loss": 0.9879,
+      "step": 5125
+    },
+    {
+      "epoch": 1.5380748743341586,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001235427022141889,
+      "loss": 1.0602,
+      "step": 5126
+    },
+    {
+      "epoch": 1.5383749718658564,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00012352763970477481,
+      "loss": 0.9673,
+      "step": 5127
+    },
+    {
+      "epoch": 1.5386750693975542,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012351257719536075,
+      "loss": 1.0449,
+      "step": 5128
+    },
+    {
+      "epoch": 1.538975166929252,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012349751468594667,
+      "loss": 0.9664,
+      "step": 5129
+    },
+    {
+      "epoch": 1.5392752644609498,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012348245217653263,
+      "loss": 1.0322,
+      "step": 5130
+    },
+    {
+      "epoch": 1.5395753619926476,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00012346738966711855,
+      "loss": 0.9945,
+      "step": 5131
+    },
+    {
+      "epoch": 1.5398754595243456,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012345232715770446,
+      "loss": 0.9086,
+      "step": 5132
+    },
+    {
+      "epoch": 1.5401755570560431,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012343726464829043,
+      "loss": 1.0148,
+      "step": 5133
+    },
+    {
+      "epoch": 1.5404756545877412,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012342220213887634,
+      "loss": 1.1268,
+      "step": 5134
+    },
+    {
+      "epoch": 1.5407757521194387,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012340713962946225,
+      "loss": 1.0982,
+      "step": 5135
+    },
+    {
+      "epoch": 1.5410758496511368,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012339207712004822,
+      "loss": 0.9988,
+      "step": 5136
+    },
+    {
+      "epoch": 1.5413759471828343,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012337701461063414,
+      "loss": 1.0259,
+      "step": 5137
+    },
+    {
+      "epoch": 1.5416760447145323,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012336195210122008,
+      "loss": 1.0475,
+      "step": 5138
+    },
+    {
+      "epoch": 1.54197614224623,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012334688959180602,
+      "loss": 1.0412,
+      "step": 5139
+    },
+    {
+      "epoch": 1.542276239777928,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00012333182708239193,
+      "loss": 0.9795,
+      "step": 5140
+    },
+    {
+      "epoch": 1.5425763373096255,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012331676457297787,
+      "loss": 0.9727,
+      "step": 5141
+    },
+    {
+      "epoch": 1.5428764348413235,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00012330170206356378,
+      "loss": 0.9782,
+      "step": 5142
+    },
+    {
+      "epoch": 1.543176532373021,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012328663955414972,
+      "loss": 1.0613,
+      "step": 5143
+    },
+    {
+      "epoch": 1.5434766299047191,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012327157704473566,
+      "loss": 1.1287,
+      "step": 5144
+    },
+    {
+      "epoch": 1.5437767274364167,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00012325651453532158,
+      "loss": 1.0785,
+      "step": 5145
+    },
+    {
+      "epoch": 1.5440768249681147,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00012324145202590752,
+      "loss": 1.1445,
+      "step": 5146
+    },
+    {
+      "epoch": 1.5443769224998123,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012322638951649346,
+      "loss": 1.0681,
+      "step": 5147
+    },
+    {
+      "epoch": 1.5446770200315103,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012321132700707937,
+      "loss": 1.0105,
+      "step": 5148
+    },
+    {
+      "epoch": 1.5449771175632079,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00012319626449766534,
+      "loss": 0.9129,
+      "step": 5149
+    },
+    {
+      "epoch": 1.545277215094906,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012318120198825125,
+      "loss": 1.0011,
+      "step": 5150
+    },
+    {
+      "epoch": 1.5455773126266037,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00012316613947883716,
+      "loss": 1.0156,
+      "step": 5151
+    },
+    {
+      "epoch": 1.5458774101583015,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012315107696942313,
+      "loss": 1.0164,
+      "step": 5152
+    },
+    {
+      "epoch": 1.5461775076899993,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00012313601446000904,
+      "loss": 0.9314,
+      "step": 5153
+    },
+    {
+      "epoch": 1.546477605221697,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012312095195059496,
+      "loss": 0.9809,
+      "step": 5154
+    },
+    {
+      "epoch": 1.5467777027533949,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001231058894411809,
+      "loss": 1.0034,
+      "step": 5155
+    },
+    {
+      "epoch": 1.5470778002850927,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012309082693176684,
+      "loss": 1.1758,
+      "step": 5156
+    },
+    {
+      "epoch": 1.5473778978167905,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00012307576442235278,
+      "loss": 0.9658,
+      "step": 5157
+    },
+    {
+      "epoch": 1.5476779953484883,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001230607019129387,
+      "loss": 0.9624,
+      "step": 5158
+    },
+    {
+      "epoch": 1.547978092880186,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00012304563940352463,
+      "loss": 0.9373,
+      "step": 5159
+    },
+    {
+      "epoch": 1.5482781904118839,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012303057689411057,
+      "loss": 1.0079,
+      "step": 5160
+    },
+    {
+      "epoch": 1.5485782879435817,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012301551438469649,
+      "loss": 1.1122,
+      "step": 5161
+    },
+    {
+      "epoch": 1.5488783854752795,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012300045187528243,
+      "loss": 1.1164,
+      "step": 5162
+    },
+    {
+      "epoch": 1.5491784830069772,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012298538936586837,
+      "loss": 0.9653,
+      "step": 5163
+    },
+    {
+      "epoch": 1.549478580538675,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012297032685645428,
+      "loss": 1.0632,
+      "step": 5164
+    },
+    {
+      "epoch": 1.5497786780703728,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012295526434704022,
+      "loss": 1.1142,
+      "step": 5165
+    },
+    {
+      "epoch": 1.5500787756020706,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012294020183762616,
+      "loss": 1.1531,
+      "step": 5166
+    },
+    {
+      "epoch": 1.5503788731337684,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00012292513932821207,
+      "loss": 1.047,
+      "step": 5167
+    },
+    {
+      "epoch": 1.5506789706654662,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012291007681879801,
+      "loss": 1.0978,
+      "step": 5168
+    },
+    {
+      "epoch": 1.550979068197164,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00012289501430938395,
+      "loss": 0.9251,
+      "step": 5169
+    },
+    {
+      "epoch": 1.5512791657288618,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00012287995179996987,
+      "loss": 1.0982,
+      "step": 5170
+    },
+    {
+      "epoch": 1.5515792632605598,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001228648892905558,
+      "loss": 0.9792,
+      "step": 5171
+    },
+    {
+      "epoch": 1.5518793607922574,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00012284982678114175,
+      "loss": 1.0515,
+      "step": 5172
+    },
+    {
+      "epoch": 1.5521794583239554,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012283476427172766,
+      "loss": 0.9583,
+      "step": 5173
+    },
+    {
+      "epoch": 1.552479555855653,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001228197017623136,
+      "loss": 1.0282,
+      "step": 5174
+    },
+    {
+      "epoch": 1.552779653387351,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00012280463925289954,
+      "loss": 1.1104,
+      "step": 5175
+    },
+    {
+      "epoch": 1.5530797509190486,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012278957674348548,
+      "loss": 0.9607,
+      "step": 5176
+    },
+    {
+      "epoch": 1.5533798484507466,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001227745142340714,
+      "loss": 0.9572,
+      "step": 5177
+    },
+    {
+      "epoch": 1.5536799459824442,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012275945172465734,
+      "loss": 1.0454,
+      "step": 5178
+    },
+    {
+      "epoch": 1.5539800435141422,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012274438921524328,
+      "loss": 1.0256,
+      "step": 5179
+    },
+    {
+      "epoch": 1.5542801410458398,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001227293267058292,
+      "loss": 1.0223,
+      "step": 5180
+    },
+    {
+      "epoch": 1.5545802385775378,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00012271426419641513,
+      "loss": 0.9387,
+      "step": 5181
+    },
+    {
+      "epoch": 1.5548803361092354,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012269920168700107,
+      "loss": 1.0829,
+      "step": 5182
+    },
+    {
+      "epoch": 1.5551804336409334,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00012268413917758698,
+      "loss": 1.1086,
+      "step": 5183
+    },
+    {
+      "epoch": 1.555480531172631,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012266907666817292,
+      "loss": 1.066,
+      "step": 5184
+    },
+    {
+      "epoch": 1.555780628704329,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012265401415875886,
+      "loss": 0.9704,
+      "step": 5185
+    },
+    {
+      "epoch": 1.5560807262360266,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012263895164934478,
+      "loss": 1.0297,
+      "step": 5186
+    },
+    {
+      "epoch": 1.5563808237677246,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012262388913993072,
+      "loss": 1.1339,
+      "step": 5187
+    },
+    {
+      "epoch": 1.5566809212994221,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012260882663051666,
+      "loss": 1.0288,
+      "step": 5188
+    },
+    {
+      "epoch": 1.5569810188311202,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012259376412110257,
+      "loss": 1.0583,
+      "step": 5189
+    },
+    {
+      "epoch": 1.557281116362818,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001225787016116885,
+      "loss": 1.2154,
+      "step": 5190
+    },
+    {
+      "epoch": 1.5575812138945158,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00012256363910227445,
+      "loss": 1.1012,
+      "step": 5191
+    },
+    {
+      "epoch": 1.5578813114262136,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001225485765928604,
+      "loss": 0.9475,
+      "step": 5192
+    },
+    {
+      "epoch": 1.5581814089579114,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001225335140834463,
+      "loss": 1.2704,
+      "step": 5193
+    },
+    {
+      "epoch": 1.5584815064896091,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00012251845157403224,
+      "loss": 0.8758,
+      "step": 5194
+    },
+    {
+      "epoch": 1.558781604021307,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012250338906461819,
+      "loss": 1.114,
+      "step": 5195
+    },
+    {
+      "epoch": 1.5590817015530047,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001224883265552041,
+      "loss": 0.8449,
+      "step": 5196
+    },
+    {
+      "epoch": 1.5593817990847025,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012247326404579004,
+      "loss": 1.0866,
+      "step": 5197
+    },
+    {
+      "epoch": 1.5596818966164003,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00012245820153637598,
+      "loss": 0.9713,
+      "step": 5198
+    },
+    {
+      "epoch": 1.5599819941480981,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001224431390269619,
+      "loss": 0.9345,
+      "step": 5199
+    },
+    {
+      "epoch": 1.560282091679796,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001224280765175478,
+      "loss": 0.931,
+      "step": 5200
+    },
+    {
+      "epoch": 1.5605821892114937,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00012241301400813377,
+      "loss": 1.0098,
+      "step": 5201
+    },
+    {
+      "epoch": 1.5608822867431915,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012239795149871969,
+      "loss": 1.0506,
+      "step": 5202
+    },
+    {
+      "epoch": 1.5611823842748893,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012238288898930563,
+      "loss": 0.9932,
+      "step": 5203
+    },
+    {
+      "epoch": 1.561482481806587,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012236782647989157,
+      "loss": 0.9911,
+      "step": 5204
+    },
+    {
+      "epoch": 1.561782579338285,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00012235276397047748,
+      "loss": 1.0089,
+      "step": 5205
+    },
+    {
+      "epoch": 1.5620826768699827,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012233770146106342,
+      "loss": 1.0379,
+      "step": 5206
+    },
+    {
+      "epoch": 1.5623827744016805,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012232263895164936,
+      "loss": 1.2224,
+      "step": 5207
+    },
+    {
+      "epoch": 1.5626828719333785,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00012230757644223527,
+      "loss": 1.3069,
+      "step": 5208
+    },
+    {
+      "epoch": 1.562982969465076,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012229251393282121,
+      "loss": 1.0207,
+      "step": 5209
+    },
+    {
+      "epoch": 1.563283066996774,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00012227745142340715,
+      "loss": 1.1338,
+      "step": 5210
+    },
+    {
+      "epoch": 1.5635831645284717,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001222623889139931,
+      "loss": 1.1221,
+      "step": 5211
+    },
+    {
+      "epoch": 1.5638832620601697,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.000122247326404579,
+      "loss": 1.0835,
+      "step": 5212
+    },
+    {
+      "epoch": 1.5641833595918673,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00012223226389516492,
+      "loss": 0.951,
+      "step": 5213
+    },
+    {
+      "epoch": 1.5644834571235653,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001222172013857509,
+      "loss": 1.0912,
+      "step": 5214
+    },
+    {
+      "epoch": 1.5647835546552629,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001222021388763368,
+      "loss": 1.0198,
+      "step": 5215
+    },
+    {
+      "epoch": 1.5650836521869609,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00012218707636692271,
+      "loss": 1.0857,
+      "step": 5216
+    },
+    {
+      "epoch": 1.5653837497186585,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00012217201385750868,
+      "loss": 0.8783,
+      "step": 5217
+    },
+    {
+      "epoch": 1.5656838472503565,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001221569513480946,
+      "loss": 1.1199,
+      "step": 5218
+    },
+    {
+      "epoch": 1.565983944782054,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012214188883868054,
+      "loss": 0.9602,
+      "step": 5219
+    },
+    {
+      "epoch": 1.566284042313752,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012212682632926648,
+      "loss": 0.9599,
+      "step": 5220
+    },
+    {
+      "epoch": 1.5665841398454496,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001221117638198524,
+      "loss": 1.0217,
+      "step": 5221
+    },
+    {
+      "epoch": 1.5668842373771477,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00012209670131043833,
+      "loss": 0.8936,
+      "step": 5222
+    },
+    {
+      "epoch": 1.5671843349088452,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012208163880102424,
+      "loss": 0.8972,
+      "step": 5223
+    },
+    {
+      "epoch": 1.5674844324405433,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012206657629161018,
+      "loss": 1.0479,
+      "step": 5224
+    },
+    {
+      "epoch": 1.5677845299722408,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00012205151378219612,
+      "loss": 1.0079,
+      "step": 5225
+    },
+    {
+      "epoch": 1.5680846275039388,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012203645127278205,
+      "loss": 0.9798,
+      "step": 5226
+    },
+    {
+      "epoch": 1.5683847250356364,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012202138876336798,
+      "loss": 1.0502,
+      "step": 5227
+    },
+    {
+      "epoch": 1.5686848225673344,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012200632625395392,
+      "loss": 1.0766,
+      "step": 5228
+    },
+    {
+      "epoch": 1.5689849200990322,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00012199126374453984,
+      "loss": 1.0093,
+      "step": 5229
+    },
+    {
+      "epoch": 1.56928501763073,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012197620123512578,
+      "loss": 1.0224,
+      "step": 5230
+    },
+    {
+      "epoch": 1.5695851151624278,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00012196113872571171,
+      "loss": 1.0255,
+      "step": 5231
+    },
+    {
+      "epoch": 1.5698852126941256,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012194607621629764,
+      "loss": 1.1113,
+      "step": 5232
+    },
+    {
+      "epoch": 1.5701853102258234,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012193101370688358,
+      "loss": 1.0041,
+      "step": 5233
+    },
+    {
+      "epoch": 1.5704854077575212,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001219159511974695,
+      "loss": 1.0487,
+      "step": 5234
+    },
+    {
+      "epoch": 1.570785505289219,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00012190088868805543,
+      "loss": 0.9243,
+      "step": 5235
+    },
+    {
+      "epoch": 1.5710856028209168,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012188582617864137,
+      "loss": 1.0501,
+      "step": 5236
+    },
+    {
+      "epoch": 1.5713857003526146,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001218707636692273,
+      "loss": 1.1351,
+      "step": 5237
+    },
+    {
+      "epoch": 1.5716857978843124,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00012185570115981324,
+      "loss": 1.0674,
+      "step": 5238
+    },
+    {
+      "epoch": 1.5719858954160102,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012184063865039917,
+      "loss": 1.0787,
+      "step": 5239
+    },
+    {
+      "epoch": 1.572285992947708,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012182557614098509,
+      "loss": 1.195,
+      "step": 5240
+    },
+    {
+      "epoch": 1.5725860904794058,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012181051363157103,
+      "loss": 1.0863,
+      "step": 5241
+    },
+    {
+      "epoch": 1.5728861880111036,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012179545112215696,
+      "loss": 1.1295,
+      "step": 5242
+    },
+    {
+      "epoch": 1.5731862855428014,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00012178038861274289,
+      "loss": 0.918,
+      "step": 5243
+    },
+    {
+      "epoch": 1.5734863830744992,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012176532610332883,
+      "loss": 0.9657,
+      "step": 5244
+    },
+    {
+      "epoch": 1.573786480606197,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012175026359391475,
+      "loss": 0.941,
+      "step": 5245
+    },
+    {
+      "epoch": 1.5740865781378948,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001217352010845007,
+      "loss": 0.9955,
+      "step": 5246
+    },
+    {
+      "epoch": 1.5743866756695928,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00012172013857508662,
+      "loss": 1.1136,
+      "step": 5247
+    },
+    {
+      "epoch": 1.5746867732012904,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012170507606567255,
+      "loss": 1.0118,
+      "step": 5248
+    },
+    {
+      "epoch": 1.5749868707329884,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012169001355625849,
+      "loss": 0.8945,
+      "step": 5249
+    },
+    {
+      "epoch": 1.575286968264686,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012167495104684441,
+      "loss": 1.0329,
+      "step": 5250
+    },
+    {
+      "epoch": 1.575587065796384,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00012165988853743033,
+      "loss": 0.9707,
+      "step": 5251
+    },
+    {
+      "epoch": 1.5758871633280815,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00012164482602801628,
+      "loss": 1.187,
+      "step": 5252
+    },
+    {
+      "epoch": 1.5761872608597796,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00012162976351860221,
+      "loss": 1.0577,
+      "step": 5253
+    },
+    {
+      "epoch": 1.5764873583914771,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012161470100918812,
+      "loss": 1.0935,
+      "step": 5254
+    },
+    {
+      "epoch": 1.5767874559231752,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012159963849977408,
+      "loss": 0.9469,
+      "step": 5255
+    },
+    {
+      "epoch": 1.5770875534548727,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00012158457599035999,
+      "loss": 1.0808,
+      "step": 5256
+    },
+    {
+      "epoch": 1.5773876509865707,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00012156951348094594,
+      "loss": 1.1997,
+      "step": 5257
+    },
+    {
+      "epoch": 1.5776877485182683,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012155445097153187,
+      "loss": 1.056,
+      "step": 5258
+    },
+    {
+      "epoch": 1.5779878460499663,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012153938846211778,
+      "loss": 1.0824,
+      "step": 5259
+    },
+    {
+      "epoch": 1.578287943581664,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00012152432595270374,
+      "loss": 1.152,
+      "step": 5260
+    },
+    {
+      "epoch": 1.578588041113362,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012150926344328966,
+      "loss": 1.0537,
+      "step": 5261
+    },
+    {
+      "epoch": 1.5788881386450595,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012149420093387558,
+      "loss": 1.1131,
+      "step": 5262
+    },
+    {
+      "epoch": 1.5791882361767575,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012147913842446153,
+      "loss": 1.0679,
+      "step": 5263
+    },
+    {
+      "epoch": 1.579488333708455,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012146407591504744,
+      "loss": 0.9533,
+      "step": 5264
+    },
+    {
+      "epoch": 1.5797884312401531,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001214490134056334,
+      "loss": 1.1093,
+      "step": 5265
+    },
+    {
+      "epoch": 1.5800885287718507,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012143395089621932,
+      "loss": 1.0228,
+      "step": 5266
+    },
+    {
+      "epoch": 1.5803886263035487,
+      "grad_norm": 1.8046875,
+      "learning_rate": 0.00012141888838680524,
+      "loss": 1.0928,
+      "step": 5267
+    },
+    {
+      "epoch": 1.5806887238352465,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012140382587739119,
+      "loss": 1.1198,
+      "step": 5268
+    },
+    {
+      "epoch": 1.5809888213669443,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001213887633679771,
+      "loss": 1.0117,
+      "step": 5269
+    },
+    {
+      "epoch": 1.581288918898642,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00012137370085856303,
+      "loss": 0.9741,
+      "step": 5270
+    },
+    {
+      "epoch": 1.58158901643034,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012135863834914898,
+      "loss": 1.1916,
+      "step": 5271
+    },
+    {
+      "epoch": 1.5818891139620377,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001213435758397349,
+      "loss": 1.1481,
+      "step": 5272
+    },
+    {
+      "epoch": 1.5821892114937355,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012132851333032085,
+      "loss": 1.2348,
+      "step": 5273
+    },
+    {
+      "epoch": 1.5824893090254333,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00012131345082090676,
+      "loss": 1.0863,
+      "step": 5274
+    },
+    {
+      "epoch": 1.582789406557131,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012129838831149269,
+      "loss": 1.0947,
+      "step": 5275
+    },
+    {
+      "epoch": 1.5830895040888289,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00012128332580207865,
+      "loss": 0.9962,
+      "step": 5276
+    },
+    {
+      "epoch": 1.5833896016205267,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00012126826329266456,
+      "loss": 1.0112,
+      "step": 5277
+    },
+    {
+      "epoch": 1.5836896991522245,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00012125320078325049,
+      "loss": 1.0993,
+      "step": 5278
+    },
+    {
+      "epoch": 1.5839897966839223,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012123813827383644,
+      "loss": 1.0482,
+      "step": 5279
+    },
+    {
+      "epoch": 1.58428989421562,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00012122307576442235,
+      "loss": 1.0565,
+      "step": 5280
+    },
+    {
+      "epoch": 1.5845899917473178,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012120801325500828,
+      "loss": 1.0443,
+      "step": 5281
+    },
+    {
+      "epoch": 1.5848900892790156,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00012119295074559422,
+      "loss": 1.2034,
+      "step": 5282
+    },
+    {
+      "epoch": 1.5851901868107134,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012117788823618015,
+      "loss": 1.0165,
+      "step": 5283
+    },
+    {
+      "epoch": 1.5854902843424112,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.0001211628257267661,
+      "loss": 1.0034,
+      "step": 5284
+    },
+    {
+      "epoch": 1.585790381874109,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00012114776321735201,
+      "loss": 1.0413,
+      "step": 5285
+    },
+    {
+      "epoch": 1.586090479405807,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012113270070793794,
+      "loss": 1.1178,
+      "step": 5286
+    },
+    {
+      "epoch": 1.5863905769375046,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00012111763819852388,
+      "loss": 0.8831,
+      "step": 5287
+    },
+    {
+      "epoch": 1.5866906744692026,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012110257568910981,
+      "loss": 0.9846,
+      "step": 5288
+    },
+    {
+      "epoch": 1.5869907720009002,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012108751317969573,
+      "loss": 1.0805,
+      "step": 5289
+    },
+    {
+      "epoch": 1.5872908695325982,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012107245067028167,
+      "loss": 1.0914,
+      "step": 5290
+    },
+    {
+      "epoch": 1.5875909670642958,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001210573881608676,
+      "loss": 1.0535,
+      "step": 5291
+    },
+    {
+      "epoch": 1.5878910645959938,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012104232565145354,
+      "loss": 1.0505,
+      "step": 5292
+    },
+    {
+      "epoch": 1.5881911621276914,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012102726314203947,
+      "loss": 1.1168,
+      "step": 5293
+    },
+    {
+      "epoch": 1.5884912596593894,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001210122006326254,
+      "loss": 1.0444,
+      "step": 5294
+    },
+    {
+      "epoch": 1.588791357191087,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00012099713812321133,
+      "loss": 1.1495,
+      "step": 5295
+    },
+    {
+      "epoch": 1.589091454722785,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012098207561379726,
+      "loss": 1.1714,
+      "step": 5296
+    },
+    {
+      "epoch": 1.5893915522544826,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012096701310438319,
+      "loss": 0.9193,
+      "step": 5297
+    },
+    {
+      "epoch": 1.5896916497861806,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012095195059496913,
+      "loss": 1.1136,
+      "step": 5298
+    },
+    {
+      "epoch": 1.5899917473178782,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012093688808555506,
+      "loss": 1.0738,
+      "step": 5299
+    },
+    {
+      "epoch": 1.5902918448495762,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.000120921825576141,
+      "loss": 1.2235,
+      "step": 5300
+    },
+    {
+      "epoch": 1.5905919423812738,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00012090676306672692,
+      "loss": 0.9575,
+      "step": 5301
+    },
+    {
+      "epoch": 1.5908920399129718,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012089170055731285,
+      "loss": 0.9311,
+      "step": 5302
+    },
+    {
+      "epoch": 1.5911921374446694,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012087663804789879,
+      "loss": 1.0533,
+      "step": 5303
+    },
+    {
+      "epoch": 1.5914922349763674,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00012086157553848472,
+      "loss": 1.0737,
+      "step": 5304
+    },
+    {
+      "epoch": 1.591792332508065,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012084651302907064,
+      "loss": 1.1245,
+      "step": 5305
+    },
+    {
+      "epoch": 1.592092430039763,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012083145051965658,
+      "loss": 0.9474,
+      "step": 5306
+    },
+    {
+      "epoch": 1.5923925275714608,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012081638801024251,
+      "loss": 1.1955,
+      "step": 5307
+    },
+    {
+      "epoch": 1.5926926251031586,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00012080132550082844,
+      "loss": 0.9958,
+      "step": 5308
+    },
+    {
+      "epoch": 1.5929927226348564,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012078626299141438,
+      "loss": 1.1492,
+      "step": 5309
+    },
+    {
+      "epoch": 1.5932928201665542,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001207712004820003,
+      "loss": 1.1771,
+      "step": 5310
+    },
+    {
+      "epoch": 1.593592917698252,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012075613797258624,
+      "loss": 1.0555,
+      "step": 5311
+    },
+    {
+      "epoch": 1.5938930152299498,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00012074107546317217,
+      "loss": 0.9328,
+      "step": 5312
+    },
+    {
+      "epoch": 1.5941931127616475,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001207260129537581,
+      "loss": 1.0488,
+      "step": 5313
+    },
+    {
+      "epoch": 1.5944932102933453,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012071095044434404,
+      "loss": 1.0027,
+      "step": 5314
+    },
+    {
+      "epoch": 1.5947933078250431,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012069588793492996,
+      "loss": 0.9649,
+      "step": 5315
+    },
+    {
+      "epoch": 1.595093405356741,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012068082542551589,
+      "loss": 1.0259,
+      "step": 5316
+    },
+    {
+      "epoch": 1.5953935028884387,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012066576291610183,
+      "loss": 0.9036,
+      "step": 5317
+    },
+    {
+      "epoch": 1.5956936004201365,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012065070040668776,
+      "loss": 1.0605,
+      "step": 5318
+    },
+    {
+      "epoch": 1.5959936979518343,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001206356378972737,
+      "loss": 1.001,
+      "step": 5319
+    },
+    {
+      "epoch": 1.5962937954835321,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012062057538785963,
+      "loss": 1.1092,
+      "step": 5320
+    },
+    {
+      "epoch": 1.59659389301523,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012060551287844555,
+      "loss": 1.1036,
+      "step": 5321
+    },
+    {
+      "epoch": 1.5968939905469277,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012059045036903149,
+      "loss": 0.9339,
+      "step": 5322
+    },
+    {
+      "epoch": 1.5971940880786255,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012057538785961742,
+      "loss": 0.9815,
+      "step": 5323
+    },
+    {
+      "epoch": 1.5974941856103233,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00012056032535020335,
+      "loss": 1.2858,
+      "step": 5324
+    },
+    {
+      "epoch": 1.5977942831420213,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012054526284078929,
+      "loss": 1.0747,
+      "step": 5325
+    },
+    {
+      "epoch": 1.598094380673719,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00012053020033137521,
+      "loss": 1.1103,
+      "step": 5326
+    },
+    {
+      "epoch": 1.598394478205417,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00012051513782196115,
+      "loss": 1.0833,
+      "step": 5327
+    },
+    {
+      "epoch": 1.5986945757371145,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00012050007531254708,
+      "loss": 0.9725,
+      "step": 5328
+    },
+    {
+      "epoch": 1.5989946732688125,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00012048501280313301,
+      "loss": 0.8614,
+      "step": 5329
+    },
+    {
+      "epoch": 1.59929477080051,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00012046995029371895,
+      "loss": 1.0714,
+      "step": 5330
+    },
+    {
+      "epoch": 1.599594868332208,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012045488778430487,
+      "loss": 1.1188,
+      "step": 5331
+    },
+    {
+      "epoch": 1.5998949658639057,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012043982527489079,
+      "loss": 0.8423,
+      "step": 5332
+    },
+    {
+      "epoch": 1.6001950633956037,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012042476276547674,
+      "loss": 1.1846,
+      "step": 5333
+    },
+    {
+      "epoch": 1.6004951609273013,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00012040970025606267,
+      "loss": 1.0046,
+      "step": 5334
+    },
+    {
+      "epoch": 1.6007952584589993,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00012039463774664858,
+      "loss": 1.0011,
+      "step": 5335
+    },
+    {
+      "epoch": 1.6010953559906969,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00012037957523723454,
+      "loss": 1.0482,
+      "step": 5336
+    },
+    {
+      "epoch": 1.6013954535223949,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00012036451272782046,
+      "loss": 1.1025,
+      "step": 5337
+    },
+    {
+      "epoch": 1.6016955510540924,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001203494502184064,
+      "loss": 1.0992,
+      "step": 5338
+    },
+    {
+      "epoch": 1.6019956485857905,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00012033438770899233,
+      "loss": 0.9555,
+      "step": 5339
+    },
+    {
+      "epoch": 1.602295746117488,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00012031932519957824,
+      "loss": 0.98,
+      "step": 5340
+    },
+    {
+      "epoch": 1.602595843649186,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001203042626901642,
+      "loss": 1.0067,
+      "step": 5341
+    },
+    {
+      "epoch": 1.6028959411808836,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00012028920018075012,
+      "loss": 1.046,
+      "step": 5342
+    },
+    {
+      "epoch": 1.6031960387125817,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00012027413767133604,
+      "loss": 1.1024,
+      "step": 5343
+    },
+    {
+      "epoch": 1.6034961362442792,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00012025907516192199,
+      "loss": 1.0862,
+      "step": 5344
+    },
+    {
+      "epoch": 1.6037962337759772,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001202440126525079,
+      "loss": 1.1251,
+      "step": 5345
+    },
+    {
+      "epoch": 1.604096331307675,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012022895014309386,
+      "loss": 0.9962,
+      "step": 5346
+    },
+    {
+      "epoch": 1.6043964288393728,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00012021388763367978,
+      "loss": 1.1323,
+      "step": 5347
+    },
+    {
+      "epoch": 1.6046965263710706,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001201988251242657,
+      "loss": 1.1584,
+      "step": 5348
+    },
+    {
+      "epoch": 1.6049966239027684,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00012018376261485165,
+      "loss": 1.0024,
+      "step": 5349
+    },
+    {
+      "epoch": 1.6052967214344662,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00012016870010543756,
+      "loss": 1.0808,
+      "step": 5350
+    },
+    {
+      "epoch": 1.605596818966164,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00012015363759602349,
+      "loss": 0.9702,
+      "step": 5351
+    },
+    {
+      "epoch": 1.6058969164978618,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00012013857508660944,
+      "loss": 1.1139,
+      "step": 5352
+    },
+    {
+      "epoch": 1.6061970140295596,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00012012351257719536,
+      "loss": 1.0971,
+      "step": 5353
+    },
+    {
+      "epoch": 1.6064971115612574,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00012010845006778131,
+      "loss": 1.0302,
+      "step": 5354
+    },
+    {
+      "epoch": 1.6067972090929552,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00012009338755836724,
+      "loss": 1.1363,
+      "step": 5355
+    },
+    {
+      "epoch": 1.607097306624653,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00012007832504895315,
+      "loss": 1.1088,
+      "step": 5356
+    },
+    {
+      "epoch": 1.6073974041563508,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001200632625395391,
+      "loss": 0.9245,
+      "step": 5357
+    },
+    {
+      "epoch": 1.6076975016880486,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00012004820003012502,
+      "loss": 1.05,
+      "step": 5358
+    },
+    {
+      "epoch": 1.6079975992197464,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00012003313752071095,
+      "loss": 0.9933,
+      "step": 5359
+    },
+    {
+      "epoch": 1.6082976967514442,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001200180750112969,
+      "loss": 1.1693,
+      "step": 5360
+    },
+    {
+      "epoch": 1.608597794283142,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00012000301250188281,
+      "loss": 1.0804,
+      "step": 5361
+    },
+    {
+      "epoch": 1.6088978918148398,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00011998794999246874,
+      "loss": 0.9651,
+      "step": 5362
+    },
+    {
+      "epoch": 1.6091979893465376,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011997288748305468,
+      "loss": 1.0085,
+      "step": 5363
+    },
+    {
+      "epoch": 1.6094980868782356,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001199578249736406,
+      "loss": 0.9374,
+      "step": 5364
+    },
+    {
+      "epoch": 1.6097981844099332,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011994276246422656,
+      "loss": 1.0158,
+      "step": 5365
+    },
+    {
+      "epoch": 1.6100982819416312,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011992769995481247,
+      "loss": 1.1958,
+      "step": 5366
+    },
+    {
+      "epoch": 1.6103983794733288,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001199126374453984,
+      "loss": 1.0106,
+      "step": 5367
+    },
+    {
+      "epoch": 1.6106984770050268,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011989757493598435,
+      "loss": 1.1282,
+      "step": 5368
+    },
+    {
+      "epoch": 1.6109985745367243,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011988251242657027,
+      "loss": 0.9255,
+      "step": 5369
+    },
+    {
+      "epoch": 1.6112986720684224,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001198674499171562,
+      "loss": 0.9794,
+      "step": 5370
+    },
+    {
+      "epoch": 1.61159876960012,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011985238740774213,
+      "loss": 1.0128,
+      "step": 5371
+    },
+    {
+      "epoch": 1.611898867131818,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011983732489832806,
+      "loss": 0.9596,
+      "step": 5372
+    },
+    {
+      "epoch": 1.6121989646635155,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00011982226238891401,
+      "loss": 1.1339,
+      "step": 5373
+    },
+    {
+      "epoch": 1.6124990621952136,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011980719987949993,
+      "loss": 0.9484,
+      "step": 5374
+    },
+    {
+      "epoch": 1.6127991597269111,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011979213737008585,
+      "loss": 1.0844,
+      "step": 5375
+    },
+    {
+      "epoch": 1.6130992572586091,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001197770748606718,
+      "loss": 1.0575,
+      "step": 5376
+    },
+    {
+      "epoch": 1.6133993547903067,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00011976201235125772,
+      "loss": 1.0816,
+      "step": 5377
+    },
+    {
+      "epoch": 1.6136994523220047,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011974694984184365,
+      "loss": 0.9915,
+      "step": 5378
+    },
+    {
+      "epoch": 1.6139995498537023,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011973188733242959,
+      "loss": 1.0232,
+      "step": 5379
+    },
+    {
+      "epoch": 1.6142996473854003,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011971682482301552,
+      "loss": 1.2161,
+      "step": 5380
+    },
+    {
+      "epoch": 1.614599744917098,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011970176231360146,
+      "loss": 1.0175,
+      "step": 5381
+    },
+    {
+      "epoch": 1.614899842448796,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00011968669980418738,
+      "loss": 0.9889,
+      "step": 5382
+    },
+    {
+      "epoch": 1.6151999399804937,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011967163729477331,
+      "loss": 0.9956,
+      "step": 5383
+    },
+    {
+      "epoch": 1.6155000375121915,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011965657478535925,
+      "loss": 1.0757,
+      "step": 5384
+    },
+    {
+      "epoch": 1.6158001350438893,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011964151227594518,
+      "loss": 1.0005,
+      "step": 5385
+    },
+    {
+      "epoch": 1.616100232575587,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001196264497665311,
+      "loss": 1.0727,
+      "step": 5386
+    },
+    {
+      "epoch": 1.616400330107285,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011961138725711704,
+      "loss": 1.0759,
+      "step": 5387
+    },
+    {
+      "epoch": 1.6167004276389827,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011959632474770297,
+      "loss": 1.0219,
+      "step": 5388
+    },
+    {
+      "epoch": 1.6170005251706805,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001195812622382889,
+      "loss": 1.1434,
+      "step": 5389
+    },
+    {
+      "epoch": 1.6173006227023783,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011956619972887484,
+      "loss": 1.0824,
+      "step": 5390
+    },
+    {
+      "epoch": 1.617600720234076,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011955113721946076,
+      "loss": 0.9704,
+      "step": 5391
+    },
+    {
+      "epoch": 1.6179008177657739,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001195360747100467,
+      "loss": 1.148,
+      "step": 5392
+    },
+    {
+      "epoch": 1.6182009152974717,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011952101220063263,
+      "loss": 1.0096,
+      "step": 5393
+    },
+    {
+      "epoch": 1.6185010128291695,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011950594969121856,
+      "loss": 0.9491,
+      "step": 5394
+    },
+    {
+      "epoch": 1.6188011103608673,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001194908871818045,
+      "loss": 1.1401,
+      "step": 5395
+    },
+    {
+      "epoch": 1.619101207892565,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011947582467239042,
+      "loss": 1.0091,
+      "step": 5396
+    },
+    {
+      "epoch": 1.6194013054242629,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011946076216297635,
+      "loss": 0.9527,
+      "step": 5397
+    },
+    {
+      "epoch": 1.6197014029559607,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011944569965356229,
+      "loss": 0.9278,
+      "step": 5398
+    },
+    {
+      "epoch": 1.6200015004876585,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011943063714414822,
+      "loss": 1.0199,
+      "step": 5399
+    },
+    {
+      "epoch": 1.6203015980193562,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00011941557463473416,
+      "loss": 1.1208,
+      "step": 5400
+    },
+    {
+      "epoch": 1.620601695551054,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011940051212532009,
+      "loss": 0.9658,
+      "step": 5401
+    },
+    {
+      "epoch": 1.6209017930827518,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011938544961590601,
+      "loss": 1.0763,
+      "step": 5402
+    },
+    {
+      "epoch": 1.6212018906144499,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00011937038710649195,
+      "loss": 0.9923,
+      "step": 5403
+    },
+    {
+      "epoch": 1.6215019881461474,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011935532459707788,
+      "loss": 1.0312,
+      "step": 5404
+    },
+    {
+      "epoch": 1.6218020856778455,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001193402620876638,
+      "loss": 1.0674,
+      "step": 5405
+    },
+    {
+      "epoch": 1.622102183209543,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011932519957824975,
+      "loss": 1.0148,
+      "step": 5406
+    },
+    {
+      "epoch": 1.622402280741241,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011931013706883567,
+      "loss": 1.0445,
+      "step": 5407
+    },
+    {
+      "epoch": 1.6227023782729386,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011929507455942161,
+      "loss": 0.94,
+      "step": 5408
+    },
+    {
+      "epoch": 1.6230024758046366,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011928001205000754,
+      "loss": 1.0535,
+      "step": 5409
+    },
+    {
+      "epoch": 1.6233025733363342,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011926494954059347,
+      "loss": 1.0063,
+      "step": 5410
+    },
+    {
+      "epoch": 1.6236026708680322,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011924988703117941,
+      "loss": 0.9823,
+      "step": 5411
+    },
+    {
+      "epoch": 1.6239027683997298,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011923482452176533,
+      "loss": 1.1024,
+      "step": 5412
+    },
+    {
+      "epoch": 1.6242028659314278,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011921976201235126,
+      "loss": 0.9882,
+      "step": 5413
+    },
+    {
+      "epoch": 1.6245029634631254,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001192046995029372,
+      "loss": 1.042,
+      "step": 5414
+    },
+    {
+      "epoch": 1.6248030609948234,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011918963699352313,
+      "loss": 0.9919,
+      "step": 5415
+    },
+    {
+      "epoch": 1.625103158526521,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011917457448410904,
+      "loss": 1.0843,
+      "step": 5416
+    },
+    {
+      "epoch": 1.625403256058219,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.000119159511974695,
+      "loss": 1.1537,
+      "step": 5417
+    },
+    {
+      "epoch": 1.6257033535899166,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011914444946528092,
+      "loss": 1.1644,
+      "step": 5418
+    },
+    {
+      "epoch": 1.6260034511216146,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00011912938695586686,
+      "loss": 1.0443,
+      "step": 5419
+    },
+    {
+      "epoch": 1.6263035486533122,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011911432444645279,
+      "loss": 0.9495,
+      "step": 5420
+    },
+    {
+      "epoch": 1.6266036461850102,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001190992619370387,
+      "loss": 1.0111,
+      "step": 5421
+    },
+    {
+      "epoch": 1.626903743716708,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011908419942762466,
+      "loss": 1.1138,
+      "step": 5422
+    },
+    {
+      "epoch": 1.6272038412484058,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011906913691821058,
+      "loss": 1.0176,
+      "step": 5423
+    },
+    {
+      "epoch": 1.6275039387801036,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001190540744087965,
+      "loss": 1.277,
+      "step": 5424
+    },
+    {
+      "epoch": 1.6278040363118014,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011903901189938245,
+      "loss": 1.1104,
+      "step": 5425
+    },
+    {
+      "epoch": 1.6281041338434992,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011902394938996838,
+      "loss": 1.1035,
+      "step": 5426
+    },
+    {
+      "epoch": 1.628404231375197,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011900888688055432,
+      "loss": 1.0396,
+      "step": 5427
+    },
+    {
+      "epoch": 1.6287043289068948,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011899382437114024,
+      "loss": 1.1552,
+      "step": 5428
+    },
+    {
+      "epoch": 1.6290044264385926,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011897876186172616,
+      "loss": 1.11,
+      "step": 5429
+    },
+    {
+      "epoch": 1.6293045239702904,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011896369935231211,
+      "loss": 1.0599,
+      "step": 5430
+    },
+    {
+      "epoch": 1.6296046215019881,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011894863684289804,
+      "loss": 1.0608,
+      "step": 5431
+    },
+    {
+      "epoch": 1.629904719033686,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011893357433348395,
+      "loss": 0.9532,
+      "step": 5432
+    },
+    {
+      "epoch": 1.6302048165653837,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001189185118240699,
+      "loss": 1.0564,
+      "step": 5433
+    },
+    {
+      "epoch": 1.6305049140970815,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011890344931465582,
+      "loss": 0.9162,
+      "step": 5434
+    },
+    {
+      "epoch": 1.6308050116287793,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011888838680524177,
+      "loss": 1.0222,
+      "step": 5435
+    },
+    {
+      "epoch": 1.6311051091604771,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001188733242958277,
+      "loss": 0.9522,
+      "step": 5436
+    },
+    {
+      "epoch": 1.631405206692175,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011885826178641361,
+      "loss": 0.9674,
+      "step": 5437
+    },
+    {
+      "epoch": 1.6317053042238727,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00011884319927699957,
+      "loss": 0.8111,
+      "step": 5438
+    },
+    {
+      "epoch": 1.6320054017555705,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011882813676758548,
+      "loss": 0.9606,
+      "step": 5439
+    },
+    {
+      "epoch": 1.6323054992872683,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001188130742581714,
+      "loss": 1.0682,
+      "step": 5440
+    },
+    {
+      "epoch": 1.632605596818966,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011879801174875736,
+      "loss": 1.0967,
+      "step": 5441
+    },
+    {
+      "epoch": 1.6329056943506641,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011878294923934327,
+      "loss": 1.0793,
+      "step": 5442
+    },
+    {
+      "epoch": 1.6332057918823617,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001187678867299292,
+      "loss": 1.0546,
+      "step": 5443
+    },
+    {
+      "epoch": 1.6335058894140597,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011875282422051515,
+      "loss": 1.0908,
+      "step": 5444
+    },
+    {
+      "epoch": 1.6338059869457573,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011873776171110107,
+      "loss": 1.0453,
+      "step": 5445
+    },
+    {
+      "epoch": 1.6341060844774553,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011872269920168702,
+      "loss": 1.0398,
+      "step": 5446
+    },
+    {
+      "epoch": 1.6344061820091529,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011870763669227293,
+      "loss": 0.9,
+      "step": 5447
+    },
+    {
+      "epoch": 1.634706279540851,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011869257418285886,
+      "loss": 1.0391,
+      "step": 5448
+    },
+    {
+      "epoch": 1.6350063770725485,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011867751167344481,
+      "loss": 0.948,
+      "step": 5449
+    },
+    {
+      "epoch": 1.6353064746042465,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011866244916403073,
+      "loss": 0.9743,
+      "step": 5450
+    },
+    {
+      "epoch": 1.635606572135944,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011864738665461665,
+      "loss": 1.0527,
+      "step": 5451
+    },
+    {
+      "epoch": 1.635906669667642,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001186323241452026,
+      "loss": 1.1232,
+      "step": 5452
+    },
+    {
+      "epoch": 1.6362067671993397,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00011861726163578852,
+      "loss": 0.9827,
+      "step": 5453
+    },
+    {
+      "epoch": 1.6365068647310377,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011860219912637447,
+      "loss": 1.1423,
+      "step": 5454
+    },
+    {
+      "epoch": 1.6368069622627353,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011858713661696039,
+      "loss": 1.07,
+      "step": 5455
+    },
+    {
+      "epoch": 1.6371070597944333,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011857207410754631,
+      "loss": 1.1,
+      "step": 5456
+    },
+    {
+      "epoch": 1.6374071573261308,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00011855701159813225,
+      "loss": 0.9399,
+      "step": 5457
+    },
+    {
+      "epoch": 1.6377072548578289,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011854194908871818,
+      "loss": 1.0064,
+      "step": 5458
+    },
+    {
+      "epoch": 1.6380073523895264,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011852688657930411,
+      "loss": 1.0417,
+      "step": 5459
+    },
+    {
+      "epoch": 1.6383074499212245,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011851182406989005,
+      "loss": 1.0747,
+      "step": 5460
+    },
+    {
+      "epoch": 1.6386075474529223,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011849676156047598,
+      "loss": 0.9535,
+      "step": 5461
+    },
+    {
+      "epoch": 1.63890764498462,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011848169905106193,
+      "loss": 1.1841,
+      "step": 5462
+    },
+    {
+      "epoch": 1.6392077425163178,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011846663654164784,
+      "loss": 0.9939,
+      "step": 5463
+    },
+    {
+      "epoch": 1.6395078400480156,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011845157403223377,
+      "loss": 1.0701,
+      "step": 5464
+    },
+    {
+      "epoch": 1.6398079375797134,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011843651152281971,
+      "loss": 0.9841,
+      "step": 5465
+    },
+    {
+      "epoch": 1.6401080351114112,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011842144901340564,
+      "loss": 0.9968,
+      "step": 5466
+    },
+    {
+      "epoch": 1.640408132643109,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011840638650399156,
+      "loss": 1.003,
+      "step": 5467
+    },
+    {
+      "epoch": 1.6407082301748068,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001183913239945775,
+      "loss": 0.9897,
+      "step": 5468
+    },
+    {
+      "epoch": 1.6410083277065046,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011837626148516343,
+      "loss": 1.1211,
+      "step": 5469
+    },
+    {
+      "epoch": 1.6413084252382024,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00011836119897574936,
+      "loss": 0.871,
+      "step": 5470
+    },
+    {
+      "epoch": 1.6416085227699002,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001183461364663353,
+      "loss": 1.1325,
+      "step": 5471
+    },
+    {
+      "epoch": 1.641908620301598,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011833107395692122,
+      "loss": 0.9866,
+      "step": 5472
+    },
+    {
+      "epoch": 1.6422087178332958,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011831601144750716,
+      "loss": 1.0066,
+      "step": 5473
+    },
+    {
+      "epoch": 1.6425088153649936,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011830094893809309,
+      "loss": 1.0278,
+      "step": 5474
+    },
+    {
+      "epoch": 1.6428089128966914,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00011828588642867902,
+      "loss": 0.8017,
+      "step": 5475
+    },
+    {
+      "epoch": 1.6431090104283892,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011827082391926496,
+      "loss": 1.0268,
+      "step": 5476
+    },
+    {
+      "epoch": 1.643409107960087,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011825576140985088,
+      "loss": 0.9003,
+      "step": 5477
+    },
+    {
+      "epoch": 1.6437092054917848,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011824069890043681,
+      "loss": 1.0874,
+      "step": 5478
+    },
+    {
+      "epoch": 1.6440093030234826,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00011822563639102275,
+      "loss": 1.0225,
+      "step": 5479
+    },
+    {
+      "epoch": 1.6443094005551804,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011821057388160868,
+      "loss": 0.9871,
+      "step": 5480
+    },
+    {
+      "epoch": 1.6446094980868784,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011819551137219462,
+      "loss": 1.1709,
+      "step": 5481
+    },
+    {
+      "epoch": 1.644909595618576,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011818044886278055,
+      "loss": 0.977,
+      "step": 5482
+    },
+    {
+      "epoch": 1.645209693150274,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011816538635336647,
+      "loss": 1.0864,
+      "step": 5483
+    },
+    {
+      "epoch": 1.6455097906819716,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011815032384395241,
+      "loss": 1.0913,
+      "step": 5484
+    },
+    {
+      "epoch": 1.6458098882136696,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00011813526133453834,
+      "loss": 0.9772,
+      "step": 5485
+    },
+    {
+      "epoch": 1.6461099857453672,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011812019882512427,
+      "loss": 1.0599,
+      "step": 5486
+    },
+    {
+      "epoch": 1.6464100832770652,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001181051363157102,
+      "loss": 1.0945,
+      "step": 5487
+    },
+    {
+      "epoch": 1.6467101808087627,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011809007380629613,
+      "loss": 1.2087,
+      "step": 5488
+    },
+    {
+      "epoch": 1.6470102783404608,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00011807501129688207,
+      "loss": 1.0086,
+      "step": 5489
+    },
+    {
+      "epoch": 1.6473103758721583,
+      "grad_norm": 1.0,
+      "learning_rate": 0.000118059948787468,
+      "loss": 1.0374,
+      "step": 5490
+    },
+    {
+      "epoch": 1.6476104734038564,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011804488627805393,
+      "loss": 1.0587,
+      "step": 5491
+    },
+    {
+      "epoch": 1.647910570935554,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00011802982376863987,
+      "loss": 0.9637,
+      "step": 5492
+    },
+    {
+      "epoch": 1.648210668467252,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001180147612592258,
+      "loss": 0.893,
+      "step": 5493
+    },
+    {
+      "epoch": 1.6485107659989495,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011799969874981172,
+      "loss": 1.036,
+      "step": 5494
+    },
+    {
+      "epoch": 1.6488108635306475,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00011798463624039766,
+      "loss": 1.1868,
+      "step": 5495
+    },
+    {
+      "epoch": 1.6491109610623451,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011796957373098359,
+      "loss": 1.0221,
+      "step": 5496
+    },
+    {
+      "epoch": 1.6494110585940431,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011795451122156951,
+      "loss": 0.9717,
+      "step": 5497
+    },
+    {
+      "epoch": 1.6497111561257407,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00011793944871215546,
+      "loss": 1.1067,
+      "step": 5498
+    },
+    {
+      "epoch": 1.6500112536574387,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011792438620274138,
+      "loss": 1.0001,
+      "step": 5499
+    },
+    {
+      "epoch": 1.6503113511891365,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011790932369332732,
+      "loss": 1.0409,
+      "step": 5500
+    },
+    {
+      "epoch": 1.6506114487208343,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011789426118391325,
+      "loss": 1.076,
+      "step": 5501
+    },
+    {
+      "epoch": 1.6509115462525321,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011787919867449918,
+      "loss": 1.0047,
+      "step": 5502
+    },
+    {
+      "epoch": 1.65121164378423,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011786413616508512,
+      "loss": 0.9132,
+      "step": 5503
+    },
+    {
+      "epoch": 1.6515117413159277,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011784907365567104,
+      "loss": 1.0849,
+      "step": 5504
+    },
+    {
+      "epoch": 1.6518118388476255,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011783401114625696,
+      "loss": 1.0103,
+      "step": 5505
+    },
+    {
+      "epoch": 1.6521119363793233,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011781894863684291,
+      "loss": 0.9695,
+      "step": 5506
+    },
+    {
+      "epoch": 1.652412033911021,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011780388612742884,
+      "loss": 1.1087,
+      "step": 5507
+    },
+    {
+      "epoch": 1.652712131442719,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011778882361801478,
+      "loss": 1.0483,
+      "step": 5508
+    },
+    {
+      "epoch": 1.6530122289744167,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001177737611086007,
+      "loss": 0.9686,
+      "step": 5509
+    },
+    {
+      "epoch": 1.6533123265061145,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011775869859918662,
+      "loss": 1.1382,
+      "step": 5510
+    },
+    {
+      "epoch": 1.6536124240378123,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011774363608977257,
+      "loss": 1.0497,
+      "step": 5511
+    },
+    {
+      "epoch": 1.65391252156951,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001177285735803585,
+      "loss": 1.136,
+      "step": 5512
+    },
+    {
+      "epoch": 1.6542126191012079,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011771351107094441,
+      "loss": 0.8739,
+      "step": 5513
+    },
+    {
+      "epoch": 1.6545127166329057,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011769844856153036,
+      "loss": 0.9347,
+      "step": 5514
+    },
+    {
+      "epoch": 1.6548128141646035,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011768338605211629,
+      "loss": 1.0917,
+      "step": 5515
+    },
+    {
+      "epoch": 1.6551129116963013,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011766832354270223,
+      "loss": 0.9241,
+      "step": 5516
+    },
+    {
+      "epoch": 1.655413009227999,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011765326103328816,
+      "loss": 1.0383,
+      "step": 5517
+    },
+    {
+      "epoch": 1.6557131067596969,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00011763819852387407,
+      "loss": 1.0589,
+      "step": 5518
+    },
+    {
+      "epoch": 1.6560132042913946,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00011762313601446003,
+      "loss": 0.9527,
+      "step": 5519
+    },
+    {
+      "epoch": 1.6563133018230927,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011760807350504595,
+      "loss": 0.9775,
+      "step": 5520
+    },
+    {
+      "epoch": 1.6566133993547902,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011759301099563187,
+      "loss": 1.1667,
+      "step": 5521
+    },
+    {
+      "epoch": 1.6569134968864883,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011757794848621782,
+      "loss": 1.0756,
+      "step": 5522
+    },
+    {
+      "epoch": 1.6572135944181858,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011756288597680373,
+      "loss": 0.9303,
+      "step": 5523
+    },
+    {
+      "epoch": 1.6575136919498838,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011754782346738966,
+      "loss": 0.9819,
+      "step": 5524
+    },
+    {
+      "epoch": 1.6578137894815814,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011753276095797561,
+      "loss": 1.054,
+      "step": 5525
+    },
+    {
+      "epoch": 1.6581138870132794,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011751769844856153,
+      "loss": 1.0493,
+      "step": 5526
+    },
+    {
+      "epoch": 1.658413984544977,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00011750263593914748,
+      "loss": 0.9048,
+      "step": 5527
+    },
+    {
+      "epoch": 1.658714082076675,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011748757342973339,
+      "loss": 0.9494,
+      "step": 5528
+    },
+    {
+      "epoch": 1.6590141796083726,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011747251092031932,
+      "loss": 1.0415,
+      "step": 5529
+    },
+    {
+      "epoch": 1.6593142771400706,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011745744841090527,
+      "loss": 1.0324,
+      "step": 5530
+    },
+    {
+      "epoch": 1.6596143746717682,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011744238590149119,
+      "loss": 0.9385,
+      "step": 5531
+    },
+    {
+      "epoch": 1.6599144722034662,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011742732339207711,
+      "loss": 1.0688,
+      "step": 5532
+    },
+    {
+      "epoch": 1.6602145697351638,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00011741226088266307,
+      "loss": 0.887,
+      "step": 5533
+    },
+    {
+      "epoch": 1.6605146672668618,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011739719837324898,
+      "loss": 0.9345,
+      "step": 5534
+    },
+    {
+      "epoch": 1.6608147647985594,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011738213586383493,
+      "loss": 0.975,
+      "step": 5535
+    },
+    {
+      "epoch": 1.6611148623302574,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011736707335442085,
+      "loss": 1.0083,
+      "step": 5536
+    },
+    {
+      "epoch": 1.661414959861955,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011735201084500677,
+      "loss": 1.0249,
+      "step": 5537
+    },
+    {
+      "epoch": 1.661715057393653,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011733694833559273,
+      "loss": 1.0027,
+      "step": 5538
+    },
+    {
+      "epoch": 1.6620151549253508,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011732188582617864,
+      "loss": 1.0525,
+      "step": 5539
+    },
+    {
+      "epoch": 1.6623152524570486,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011730682331676457,
+      "loss": 1.0334,
+      "step": 5540
+    },
+    {
+      "epoch": 1.6626153499887464,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011729176080735051,
+      "loss": 1.1045,
+      "step": 5541
+    },
+    {
+      "epoch": 1.6629154475204442,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011727669829793644,
+      "loss": 1.0182,
+      "step": 5542
+    },
+    {
+      "epoch": 1.663215545052142,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011726163578852239,
+      "loss": 1.0938,
+      "step": 5543
+    },
+    {
+      "epoch": 1.6635156425838398,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001172465732791083,
+      "loss": 1.098,
+      "step": 5544
+    },
+    {
+      "epoch": 1.6638157401155376,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011723151076969423,
+      "loss": 0.8947,
+      "step": 5545
+    },
+    {
+      "epoch": 1.6641158376472354,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011721644826028017,
+      "loss": 1.058,
+      "step": 5546
+    },
+    {
+      "epoch": 1.6644159351789332,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001172013857508661,
+      "loss": 1.0463,
+      "step": 5547
+    },
+    {
+      "epoch": 1.664716032710631,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011718632324145202,
+      "loss": 0.945,
+      "step": 5548
+    },
+    {
+      "epoch": 1.6650161302423288,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00011717126073203796,
+      "loss": 1.0344,
+      "step": 5549
+    },
+    {
+      "epoch": 1.6653162277740265,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00011715619822262389,
+      "loss": 0.8818,
+      "step": 5550
+    },
+    {
+      "epoch": 1.6656163253057243,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011714113571320982,
+      "loss": 0.9849,
+      "step": 5551
+    },
+    {
+      "epoch": 1.6659164228374221,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011712607320379576,
+      "loss": 0.9982,
+      "step": 5552
+    },
+    {
+      "epoch": 1.66621652036912,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00011711101069438168,
+      "loss": 1.276,
+      "step": 5553
+    },
+    {
+      "epoch": 1.6665166179008177,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011709594818496762,
+      "loss": 1.014,
+      "step": 5554
+    },
+    {
+      "epoch": 1.6668167154325155,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011708088567555355,
+      "loss": 1.0284,
+      "step": 5555
+    },
+    {
+      "epoch": 1.6671168129642133,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011706582316613948,
+      "loss": 1.0737,
+      "step": 5556
+    },
+    {
+      "epoch": 1.6674169104959111,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011705076065672542,
+      "loss": 0.9853,
+      "step": 5557
+    },
+    {
+      "epoch": 1.667717008027609,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011703569814731134,
+      "loss": 1.0088,
+      "step": 5558
+    },
+    {
+      "epoch": 1.668017105559307,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011702063563789727,
+      "loss": 1.0305,
+      "step": 5559
+    },
+    {
+      "epoch": 1.6683172030910045,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00011700557312848321,
+      "loss": 1.0892,
+      "step": 5560
+    },
+    {
+      "epoch": 1.6686173006227025,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011699051061906914,
+      "loss": 1.0834,
+      "step": 5561
+    },
+    {
+      "epoch": 1.6689173981544,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011697544810965508,
+      "loss": 1.1467,
+      "step": 5562
+    },
+    {
+      "epoch": 1.6692174956860981,
+      "grad_norm": 1.0,
+      "learning_rate": 0.000116960385600241,
+      "loss": 1.1146,
+      "step": 5563
+    },
+    {
+      "epoch": 1.6695175932177957,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011694532309082693,
+      "loss": 1.0224,
+      "step": 5564
+    },
+    {
+      "epoch": 1.6698176907494937,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011693026058141287,
+      "loss": 1.0532,
+      "step": 5565
+    },
+    {
+      "epoch": 1.6701177882811913,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001169151980719988,
+      "loss": 1.1647,
+      "step": 5566
+    },
+    {
+      "epoch": 1.6704178858128893,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011690013556258473,
+      "loss": 1.0533,
+      "step": 5567
+    },
+    {
+      "epoch": 1.6707179833445869,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00011688507305317067,
+      "loss": 1.0169,
+      "step": 5568
+    },
+    {
+      "epoch": 1.671018080876285,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.0001168700105437566,
+      "loss": 1.0465,
+      "step": 5569
+    },
+    {
+      "epoch": 1.6713181784079825,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011685494803434253,
+      "loss": 1.0424,
+      "step": 5570
+    },
+    {
+      "epoch": 1.6716182759396805,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011683988552492846,
+      "loss": 1.0492,
+      "step": 5571
+    },
+    {
+      "epoch": 1.671918373471378,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00011682482301551439,
+      "loss": 1.1546,
+      "step": 5572
+    },
+    {
+      "epoch": 1.672218471003076,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00011680976050610033,
+      "loss": 0.9894,
+      "step": 5573
+    },
+    {
+      "epoch": 1.6725185685347737,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011679469799668625,
+      "loss": 0.973,
+      "step": 5574
+    },
+    {
+      "epoch": 1.6728186660664717,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011677963548727218,
+      "loss": 0.994,
+      "step": 5575
+    },
+    {
+      "epoch": 1.6731187635981692,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011676457297785812,
+      "loss": 1.0345,
+      "step": 5576
+    },
+    {
+      "epoch": 1.6734188611298673,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011674951046844405,
+      "loss": 1.0353,
+      "step": 5577
+    },
+    {
+      "epoch": 1.673718958661565,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011673444795902997,
+      "loss": 1.1621,
+      "step": 5578
+    },
+    {
+      "epoch": 1.6740190561932629,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011671938544961592,
+      "loss": 0.8957,
+      "step": 5579
+    },
+    {
+      "epoch": 1.6743191537249607,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011670432294020184,
+      "loss": 0.9932,
+      "step": 5580
+    },
+    {
+      "epoch": 1.6746192512566584,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011668926043078778,
+      "loss": 0.9625,
+      "step": 5581
+    },
+    {
+      "epoch": 1.6749193487883562,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011667419792137371,
+      "loss": 1.0694,
+      "step": 5582
+    },
+    {
+      "epoch": 1.675219446320054,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00011665913541195964,
+      "loss": 1.0759,
+      "step": 5583
+    },
+    {
+      "epoch": 1.6755195438517518,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011664407290254558,
+      "loss": 1.0541,
+      "step": 5584
+    },
+    {
+      "epoch": 1.6758196413834496,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001166290103931315,
+      "loss": 0.9495,
+      "step": 5585
+    },
+    {
+      "epoch": 1.6761197389151474,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011661394788371742,
+      "loss": 1.0205,
+      "step": 5586
+    },
+    {
+      "epoch": 1.6764198364468452,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011659888537430337,
+      "loss": 1.0113,
+      "step": 5587
+    },
+    {
+      "epoch": 1.676719933978543,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001165838228648893,
+      "loss": 1.0477,
+      "step": 5588
+    },
+    {
+      "epoch": 1.6770200315102408,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00011656876035547524,
+      "loss": 0.9408,
+      "step": 5589
+    },
+    {
+      "epoch": 1.6773201290419386,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011655369784606116,
+      "loss": 1.0378,
+      "step": 5590
+    },
+    {
+      "epoch": 1.6776202265736364,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011653863533664709,
+      "loss": 1.1593,
+      "step": 5591
+    },
+    {
+      "epoch": 1.6779203241053342,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011652357282723303,
+      "loss": 1.0849,
+      "step": 5592
+    },
+    {
+      "epoch": 1.678220421637032,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011650851031781896,
+      "loss": 1.0363,
+      "step": 5593
+    },
+    {
+      "epoch": 1.6785205191687298,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00011649344780840487,
+      "loss": 0.8379,
+      "step": 5594
+    },
+    {
+      "epoch": 1.6788206167004276,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011647838529899082,
+      "loss": 0.961,
+      "step": 5595
+    },
+    {
+      "epoch": 1.6791207142321254,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00011646332278957675,
+      "loss": 0.9115,
+      "step": 5596
+    },
+    {
+      "epoch": 1.6794208117638232,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011644826028016269,
+      "loss": 0.9621,
+      "step": 5597
+    },
+    {
+      "epoch": 1.6797209092955212,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011643319777074862,
+      "loss": 1.0273,
+      "step": 5598
+    },
+    {
+      "epoch": 1.6800210068272188,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011641813526133453,
+      "loss": 0.9837,
+      "step": 5599
+    },
+    {
+      "epoch": 1.6803211043589168,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011640307275192049,
+      "loss": 0.8675,
+      "step": 5600
+    },
+    {
+      "epoch": 1.6806212018906144,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011638801024250641,
+      "loss": 0.9082,
+      "step": 5601
+    },
+    {
+      "epoch": 1.6809212994223124,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011637294773309233,
+      "loss": 1.0657,
+      "step": 5602
+    },
+    {
+      "epoch": 1.68122139695401,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011635788522367828,
+      "loss": 0.8889,
+      "step": 5603
+    },
+    {
+      "epoch": 1.681521494485708,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001163428227142642,
+      "loss": 0.9583,
+      "step": 5604
+    },
+    {
+      "epoch": 1.6818215920174056,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011632776020485015,
+      "loss": 1.1142,
+      "step": 5605
+    },
+    {
+      "epoch": 1.6821216895491036,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011631269769543607,
+      "loss": 1.0385,
+      "step": 5606
+    },
+    {
+      "epoch": 1.6824217870808011,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00011629763518602199,
+      "loss": 1.1073,
+      "step": 5607
+    },
+    {
+      "epoch": 1.6827218846124992,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011628257267660794,
+      "loss": 1.0515,
+      "step": 5608
+    },
+    {
+      "epoch": 1.6830219821441967,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011626751016719387,
+      "loss": 1.0954,
+      "step": 5609
+    },
+    {
+      "epoch": 1.6833220796758948,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011625244765777978,
+      "loss": 1.0447,
+      "step": 5610
+    },
+    {
+      "epoch": 1.6836221772075923,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011623738514836573,
+      "loss": 1.1106,
+      "step": 5611
+    },
+    {
+      "epoch": 1.6839222747392903,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011622232263895165,
+      "loss": 1.0819,
+      "step": 5612
+    },
+    {
+      "epoch": 1.684222372270988,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011620726012953757,
+      "loss": 1.162,
+      "step": 5613
+    },
+    {
+      "epoch": 1.684522469802686,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00011619219762012353,
+      "loss": 1.071,
+      "step": 5614
+    },
+    {
+      "epoch": 1.6848225673343835,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011617713511070944,
+      "loss": 1.1506,
+      "step": 5615
+    },
+    {
+      "epoch": 1.6851226648660815,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.0001161620726012954,
+      "loss": 0.9622,
+      "step": 5616
+    },
+    {
+      "epoch": 1.6854227623977793,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00011614701009188131,
+      "loss": 0.8561,
+      "step": 5617
+    },
+    {
+      "epoch": 1.6857228599294771,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011613194758246723,
+      "loss": 1.1936,
+      "step": 5618
+    },
+    {
+      "epoch": 1.686022957461175,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011611688507305319,
+      "loss": 0.974,
+      "step": 5619
+    },
+    {
+      "epoch": 1.6863230549928727,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001161018225636391,
+      "loss": 1.091,
+      "step": 5620
+    },
+    {
+      "epoch": 1.6866231525245705,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011608676005422503,
+      "loss": 1.0244,
+      "step": 5621
+    },
+    {
+      "epoch": 1.6869232500562683,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011607169754481098,
+      "loss": 0.9765,
+      "step": 5622
+    },
+    {
+      "epoch": 1.687223347587966,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001160566350353969,
+      "loss": 1.0911,
+      "step": 5623
+    },
+    {
+      "epoch": 1.687523445119664,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011604157252598285,
+      "loss": 1.0415,
+      "step": 5624
+    },
+    {
+      "epoch": 1.6878235426513617,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011602651001656876,
+      "loss": 0.9905,
+      "step": 5625
+    },
+    {
+      "epoch": 1.6881236401830595,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00011601144750715469,
+      "loss": 0.9229,
+      "step": 5626
+    },
+    {
+      "epoch": 1.6884237377147573,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011599638499774064,
+      "loss": 0.9545,
+      "step": 5627
+    },
+    {
+      "epoch": 1.688723835246455,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00011598132248832656,
+      "loss": 0.7346,
+      "step": 5628
+    },
+    {
+      "epoch": 1.6890239327781529,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011596625997891248,
+      "loss": 1.0247,
+      "step": 5629
+    },
+    {
+      "epoch": 1.6893240303098507,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00011595119746949842,
+      "loss": 0.9613,
+      "step": 5630
+    },
+    {
+      "epoch": 1.6896241278415485,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011593613496008435,
+      "loss": 0.9317,
+      "step": 5631
+    },
+    {
+      "epoch": 1.6899242253732463,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001159210724506703,
+      "loss": 1.1727,
+      "step": 5632
+    },
+    {
+      "epoch": 1.690224322904944,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011590600994125622,
+      "loss": 0.9298,
+      "step": 5633
+    },
+    {
+      "epoch": 1.6905244204366419,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011589094743184214,
+      "loss": 1.0612,
+      "step": 5634
+    },
+    {
+      "epoch": 1.6908245179683397,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00011587588492242808,
+      "loss": 1.2412,
+      "step": 5635
+    },
+    {
+      "epoch": 1.6911246155000375,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011586082241301401,
+      "loss": 1.0565,
+      "step": 5636
+    },
+    {
+      "epoch": 1.6914247130317355,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011584575990359994,
+      "loss": 1.0947,
+      "step": 5637
+    },
+    {
+      "epoch": 1.691724810563433,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011583069739418588,
+      "loss": 0.991,
+      "step": 5638
+    },
+    {
+      "epoch": 1.692024908095131,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001158156348847718,
+      "loss": 0.9316,
+      "step": 5639
+    },
+    {
+      "epoch": 1.6923250056268286,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011580057237535773,
+      "loss": 1.0366,
+      "step": 5640
+    },
+    {
+      "epoch": 1.6926251031585267,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011578550986594367,
+      "loss": 0.9435,
+      "step": 5641
+    },
+    {
+      "epoch": 1.6929252006902242,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001157704473565296,
+      "loss": 1.1433,
+      "step": 5642
+    },
+    {
+      "epoch": 1.6932252982219222,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011575538484711554,
+      "loss": 0.9952,
+      "step": 5643
+    },
+    {
+      "epoch": 1.6935253957536198,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011574032233770147,
+      "loss": 0.9732,
+      "step": 5644
+    },
+    {
+      "epoch": 1.6938254932853178,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00011572525982828739,
+      "loss": 0.9834,
+      "step": 5645
+    },
+    {
+      "epoch": 1.6941255908170154,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011571019731887333,
+      "loss": 1.0347,
+      "step": 5646
+    },
+    {
+      "epoch": 1.6944256883487134,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011569513480945926,
+      "loss": 0.9658,
+      "step": 5647
+    },
+    {
+      "epoch": 1.694725785880411,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011568007230004519,
+      "loss": 0.9244,
+      "step": 5648
+    },
+    {
+      "epoch": 1.695025883412109,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011566500979063113,
+      "loss": 0.9339,
+      "step": 5649
+    },
+    {
+      "epoch": 1.6953259809438066,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011564994728121705,
+      "loss": 0.9896,
+      "step": 5650
+    },
+    {
+      "epoch": 1.6956260784755046,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.000115634884771803,
+      "loss": 1.0039,
+      "step": 5651
+    },
+    {
+      "epoch": 1.6959261760072022,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011561982226238892,
+      "loss": 0.9469,
+      "step": 5652
+    },
+    {
+      "epoch": 1.6962262735389002,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00011560475975297485,
+      "loss": 1.1093,
+      "step": 5653
+    },
+    {
+      "epoch": 1.6965263710705978,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011558969724356079,
+      "loss": 1.0533,
+      "step": 5654
+    },
+    {
+      "epoch": 1.6968264686022958,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00011557463473414671,
+      "loss": 0.9555,
+      "step": 5655
+    },
+    {
+      "epoch": 1.6971265661339936,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011555957222473264,
+      "loss": 1.1648,
+      "step": 5656
+    },
+    {
+      "epoch": 1.6974266636656914,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00011554450971531858,
+      "loss": 0.915,
+      "step": 5657
+    },
+    {
+      "epoch": 1.6977267611973892,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011552944720590451,
+      "loss": 1.0693,
+      "step": 5658
+    },
+    {
+      "epoch": 1.698026858729087,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011551438469649045,
+      "loss": 1.094,
+      "step": 5659
+    },
+    {
+      "epoch": 1.6983269562607848,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011549932218707638,
+      "loss": 1.0546,
+      "step": 5660
+    },
+    {
+      "epoch": 1.6986270537924826,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001154842596776623,
+      "loss": 0.9981,
+      "step": 5661
+    },
+    {
+      "epoch": 1.6989271513241804,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011546919716824824,
+      "loss": 0.9811,
+      "step": 5662
+    },
+    {
+      "epoch": 1.6992272488558782,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011545413465883417,
+      "loss": 1.0955,
+      "step": 5663
+    },
+    {
+      "epoch": 1.699527346387576,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001154390721494201,
+      "loss": 1.1908,
+      "step": 5664
+    },
+    {
+      "epoch": 1.6998274439192738,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011542400964000604,
+      "loss": 0.982,
+      "step": 5665
+    },
+    {
+      "epoch": 1.7001275414509716,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011540894713059196,
+      "loss": 0.899,
+      "step": 5666
+    },
+    {
+      "epoch": 1.7004276389826694,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011539388462117789,
+      "loss": 0.981,
+      "step": 5667
+    },
+    {
+      "epoch": 1.7007277365143671,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011537882211176383,
+      "loss": 1.0696,
+      "step": 5668
+    },
+    {
+      "epoch": 1.701027834046065,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011536375960234976,
+      "loss": 1.1291,
+      "step": 5669
+    },
+    {
+      "epoch": 1.7013279315777627,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001153486970929357,
+      "loss": 0.8798,
+      "step": 5670
+    },
+    {
+      "epoch": 1.7016280291094605,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011533363458352162,
+      "loss": 1.015,
+      "step": 5671
+    },
+    {
+      "epoch": 1.7019281266411583,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011531857207410755,
+      "loss": 0.9153,
+      "step": 5672
+    },
+    {
+      "epoch": 1.7022282241728561,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011530350956469349,
+      "loss": 0.9184,
+      "step": 5673
+    },
+    {
+      "epoch": 1.702528321704554,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011528844705527942,
+      "loss": 0.9507,
+      "step": 5674
+    },
+    {
+      "epoch": 1.7028284192362517,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00011527338454586533,
+      "loss": 1.1702,
+      "step": 5675
+    },
+    {
+      "epoch": 1.7031285167679497,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011525832203645128,
+      "loss": 0.8558,
+      "step": 5676
+    },
+    {
+      "epoch": 1.7034286142996473,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00011524325952703721,
+      "loss": 0.9834,
+      "step": 5677
+    },
+    {
+      "epoch": 1.7037287118313453,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011522819701762315,
+      "loss": 1.0164,
+      "step": 5678
+    },
+    {
+      "epoch": 1.704028809363043,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011521313450820908,
+      "loss": 1.0395,
+      "step": 5679
+    },
+    {
+      "epoch": 1.704328906894741,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.000115198071998795,
+      "loss": 1.0233,
+      "step": 5680
+    },
+    {
+      "epoch": 1.7046290044264385,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011518300948938095,
+      "loss": 0.9174,
+      "step": 5681
+    },
+    {
+      "epoch": 1.7049291019581365,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011516794697996687,
+      "loss": 1.0231,
+      "step": 5682
+    },
+    {
+      "epoch": 1.705229199489834,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011515288447055279,
+      "loss": 1.0498,
+      "step": 5683
+    },
+    {
+      "epoch": 1.705529297021532,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011513782196113874,
+      "loss": 1.095,
+      "step": 5684
+    },
+    {
+      "epoch": 1.7058293945532297,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011512275945172467,
+      "loss": 1.0959,
+      "step": 5685
+    },
+    {
+      "epoch": 1.7061294920849277,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001151076969423106,
+      "loss": 0.9992,
+      "step": 5686
+    },
+    {
+      "epoch": 1.7064295896166253,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00011509263443289653,
+      "loss": 0.9353,
+      "step": 5687
+    },
+    {
+      "epoch": 1.7067296871483233,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00011507757192348245,
+      "loss": 1.0151,
+      "step": 5688
+    },
+    {
+      "epoch": 1.7070297846800209,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001150625094140684,
+      "loss": 1.1042,
+      "step": 5689
+    },
+    {
+      "epoch": 1.7073298822117189,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011504744690465433,
+      "loss": 0.9766,
+      "step": 5690
+    },
+    {
+      "epoch": 1.7076299797434165,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011503238439524024,
+      "loss": 1.0769,
+      "step": 5691
+    },
+    {
+      "epoch": 1.7079300772751145,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001150173218858262,
+      "loss": 0.9826,
+      "step": 5692
+    },
+    {
+      "epoch": 1.708230174806812,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011500225937641211,
+      "loss": 1.0225,
+      "step": 5693
+    },
+    {
+      "epoch": 1.70853027233851,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011498719686699803,
+      "loss": 0.8656,
+      "step": 5694
+    },
+    {
+      "epoch": 1.7088303698702079,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011497213435758399,
+      "loss": 1.1439,
+      "step": 5695
+    },
+    {
+      "epoch": 1.7091304674019057,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001149570718481699,
+      "loss": 0.9394,
+      "step": 5696
+    },
+    {
+      "epoch": 1.7094305649336035,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011494200933875585,
+      "loss": 1.088,
+      "step": 5697
+    },
+    {
+      "epoch": 1.7097306624653013,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011492694682934178,
+      "loss": 1.0269,
+      "step": 5698
+    },
+    {
+      "epoch": 1.710030759996999,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001149118843199277,
+      "loss": 1.1266,
+      "step": 5699
+    },
+    {
+      "epoch": 1.7103308575286968,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011489682181051365,
+      "loss": 1.0304,
+      "step": 5700
+    },
+    {
+      "epoch": 1.7106309550603946,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011488175930109956,
+      "loss": 1.0778,
+      "step": 5701
+    },
+    {
+      "epoch": 1.7109310525920924,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00011486669679168549,
+      "loss": 0.9095,
+      "step": 5702
+    },
+    {
+      "epoch": 1.7112311501237902,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00011485163428227144,
+      "loss": 1.1772,
+      "step": 5703
+    },
+    {
+      "epoch": 1.711531247655488,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011483657177285736,
+      "loss": 1.0921,
+      "step": 5704
+    },
+    {
+      "epoch": 1.7118313451871858,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00011482150926344331,
+      "loss": 1.1072,
+      "step": 5705
+    },
+    {
+      "epoch": 1.7121314427188836,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011480644675402922,
+      "loss": 1.1556,
+      "step": 5706
+    },
+    {
+      "epoch": 1.7124315402505814,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00011479138424461515,
+      "loss": 1.0108,
+      "step": 5707
+    },
+    {
+      "epoch": 1.7127316377822792,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001147763217352011,
+      "loss": 1.0379,
+      "step": 5708
+    },
+    {
+      "epoch": 1.713031735313977,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011476125922578702,
+      "loss": 1.0931,
+      "step": 5709
+    },
+    {
+      "epoch": 1.7133318328456748,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011474619671637294,
+      "loss": 1.0469,
+      "step": 5710
+    },
+    {
+      "epoch": 1.7136319303773726,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011473113420695888,
+      "loss": 1.0595,
+      "step": 5711
+    },
+    {
+      "epoch": 1.7139320279090704,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011471607169754481,
+      "loss": 1.1875,
+      "step": 5712
+    },
+    {
+      "epoch": 1.7142321254407682,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011470100918813076,
+      "loss": 0.9512,
+      "step": 5713
+    },
+    {
+      "epoch": 1.714532222972466,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011468594667871668,
+      "loss": 1.0116,
+      "step": 5714
+    },
+    {
+      "epoch": 1.714832320504164,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001146708841693026,
+      "loss": 1.0395,
+      "step": 5715
+    },
+    {
+      "epoch": 1.7151324180358616,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011465582165988856,
+      "loss": 1.1923,
+      "step": 5716
+    },
+    {
+      "epoch": 1.7154325155675596,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00011464075915047447,
+      "loss": 1.0593,
+      "step": 5717
+    },
+    {
+      "epoch": 1.7157326130992572,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001146256966410604,
+      "loss": 1.0772,
+      "step": 5718
+    },
+    {
+      "epoch": 1.7160327106309552,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00011461063413164634,
+      "loss": 0.9621,
+      "step": 5719
+    },
+    {
+      "epoch": 1.7163328081626528,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011459557162223226,
+      "loss": 1.0649,
+      "step": 5720
+    },
+    {
+      "epoch": 1.7166329056943508,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011458050911281819,
+      "loss": 0.9262,
+      "step": 5721
+    },
+    {
+      "epoch": 1.7169330032260484,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011456544660340413,
+      "loss": 1.1254,
+      "step": 5722
+    },
+    {
+      "epoch": 1.7172331007577464,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011455038409399006,
+      "loss": 1.0148,
+      "step": 5723
+    },
+    {
+      "epoch": 1.717533198289444,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.000114535321584576,
+      "loss": 1.035,
+      "step": 5724
+    },
+    {
+      "epoch": 1.717833295821142,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00011452025907516193,
+      "loss": 1.1212,
+      "step": 5725
+    },
+    {
+      "epoch": 1.7181333933528395,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011450519656574785,
+      "loss": 1.0265,
+      "step": 5726
+    },
+    {
+      "epoch": 1.7184334908845376,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011449013405633379,
+      "loss": 1.0787,
+      "step": 5727
+    },
+    {
+      "epoch": 1.7187335884162351,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00011447507154691972,
+      "loss": 1.1029,
+      "step": 5728
+    },
+    {
+      "epoch": 1.7190336859479332,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00011446000903750565,
+      "loss": 0.9411,
+      "step": 5729
+    },
+    {
+      "epoch": 1.7193337834796307,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00011444494652809159,
+      "loss": 1.0141,
+      "step": 5730
+    },
+    {
+      "epoch": 1.7196338810113287,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011442988401867751,
+      "loss": 1.0563,
+      "step": 5731
+    },
+    {
+      "epoch": 1.7199339785430263,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011441482150926345,
+      "loss": 0.9716,
+      "step": 5732
+    },
+    {
+      "epoch": 1.7202340760747243,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011439975899984938,
+      "loss": 0.9238,
+      "step": 5733
+    },
+    {
+      "epoch": 1.7205341736064221,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00011438469649043531,
+      "loss": 1.1297,
+      "step": 5734
+    },
+    {
+      "epoch": 1.72083427113812,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011436963398102125,
+      "loss": 1.0446,
+      "step": 5735
+    },
+    {
+      "epoch": 1.7211343686698177,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011435457147160717,
+      "loss": 1.1168,
+      "step": 5736
+    },
+    {
+      "epoch": 1.7214344662015155,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001143395089621931,
+      "loss": 1.0384,
+      "step": 5737
+    },
+    {
+      "epoch": 1.7217345637332133,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011432444645277904,
+      "loss": 1.0615,
+      "step": 5738
+    },
+    {
+      "epoch": 1.7220346612649111,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00011430938394336497,
+      "loss": 1.2319,
+      "step": 5739
+    },
+    {
+      "epoch": 1.722334758796609,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011429432143395091,
+      "loss": 0.9877,
+      "step": 5740
+    },
+    {
+      "epoch": 1.7226348563283067,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00011427925892453684,
+      "loss": 0.9509,
+      "step": 5741
+    },
+    {
+      "epoch": 1.7229349538600045,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011426419641512276,
+      "loss": 1.0534,
+      "step": 5742
+    },
+    {
+      "epoch": 1.7232350513917023,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001142491339057087,
+      "loss": 0.8995,
+      "step": 5743
+    },
+    {
+      "epoch": 1.7235351489234,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011423407139629463,
+      "loss": 0.9508,
+      "step": 5744
+    },
+    {
+      "epoch": 1.723835246455098,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011421900888688056,
+      "loss": 0.9297,
+      "step": 5745
+    },
+    {
+      "epoch": 1.7241353439867957,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001142039463774665,
+      "loss": 0.9901,
+      "step": 5746
+    },
+    {
+      "epoch": 1.7244354415184935,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011418888386805242,
+      "loss": 1.0506,
+      "step": 5747
+    },
+    {
+      "epoch": 1.7247355390501913,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00011417382135863835,
+      "loss": 1.098,
+      "step": 5748
+    },
+    {
+      "epoch": 1.725035636581889,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011415875884922429,
+      "loss": 0.9493,
+      "step": 5749
+    },
+    {
+      "epoch": 1.7253357341135869,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011414369633981022,
+      "loss": 1.0011,
+      "step": 5750
+    },
+    {
+      "epoch": 1.7256358316452847,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011412863383039616,
+      "loss": 1.0787,
+      "step": 5751
+    },
+    {
+      "epoch": 1.7259359291769825,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00011411357132098208,
+      "loss": 1.1036,
+      "step": 5752
+    },
+    {
+      "epoch": 1.7262360267086803,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00011409850881156801,
+      "loss": 0.9588,
+      "step": 5753
+    },
+    {
+      "epoch": 1.7265361242403783,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00011408344630215395,
+      "loss": 0.9336,
+      "step": 5754
+    },
+    {
+      "epoch": 1.7268362217720759,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00011406838379273988,
+      "loss": 1.1569,
+      "step": 5755
+    },
+    {
+      "epoch": 1.7271363193037739,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.0001140533212833258,
+      "loss": 1.1624,
+      "step": 5756
+    },
+    {
+      "epoch": 1.7274364168354714,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00011403825877391174,
+      "loss": 1.0099,
+      "step": 5757
+    },
+    {
+      "epoch": 1.7277365143671695,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011402319626449767,
+      "loss": 1.157,
+      "step": 5758
+    },
+    {
+      "epoch": 1.728036611898867,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011400813375508361,
+      "loss": 1.0364,
+      "step": 5759
+    },
+    {
+      "epoch": 1.728336709430565,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011399307124566954,
+      "loss": 1.0329,
+      "step": 5760
+    },
+    {
+      "epoch": 1.7286368069622626,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011397800873625547,
+      "loss": 1.0013,
+      "step": 5761
+    },
+    {
+      "epoch": 1.7289369044939606,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.0001139629462268414,
+      "loss": 0.9556,
+      "step": 5762
+    },
+    {
+      "epoch": 1.7292370020256582,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00011394788371742733,
+      "loss": 1.0367,
+      "step": 5763
+    },
+    {
+      "epoch": 1.7295370995573562,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011393282120801325,
+      "loss": 1.0863,
+      "step": 5764
+    },
+    {
+      "epoch": 1.7298371970890538,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001139177586985992,
+      "loss": 0.9396,
+      "step": 5765
+    },
+    {
+      "epoch": 1.7301372946207518,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00011390269618918513,
+      "loss": 0.9932,
+      "step": 5766
+    },
+    {
+      "epoch": 1.7304373921524494,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011388763367977107,
+      "loss": 1.0608,
+      "step": 5767
+    },
+    {
+      "epoch": 1.7307374896841474,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011387257117035699,
+      "loss": 1.0474,
+      "step": 5768
+    },
+    {
+      "epoch": 1.731037587215845,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011385750866094292,
+      "loss": 0.9629,
+      "step": 5769
+    },
+    {
+      "epoch": 1.731337684747543,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011384244615152886,
+      "loss": 0.9815,
+      "step": 5770
+    },
+    {
+      "epoch": 1.7316377822792406,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011382738364211479,
+      "loss": 1.029,
+      "step": 5771
+    },
+    {
+      "epoch": 1.7319378798109386,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001138123211327007,
+      "loss": 0.9592,
+      "step": 5772
+    },
+    {
+      "epoch": 1.7322379773426364,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011379725862328665,
+      "loss": 1.0609,
+      "step": 5773
+    },
+    {
+      "epoch": 1.7325380748743342,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011378219611387258,
+      "loss": 1.135,
+      "step": 5774
+    },
+    {
+      "epoch": 1.732838172406032,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.0001137671336044585,
+      "loss": 0.9085,
+      "step": 5775
+    },
+    {
+      "epoch": 1.7331382699377298,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011375207109504445,
+      "loss": 1.057,
+      "step": 5776
+    },
+    {
+      "epoch": 1.7334383674694276,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011373700858563036,
+      "loss": 0.9937,
+      "step": 5777
+    },
+    {
+      "epoch": 1.7337384650011254,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011372194607621631,
+      "loss": 0.9173,
+      "step": 5778
+    },
+    {
+      "epoch": 1.7340385625328232,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011370688356680224,
+      "loss": 1.0338,
+      "step": 5779
+    },
+    {
+      "epoch": 1.734338660064521,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011369182105738815,
+      "loss": 1.0438,
+      "step": 5780
+    },
+    {
+      "epoch": 1.7346387575962188,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011367675854797411,
+      "loss": 0.9028,
+      "step": 5781
+    },
+    {
+      "epoch": 1.7349388551279166,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00011366169603856002,
+      "loss": 1.0255,
+      "step": 5782
+    },
+    {
+      "epoch": 1.7352389526596144,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011364663352914595,
+      "loss": 1.0887,
+      "step": 5783
+    },
+    {
+      "epoch": 1.7355390501913122,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001136315710197319,
+      "loss": 1.2191,
+      "step": 5784
+    },
+    {
+      "epoch": 1.73583914772301,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00011361650851031782,
+      "loss": 1.0035,
+      "step": 5785
+    },
+    {
+      "epoch": 1.7361392452547078,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011360144600090377,
+      "loss": 1.0183,
+      "step": 5786
+    },
+    {
+      "epoch": 1.7364393427864055,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001135863834914897,
+      "loss": 1.01,
+      "step": 5787
+    },
+    {
+      "epoch": 1.7367394403181033,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011357132098207561,
+      "loss": 1.0833,
+      "step": 5788
+    },
+    {
+      "epoch": 1.7370395378498011,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011355625847266156,
+      "loss": 0.9571,
+      "step": 5789
+    },
+    {
+      "epoch": 1.737339635381499,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011354119596324748,
+      "loss": 1.0652,
+      "step": 5790
+    },
+    {
+      "epoch": 1.737639732913197,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001135261334538334,
+      "loss": 0.955,
+      "step": 5791
+    },
+    {
+      "epoch": 1.7379398304448945,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011351107094441936,
+      "loss": 1.059,
+      "step": 5792
+    },
+    {
+      "epoch": 1.7382399279765925,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011349600843500527,
+      "loss": 0.8777,
+      "step": 5793
+    },
+    {
+      "epoch": 1.7385400255082901,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011348094592559122,
+      "loss": 0.9839,
+      "step": 5794
+    },
+    {
+      "epoch": 1.7388401230399881,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011346588341617714,
+      "loss": 1.18,
+      "step": 5795
+    },
+    {
+      "epoch": 1.7391402205716857,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011345082090676306,
+      "loss": 1.1548,
+      "step": 5796
+    },
+    {
+      "epoch": 1.7394403181033837,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011343575839734902,
+      "loss": 0.9883,
+      "step": 5797
+    },
+    {
+      "epoch": 1.7397404156350813,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011342069588793493,
+      "loss": 1.0185,
+      "step": 5798
+    },
+    {
+      "epoch": 1.7400405131667793,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011340563337852086,
+      "loss": 0.863,
+      "step": 5799
+    },
+    {
+      "epoch": 1.740340610698477,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001133905708691068,
+      "loss": 0.7617,
+      "step": 5800
+    },
+    {
+      "epoch": 1.740640708230175,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011337550835969272,
+      "loss": 0.9938,
+      "step": 5801
+    },
+    {
+      "epoch": 1.7409408057618725,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011336044585027865,
+      "loss": 0.9448,
+      "step": 5802
+    },
+    {
+      "epoch": 1.7412409032935705,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00011334538334086459,
+      "loss": 0.9137,
+      "step": 5803
+    },
+    {
+      "epoch": 1.741541000825268,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011333032083145052,
+      "loss": 1.0515,
+      "step": 5804
+    },
+    {
+      "epoch": 1.741841098356966,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011331525832203647,
+      "loss": 1.0182,
+      "step": 5805
+    },
+    {
+      "epoch": 1.7421411958886637,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00011330019581262239,
+      "loss": 1.0829,
+      "step": 5806
+    },
+    {
+      "epoch": 1.7424412934203617,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011328513330320831,
+      "loss": 1.0006,
+      "step": 5807
+    },
+    {
+      "epoch": 1.7427413909520593,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00011327007079379425,
+      "loss": 0.8421,
+      "step": 5808
+    },
+    {
+      "epoch": 1.7430414884837573,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011325500828438018,
+      "loss": 1.0355,
+      "step": 5809
+    },
+    {
+      "epoch": 1.7433415860154549,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001132399457749661,
+      "loss": 1.1037,
+      "step": 5810
+    },
+    {
+      "epoch": 1.7436416835471529,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011322488326555205,
+      "loss": 1.0861,
+      "step": 5811
+    },
+    {
+      "epoch": 1.7439417810788507,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00011320982075613797,
+      "loss": 0.8799,
+      "step": 5812
+    },
+    {
+      "epoch": 1.7442418786105485,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011319475824672391,
+      "loss": 1.0234,
+      "step": 5813
+    },
+    {
+      "epoch": 1.7445419761422463,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00011317969573730984,
+      "loss": 0.8784,
+      "step": 5814
+    },
+    {
+      "epoch": 1.744842073673944,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011316463322789577,
+      "loss": 1.1431,
+      "step": 5815
+    },
+    {
+      "epoch": 1.7451421712056419,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011314957071848171,
+      "loss": 1.0556,
+      "step": 5816
+    },
+    {
+      "epoch": 1.7454422687373397,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011313450820906763,
+      "loss": 0.8681,
+      "step": 5817
+    },
+    {
+      "epoch": 1.7457423662690374,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011311944569965356,
+      "loss": 1.0759,
+      "step": 5818
+    },
+    {
+      "epoch": 1.7460424638007352,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001131043831902395,
+      "loss": 0.9927,
+      "step": 5819
+    },
+    {
+      "epoch": 1.746342561332433,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011308932068082543,
+      "loss": 1.0866,
+      "step": 5820
+    },
+    {
+      "epoch": 1.7466426588641308,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011307425817141137,
+      "loss": 0.9298,
+      "step": 5821
+    },
+    {
+      "epoch": 1.7469427563958286,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001130591956619973,
+      "loss": 1.1021,
+      "step": 5822
+    },
+    {
+      "epoch": 1.7472428539275264,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00011304413315258322,
+      "loss": 1.0066,
+      "step": 5823
+    },
+    {
+      "epoch": 1.7475429514592242,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011302907064316916,
+      "loss": 1.1395,
+      "step": 5824
+    },
+    {
+      "epoch": 1.747843048990922,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00011301400813375509,
+      "loss": 1.0432,
+      "step": 5825
+    },
+    {
+      "epoch": 1.7481431465226198,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011299894562434102,
+      "loss": 1.0738,
+      "step": 5826
+    },
+    {
+      "epoch": 1.7484432440543176,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011298388311492696,
+      "loss": 0.9954,
+      "step": 5827
+    },
+    {
+      "epoch": 1.7487433415860154,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011296882060551288,
+      "loss": 0.9284,
+      "step": 5828
+    },
+    {
+      "epoch": 1.7490434391177132,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011295375809609881,
+      "loss": 1.0494,
+      "step": 5829
+    },
+    {
+      "epoch": 1.7493435366494112,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011293869558668475,
+      "loss": 1.0915,
+      "step": 5830
+    },
+    {
+      "epoch": 1.7496436341811088,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011292363307727068,
+      "loss": 0.9659,
+      "step": 5831
+    },
+    {
+      "epoch": 1.7499437317128068,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011290857056785662,
+      "loss": 1.0188,
+      "step": 5832
+    },
+    {
+      "epoch": 1.7502438292445044,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011289350805844254,
+      "loss": 0.8701,
+      "step": 5833
+    },
+    {
+      "epoch": 1.7505439267762024,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011287844554902847,
+      "loss": 1.0277,
+      "step": 5834
+    },
+    {
+      "epoch": 1.7508440243079,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011286338303961441,
+      "loss": 1.0491,
+      "step": 5835
+    },
+    {
+      "epoch": 1.751144121839598,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011284832053020034,
+      "loss": 1.1032,
+      "step": 5836
+    },
+    {
+      "epoch": 1.7514442193712956,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011283325802078626,
+      "loss": 1.0437,
+      "step": 5837
+    },
+    {
+      "epoch": 1.7517443169029936,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001128181955113722,
+      "loss": 1.13,
+      "step": 5838
+    },
+    {
+      "epoch": 1.7520444144346912,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011280313300195813,
+      "loss": 1.0476,
+      "step": 5839
+    },
+    {
+      "epoch": 1.7523445119663892,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011278807049254407,
+      "loss": 0.9723,
+      "step": 5840
+    },
+    {
+      "epoch": 1.7526446094980868,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011277300798313,
+      "loss": 1.0012,
+      "step": 5841
+    },
+    {
+      "epoch": 1.7529447070297848,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011275794547371593,
+      "loss": 1.0163,
+      "step": 5842
+    },
+    {
+      "epoch": 1.7532448045614824,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011274288296430187,
+      "loss": 0.9887,
+      "step": 5843
+    },
+    {
+      "epoch": 1.7535449020931804,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011272782045488779,
+      "loss": 0.9634,
+      "step": 5844
+    },
+    {
+      "epoch": 1.753844999624878,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00011271275794547372,
+      "loss": 1.0029,
+      "step": 5845
+    },
+    {
+      "epoch": 1.754145097156576,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00011269769543605966,
+      "loss": 0.9525,
+      "step": 5846
+    },
+    {
+      "epoch": 1.7544451946882735,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011268263292664559,
+      "loss": 0.9585,
+      "step": 5847
+    },
+    {
+      "epoch": 1.7547452922199716,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011266757041723153,
+      "loss": 0.949,
+      "step": 5848
+    },
+    {
+      "epoch": 1.7550453897516691,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011265250790781745,
+      "loss": 0.93,
+      "step": 5849
+    },
+    {
+      "epoch": 1.7553454872833671,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011263744539840338,
+      "loss": 1.1081,
+      "step": 5850
+    },
+    {
+      "epoch": 1.755645584815065,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011262238288898932,
+      "loss": 1.167,
+      "step": 5851
+    },
+    {
+      "epoch": 1.7559456823467627,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011260732037957525,
+      "loss": 0.9722,
+      "step": 5852
+    },
+    {
+      "epoch": 1.7562457798784605,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011259225787016116,
+      "loss": 1.0024,
+      "step": 5853
+    },
+    {
+      "epoch": 1.7565458774101583,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00011257719536074711,
+      "loss": 1.2198,
+      "step": 5854
+    },
+    {
+      "epoch": 1.7568459749418561,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011256213285133304,
+      "loss": 0.907,
+      "step": 5855
+    },
+    {
+      "epoch": 1.757146072473554,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00011254707034191895,
+      "loss": 1.0288,
+      "step": 5856
+    },
+    {
+      "epoch": 1.7574461700052517,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011253200783250491,
+      "loss": 1.155,
+      "step": 5857
+    },
+    {
+      "epoch": 1.7577462675369495,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011251694532309083,
+      "loss": 1.0628,
+      "step": 5858
+    },
+    {
+      "epoch": 1.7580463650686473,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011250188281367677,
+      "loss": 1.0147,
+      "step": 5859
+    },
+    {
+      "epoch": 1.758346462600345,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001124868203042627,
+      "loss": 1.0812,
+      "step": 5860
+    },
+    {
+      "epoch": 1.758646560132043,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011247175779484861,
+      "loss": 0.949,
+      "step": 5861
+    },
+    {
+      "epoch": 1.7589466576637407,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011245669528543457,
+      "loss": 1.0511,
+      "step": 5862
+    },
+    {
+      "epoch": 1.7592467551954385,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001124416327760205,
+      "loss": 0.9171,
+      "step": 5863
+    },
+    {
+      "epoch": 1.7595468527271363,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00011242657026660641,
+      "loss": 0.9535,
+      "step": 5864
+    },
+    {
+      "epoch": 1.759846950258834,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00011241150775719236,
+      "loss": 1.1188,
+      "step": 5865
+    },
+    {
+      "epoch": 1.7601470477905319,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011239644524777828,
+      "loss": 1.1245,
+      "step": 5866
+    },
+    {
+      "epoch": 1.7604471453222297,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011238138273836423,
+      "loss": 1.0,
+      "step": 5867
+    },
+    {
+      "epoch": 1.7607472428539275,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011236632022895016,
+      "loss": 0.8905,
+      "step": 5868
+    },
+    {
+      "epoch": 1.7610473403856255,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011235125771953607,
+      "loss": 0.7961,
+      "step": 5869
+    },
+    {
+      "epoch": 1.761347437917323,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011233619521012202,
+      "loss": 1.0845,
+      "step": 5870
+    },
+    {
+      "epoch": 1.761647535449021,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011232113270070794,
+      "loss": 0.9311,
+      "step": 5871
+    },
+    {
+      "epoch": 1.7619476329807187,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00011230607019129386,
+      "loss": 0.8095,
+      "step": 5872
+    },
+    {
+      "epoch": 1.7622477305124167,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011229100768187982,
+      "loss": 1.1738,
+      "step": 5873
+    },
+    {
+      "epoch": 1.7625478280441143,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00011227594517246573,
+      "loss": 1.075,
+      "step": 5874
+    },
+    {
+      "epoch": 1.7628479255758123,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011226088266305168,
+      "loss": 0.9873,
+      "step": 5875
+    },
+    {
+      "epoch": 1.7631480231075098,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00011224582015363761,
+      "loss": 1.0267,
+      "step": 5876
+    },
+    {
+      "epoch": 1.7634481206392079,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00011223075764422352,
+      "loss": 0.9982,
+      "step": 5877
+    },
+    {
+      "epoch": 1.7637482181709054,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011221569513480948,
+      "loss": 0.9924,
+      "step": 5878
+    },
+    {
+      "epoch": 1.7640483157026035,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011220063262539539,
+      "loss": 1.0463,
+      "step": 5879
+    },
+    {
+      "epoch": 1.764348413234301,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011218557011598132,
+      "loss": 0.9769,
+      "step": 5880
+    },
+    {
+      "epoch": 1.764648510765999,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00011217050760656727,
+      "loss": 1.0567,
+      "step": 5881
+    },
+    {
+      "epoch": 1.7649486082976966,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011215544509715318,
+      "loss": 1.0238,
+      "step": 5882
+    },
+    {
+      "epoch": 1.7652487058293946,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011214038258773911,
+      "loss": 1.0851,
+      "step": 5883
+    },
+    {
+      "epoch": 1.7655488033610922,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011212532007832505,
+      "loss": 0.9758,
+      "step": 5884
+    },
+    {
+      "epoch": 1.7658489008927902,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011211025756891098,
+      "loss": 1.0274,
+      "step": 5885
+    },
+    {
+      "epoch": 1.7661489984244878,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011209519505949693,
+      "loss": 0.976,
+      "step": 5886
+    },
+    {
+      "epoch": 1.7664490959561858,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011208013255008285,
+      "loss": 1.0094,
+      "step": 5887
+    },
+    {
+      "epoch": 1.7667491934878834,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011206507004066877,
+      "loss": 1.0971,
+      "step": 5888
+    },
+    {
+      "epoch": 1.7670492910195814,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011205000753125471,
+      "loss": 1.1145,
+      "step": 5889
+    },
+    {
+      "epoch": 1.7673493885512792,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011203494502184064,
+      "loss": 1.0566,
+      "step": 5890
+    },
+    {
+      "epoch": 1.767649486082977,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011201988251242657,
+      "loss": 1.0438,
+      "step": 5891
+    },
+    {
+      "epoch": 1.7679495836146748,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001120048200030125,
+      "loss": 0.9195,
+      "step": 5892
+    },
+    {
+      "epoch": 1.7682496811463726,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011198975749359843,
+      "loss": 1.0289,
+      "step": 5893
+    },
+    {
+      "epoch": 1.7685497786780704,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011197469498418439,
+      "loss": 1.0872,
+      "step": 5894
+    },
+    {
+      "epoch": 1.7688498762097682,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001119596324747703,
+      "loss": 0.9966,
+      "step": 5895
+    },
+    {
+      "epoch": 1.769149973741466,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011194456996535623,
+      "loss": 0.9711,
+      "step": 5896
+    },
+    {
+      "epoch": 1.7694500712731638,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011192950745594217,
+      "loss": 1.0772,
+      "step": 5897
+    },
+    {
+      "epoch": 1.7697501688048616,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001119144449465281,
+      "loss": 1.1838,
+      "step": 5898
+    },
+    {
+      "epoch": 1.7700502663365594,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011189938243711402,
+      "loss": 0.9884,
+      "step": 5899
+    },
+    {
+      "epoch": 1.7703503638682572,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00011188431992769996,
+      "loss": 1.1743,
+      "step": 5900
+    },
+    {
+      "epoch": 1.770650461399955,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011186925741828589,
+      "loss": 1.0494,
+      "step": 5901
+    },
+    {
+      "epoch": 1.7709505589316528,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011185419490887183,
+      "loss": 0.9961,
+      "step": 5902
+    },
+    {
+      "epoch": 1.7712506564633506,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011183913239945776,
+      "loss": 0.9846,
+      "step": 5903
+    },
+    {
+      "epoch": 1.7715507539950484,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011182406989004368,
+      "loss": 1.0288,
+      "step": 5904
+    },
+    {
+      "epoch": 1.7718508515267462,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011180900738062962,
+      "loss": 1.1532,
+      "step": 5905
+    },
+    {
+      "epoch": 1.772150949058444,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011179394487121555,
+      "loss": 1.0834,
+      "step": 5906
+    },
+    {
+      "epoch": 1.7724510465901417,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011177888236180148,
+      "loss": 0.8981,
+      "step": 5907
+    },
+    {
+      "epoch": 1.7727511441218398,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011176381985238742,
+      "loss": 0.8779,
+      "step": 5908
+    },
+    {
+      "epoch": 1.7730512416535373,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00011174875734297334,
+      "loss": 1.1702,
+      "step": 5909
+    },
+    {
+      "epoch": 1.7733513391852354,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00011173369483355927,
+      "loss": 1.0868,
+      "step": 5910
+    },
+    {
+      "epoch": 1.773651436716933,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011171863232414521,
+      "loss": 1.0186,
+      "step": 5911
+    },
+    {
+      "epoch": 1.773951534248631,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00011170356981473114,
+      "loss": 1.0192,
+      "step": 5912
+    },
+    {
+      "epoch": 1.7742516317803285,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011168850730531708,
+      "loss": 1.0344,
+      "step": 5913
+    },
+    {
+      "epoch": 1.7745517293120265,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.000111673444795903,
+      "loss": 0.8486,
+      "step": 5914
+    },
+    {
+      "epoch": 1.7748518268437241,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011165838228648893,
+      "loss": 1.1626,
+      "step": 5915
+    },
+    {
+      "epoch": 1.7751519243754221,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011164331977707487,
+      "loss": 0.9909,
+      "step": 5916
+    },
+    {
+      "epoch": 1.7754520219071197,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001116282572676608,
+      "loss": 1.1291,
+      "step": 5917
+    },
+    {
+      "epoch": 1.7757521194388177,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011161319475824672,
+      "loss": 1.1643,
+      "step": 5918
+    },
+    {
+      "epoch": 1.7760522169705153,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00011159813224883266,
+      "loss": 1.0562,
+      "step": 5919
+    },
+    {
+      "epoch": 1.7763523145022133,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011158306973941859,
+      "loss": 0.9815,
+      "step": 5920
+    },
+    {
+      "epoch": 1.7766524120339109,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011156800723000453,
+      "loss": 1.0431,
+      "step": 5921
+    },
+    {
+      "epoch": 1.776952509565609,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00011155294472059046,
+      "loss": 1.0113,
+      "step": 5922
+    },
+    {
+      "epoch": 1.7772526070973065,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011153788221117639,
+      "loss": 1.1099,
+      "step": 5923
+    },
+    {
+      "epoch": 1.7775527046290045,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011152281970176233,
+      "loss": 1.0454,
+      "step": 5924
+    },
+    {
+      "epoch": 1.777852802160702,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011150775719234825,
+      "loss": 1.0229,
+      "step": 5925
+    },
+    {
+      "epoch": 1.7781528996924,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00011149269468293418,
+      "loss": 1.0642,
+      "step": 5926
+    },
+    {
+      "epoch": 1.7784529972240979,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011147763217352012,
+      "loss": 0.9531,
+      "step": 5927
+    },
+    {
+      "epoch": 1.7787530947557957,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00011146256966410605,
+      "loss": 1.038,
+      "step": 5928
+    },
+    {
+      "epoch": 1.7790531922874935,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011144750715469199,
+      "loss": 1.0065,
+      "step": 5929
+    },
+    {
+      "epoch": 1.7793532898191913,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011143244464527791,
+      "loss": 0.9059,
+      "step": 5930
+    },
+    {
+      "epoch": 1.779653387350889,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011141738213586384,
+      "loss": 1.0617,
+      "step": 5931
+    },
+    {
+      "epoch": 1.7799534848825869,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011140231962644978,
+      "loss": 0.974,
+      "step": 5932
+    },
+    {
+      "epoch": 1.7802535824142847,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00011138725711703571,
+      "loss": 1.0363,
+      "step": 5933
+    },
+    {
+      "epoch": 1.7805536799459825,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011137219460762163,
+      "loss": 0.8846,
+      "step": 5934
+    },
+    {
+      "epoch": 1.7808537774776803,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011135713209820757,
+      "loss": 1.1274,
+      "step": 5935
+    },
+    {
+      "epoch": 1.781153875009378,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001113420695887935,
+      "loss": 0.9696,
+      "step": 5936
+    },
+    {
+      "epoch": 1.7814539725410758,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00011132700707937941,
+      "loss": 1.0872,
+      "step": 5937
+    },
+    {
+      "epoch": 1.7817540700727736,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00011131194456996537,
+      "loss": 1.0108,
+      "step": 5938
+    },
+    {
+      "epoch": 1.7820541676044714,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001112968820605513,
+      "loss": 1.1036,
+      "step": 5939
+    },
+    {
+      "epoch": 1.7823542651361692,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00011128181955113723,
+      "loss": 1.1135,
+      "step": 5940
+    },
+    {
+      "epoch": 1.782654362667867,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011126675704172316,
+      "loss": 1.0273,
+      "step": 5941
+    },
+    {
+      "epoch": 1.7829544601995648,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00011125169453230907,
+      "loss": 0.9471,
+      "step": 5942
+    },
+    {
+      "epoch": 1.7832545577312626,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011123663202289503,
+      "loss": 0.9557,
+      "step": 5943
+    },
+    {
+      "epoch": 1.7835546552629604,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011122156951348096,
+      "loss": 0.9999,
+      "step": 5944
+    },
+    {
+      "epoch": 1.7838547527946582,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00011120650700406687,
+      "loss": 0.9923,
+      "step": 5945
+    },
+    {
+      "epoch": 1.784154850326356,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00011119144449465282,
+      "loss": 1.0179,
+      "step": 5946
+    },
+    {
+      "epoch": 1.784454947858054,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00011117638198523874,
+      "loss": 1.016,
+      "step": 5947
+    },
+    {
+      "epoch": 1.7847550453897516,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011116131947582469,
+      "loss": 0.9319,
+      "step": 5948
+    },
+    {
+      "epoch": 1.7850551429214496,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011114625696641062,
+      "loss": 1.0707,
+      "step": 5949
+    },
+    {
+      "epoch": 1.7853552404531472,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011113119445699653,
+      "loss": 0.9791,
+      "step": 5950
+    },
+    {
+      "epoch": 1.7856553379848452,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011111613194758248,
+      "loss": 1.0698,
+      "step": 5951
+    },
+    {
+      "epoch": 1.7859554355165428,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011110106943816841,
+      "loss": 0.965,
+      "step": 5952
+    },
+    {
+      "epoch": 1.7862555330482408,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00011108600692875432,
+      "loss": 0.9312,
+      "step": 5953
+    },
+    {
+      "epoch": 1.7865556305799384,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00011107094441934028,
+      "loss": 1.0591,
+      "step": 5954
+    },
+    {
+      "epoch": 1.7868557281116364,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011105588190992619,
+      "loss": 1.0331,
+      "step": 5955
+    },
+    {
+      "epoch": 1.787155825643334,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011104081940051214,
+      "loss": 0.9466,
+      "step": 5956
+    },
+    {
+      "epoch": 1.787455923175032,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011102575689109807,
+      "loss": 0.9934,
+      "step": 5957
+    },
+    {
+      "epoch": 1.7877560207067296,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011101069438168398,
+      "loss": 0.9125,
+      "step": 5958
+    },
+    {
+      "epoch": 1.7880561182384276,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011099563187226994,
+      "loss": 1.0566,
+      "step": 5959
+    },
+    {
+      "epoch": 1.7883562157701252,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011098056936285585,
+      "loss": 1.0086,
+      "step": 5960
+    },
+    {
+      "epoch": 1.7886563133018232,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00011096550685344178,
+      "loss": 0.9962,
+      "step": 5961
+    },
+    {
+      "epoch": 1.7889564108335207,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011095044434402773,
+      "loss": 1.1475,
+      "step": 5962
+    },
+    {
+      "epoch": 1.7892565083652188,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00011093538183461364,
+      "loss": 0.9426,
+      "step": 5963
+    },
+    {
+      "epoch": 1.7895566058969163,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011092031932519957,
+      "loss": 0.9549,
+      "step": 5964
+    },
+    {
+      "epoch": 1.7898567034286144,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011090525681578551,
+      "loss": 0.9904,
+      "step": 5965
+    },
+    {
+      "epoch": 1.7901568009603122,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011089019430637144,
+      "loss": 0.9395,
+      "step": 5966
+    },
+    {
+      "epoch": 1.79045689849201,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00011087513179695739,
+      "loss": 0.8306,
+      "step": 5967
+    },
+    {
+      "epoch": 1.7907569960237077,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001108600692875433,
+      "loss": 1.127,
+      "step": 5968
+    },
+    {
+      "epoch": 1.7910570935554055,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00011084500677812923,
+      "loss": 1.0216,
+      "step": 5969
+    },
+    {
+      "epoch": 1.7913571910871033,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011082994426871519,
+      "loss": 0.9409,
+      "step": 5970
+    },
+    {
+      "epoch": 1.7916572886188011,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001108148817593011,
+      "loss": 0.8966,
+      "step": 5971
+    },
+    {
+      "epoch": 1.791957386150499,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011079981924988703,
+      "loss": 0.9815,
+      "step": 5972
+    },
+    {
+      "epoch": 1.7922574836821967,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011078475674047297,
+      "loss": 1.107,
+      "step": 5973
+    },
+    {
+      "epoch": 1.7925575812138945,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001107696942310589,
+      "loss": 1.2195,
+      "step": 5974
+    },
+    {
+      "epoch": 1.7928576787455923,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011075463172164485,
+      "loss": 0.8819,
+      "step": 5975
+    },
+    {
+      "epoch": 1.7931577762772901,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00011073956921223076,
+      "loss": 1.0972,
+      "step": 5976
+    },
+    {
+      "epoch": 1.793457873808988,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00011072450670281669,
+      "loss": 1.0329,
+      "step": 5977
+    },
+    {
+      "epoch": 1.7937579713406857,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011070944419340263,
+      "loss": 1.1325,
+      "step": 5978
+    },
+    {
+      "epoch": 1.7940580688723835,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00011069438168398855,
+      "loss": 1.0438,
+      "step": 5979
+    },
+    {
+      "epoch": 1.7943581664040813,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011067931917457448,
+      "loss": 0.9602,
+      "step": 5980
+    },
+    {
+      "epoch": 1.794658263935779,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011066425666516042,
+      "loss": 1.0399,
+      "step": 5981
+    },
+    {
+      "epoch": 1.794958361467477,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00011064919415574635,
+      "loss": 0.9232,
+      "step": 5982
+    },
+    {
+      "epoch": 1.7952584589991747,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001106341316463323,
+      "loss": 0.949,
+      "step": 5983
+    },
+    {
+      "epoch": 1.7955585565308725,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00011061906913691822,
+      "loss": 1.0514,
+      "step": 5984
+    },
+    {
+      "epoch": 1.7958586540625703,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00011060400662750414,
+      "loss": 0.9781,
+      "step": 5985
+    },
+    {
+      "epoch": 1.7961587515942683,
+      "grad_norm": 0.875,
+      "learning_rate": 0.00011058894411809008,
+      "loss": 0.8654,
+      "step": 5986
+    },
+    {
+      "epoch": 1.7964588491259659,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00011057388160867601,
+      "loss": 1.0253,
+      "step": 5987
+    },
+    {
+      "epoch": 1.796758946657664,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011055881909926194,
+      "loss": 0.8602,
+      "step": 5988
+    },
+    {
+      "epoch": 1.7970590441893615,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011054375658984788,
+      "loss": 0.9595,
+      "step": 5989
+    },
+    {
+      "epoch": 1.7973591417210595,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001105286940804338,
+      "loss": 0.9729,
+      "step": 5990
+    },
+    {
+      "epoch": 1.797659239252757,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00011051363157101973,
+      "loss": 0.964,
+      "step": 5991
+    },
+    {
+      "epoch": 1.797959336784455,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00011049856906160567,
+      "loss": 0.9561,
+      "step": 5992
+    },
+    {
+      "epoch": 1.7982594343161526,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001104835065521916,
+      "loss": 0.9354,
+      "step": 5993
+    },
+    {
+      "epoch": 1.7985595318478507,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00011046844404277754,
+      "loss": 0.8672,
+      "step": 5994
+    },
+    {
+      "epoch": 1.7988596293795482,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00011045338153336346,
+      "loss": 1.0559,
+      "step": 5995
+    },
+    {
+      "epoch": 1.7991597269112463,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011043831902394939,
+      "loss": 1.0773,
+      "step": 5996
+    },
+    {
+      "epoch": 1.7994598244429438,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00011042325651453533,
+      "loss": 1.0867,
+      "step": 5997
+    },
+    {
+      "epoch": 1.7997599219746419,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011040819400512126,
+      "loss": 0.9306,
+      "step": 5998
+    },
+    {
+      "epoch": 1.8000600195063394,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011039313149570718,
+      "loss": 0.9766,
+      "step": 5999
+    },
+    {
+      "epoch": 1.8003601170380374,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00011037806898629312,
+      "loss": 1.0893,
+      "step": 6000
+    },
+    {
+      "epoch": 1.800660214569735,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011036300647687905,
+      "loss": 1.0166,
+      "step": 6001
+    },
+    {
+      "epoch": 1.800960312101433,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011034794396746499,
+      "loss": 1.0879,
+      "step": 6002
+    },
+    {
+      "epoch": 1.8012604096331306,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011033288145805092,
+      "loss": 0.9592,
+      "step": 6003
+    },
+    {
+      "epoch": 1.8015605071648286,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00011031781894863685,
+      "loss": 0.8613,
+      "step": 6004
+    },
+    {
+      "epoch": 1.8018606046965264,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00011030275643922279,
+      "loss": 1.0638,
+      "step": 6005
+    },
+    {
+      "epoch": 1.8021607022282242,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00011028769392980871,
+      "loss": 1.2032,
+      "step": 6006
+    },
+    {
+      "epoch": 1.802460799759922,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011027263142039464,
+      "loss": 1.1809,
+      "step": 6007
+    },
+    {
+      "epoch": 1.8027608972916198,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00011025756891098058,
+      "loss": 0.8955,
+      "step": 6008
+    },
+    {
+      "epoch": 1.8030609948233176,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001102425064015665,
+      "loss": 0.8651,
+      "step": 6009
+    },
+    {
+      "epoch": 1.8033610923550154,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011022744389215245,
+      "loss": 0.9738,
+      "step": 6010
+    },
+    {
+      "epoch": 1.8036611898867132,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011021238138273837,
+      "loss": 1.0684,
+      "step": 6011
+    },
+    {
+      "epoch": 1.803961287418411,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001101973188733243,
+      "loss": 1.0247,
+      "step": 6012
+    },
+    {
+      "epoch": 1.8042613849501088,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011018225636391024,
+      "loss": 1.0315,
+      "step": 6013
+    },
+    {
+      "epoch": 1.8045614824818066,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00011016719385449617,
+      "loss": 1.0319,
+      "step": 6014
+    },
+    {
+      "epoch": 1.8048615800135044,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001101521313450821,
+      "loss": 1.0617,
+      "step": 6015
+    },
+    {
+      "epoch": 1.8051616775452022,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00011013706883566803,
+      "loss": 1.0174,
+      "step": 6016
+    },
+    {
+      "epoch": 1.8054617750769,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00011012200632625396,
+      "loss": 0.9997,
+      "step": 6017
+    },
+    {
+      "epoch": 1.8057618726085978,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011010694381683987,
+      "loss": 1.0393,
+      "step": 6018
+    },
+    {
+      "epoch": 1.8060619701402956,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00011009188130742583,
+      "loss": 1.0838,
+      "step": 6019
+    },
+    {
+      "epoch": 1.8063620676719934,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00011007681879801175,
+      "loss": 0.8967,
+      "step": 6020
+    },
+    {
+      "epoch": 1.8066621652036912,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.0001100617562885977,
+      "loss": 0.9944,
+      "step": 6021
+    },
+    {
+      "epoch": 1.806962262735389,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00011004669377918362,
+      "loss": 1.0209,
+      "step": 6022
+    },
+    {
+      "epoch": 1.8072623602670868,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00011003163126976955,
+      "loss": 0.9578,
+      "step": 6023
+    },
+    {
+      "epoch": 1.8075624577987845,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00011001656876035549,
+      "loss": 1.0682,
+      "step": 6024
+    },
+    {
+      "epoch": 1.8078625553304826,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00011000150625094142,
+      "loss": 0.9559,
+      "step": 6025
+    },
+    {
+      "epoch": 1.8081626528621801,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010998644374152733,
+      "loss": 1.0005,
+      "step": 6026
+    },
+    {
+      "epoch": 1.8084627503938782,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00010997138123211328,
+      "loss": 1.0665,
+      "step": 6027
+    },
+    {
+      "epoch": 1.8087628479255757,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010995631872269921,
+      "loss": 1.0589,
+      "step": 6028
+    },
+    {
+      "epoch": 1.8090629454572738,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010994125621328515,
+      "loss": 0.9208,
+      "step": 6029
+    },
+    {
+      "epoch": 1.8093630429889713,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010992619370387108,
+      "loss": 0.9588,
+      "step": 6030
+    },
+    {
+      "epoch": 1.8096631405206693,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00010991113119445699,
+      "loss": 0.8541,
+      "step": 6031
+    },
+    {
+      "epoch": 1.809963238052367,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010989606868504294,
+      "loss": 1.0897,
+      "step": 6032
+    },
+    {
+      "epoch": 1.810263335584065,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00010988100617562887,
+      "loss": 1.0798,
+      "step": 6033
+    },
+    {
+      "epoch": 1.8105634331157625,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010986594366621478,
+      "loss": 1.176,
+      "step": 6034
+    },
+    {
+      "epoch": 1.8108635306474605,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010985088115680074,
+      "loss": 1.0071,
+      "step": 6035
+    },
+    {
+      "epoch": 1.811163628179158,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010983581864738665,
+      "loss": 1.0902,
+      "step": 6036
+    },
+    {
+      "epoch": 1.8114637257108561,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001098207561379726,
+      "loss": 0.9411,
+      "step": 6037
+    },
+    {
+      "epoch": 1.8117638232425537,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010980569362855853,
+      "loss": 0.9757,
+      "step": 6038
+    },
+    {
+      "epoch": 1.8120639207742517,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010979063111914444,
+      "loss": 1.0,
+      "step": 6039
+    },
+    {
+      "epoch": 1.8123640183059493,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001097755686097304,
+      "loss": 1.0928,
+      "step": 6040
+    },
+    {
+      "epoch": 1.8126641158376473,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010976050610031632,
+      "loss": 1.0133,
+      "step": 6041
+    },
+    {
+      "epoch": 1.8129642133693449,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00010974544359090224,
+      "loss": 1.0097,
+      "step": 6042
+    },
+    {
+      "epoch": 1.813264310901043,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00010973038108148819,
+      "loss": 0.9998,
+      "step": 6043
+    },
+    {
+      "epoch": 1.8135644084327407,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001097153185720741,
+      "loss": 1.1927,
+      "step": 6044
+    },
+    {
+      "epoch": 1.8138645059644385,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010970025606266003,
+      "loss": 1.1109,
+      "step": 6045
+    },
+    {
+      "epoch": 1.8141646034961363,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010968519355324599,
+      "loss": 1.1551,
+      "step": 6046
+    },
+    {
+      "epoch": 1.814464701027834,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001096701310438319,
+      "loss": 1.0677,
+      "step": 6047
+    },
+    {
+      "epoch": 1.8147647985595319,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010965506853441785,
+      "loss": 1.029,
+      "step": 6048
+    },
+    {
+      "epoch": 1.8150648960912297,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010964000602500377,
+      "loss": 1.0296,
+      "step": 6049
+    },
+    {
+      "epoch": 1.8153649936229275,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010962494351558969,
+      "loss": 1.0039,
+      "step": 6050
+    },
+    {
+      "epoch": 1.8156650911546253,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010960988100617565,
+      "loss": 1.0267,
+      "step": 6051
+    },
+    {
+      "epoch": 1.815965188686323,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00010959481849676156,
+      "loss": 0.913,
+      "step": 6052
+    },
+    {
+      "epoch": 1.8162652862180209,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010957975598734749,
+      "loss": 1.0879,
+      "step": 6053
+    },
+    {
+      "epoch": 1.8165653837497187,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010956469347793343,
+      "loss": 1.0653,
+      "step": 6054
+    },
+    {
+      "epoch": 1.8168654812814165,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010954963096851935,
+      "loss": 1.0628,
+      "step": 6055
+    },
+    {
+      "epoch": 1.8171655788131142,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010953456845910531,
+      "loss": 0.9902,
+      "step": 6056
+    },
+    {
+      "epoch": 1.817465676344812,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010951950594969122,
+      "loss": 0.8921,
+      "step": 6057
+    },
+    {
+      "epoch": 1.8177657738765098,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010950444344027715,
+      "loss": 1.0335,
+      "step": 6058
+    },
+    {
+      "epoch": 1.8180658714082076,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001094893809308631,
+      "loss": 0.971,
+      "step": 6059
+    },
+    {
+      "epoch": 1.8183659689399054,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010947431842144901,
+      "loss": 1.0925,
+      "step": 6060
+    },
+    {
+      "epoch": 1.8186660664716032,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010945925591203494,
+      "loss": 1.0826,
+      "step": 6061
+    },
+    {
+      "epoch": 1.818966164003301,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010944419340262088,
+      "loss": 0.9517,
+      "step": 6062
+    },
+    {
+      "epoch": 1.8192662615349988,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010942913089320681,
+      "loss": 1.0509,
+      "step": 6063
+    },
+    {
+      "epoch": 1.8195663590666968,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010941406838379276,
+      "loss": 1.0451,
+      "step": 6064
+    },
+    {
+      "epoch": 1.8198664565983944,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010939900587437868,
+      "loss": 0.9025,
+      "step": 6065
+    },
+    {
+      "epoch": 1.8201665541300924,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001093839433649646,
+      "loss": 1.0409,
+      "step": 6066
+    },
+    {
+      "epoch": 1.82046665166179,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010936888085555054,
+      "loss": 0.9132,
+      "step": 6067
+    },
+    {
+      "epoch": 1.820766749193488,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010935381834613647,
+      "loss": 0.9321,
+      "step": 6068
+    },
+    {
+      "epoch": 1.8210668467251856,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001093387558367224,
+      "loss": 1.1071,
+      "step": 6069
+    },
+    {
+      "epoch": 1.8213669442568836,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010932369332730834,
+      "loss": 0.9608,
+      "step": 6070
+    },
+    {
+      "epoch": 1.8216670417885812,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010930863081789426,
+      "loss": 1.0453,
+      "step": 6071
+    },
+    {
+      "epoch": 1.8219671393202792,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010929356830848019,
+      "loss": 0.9218,
+      "step": 6072
+    },
+    {
+      "epoch": 1.8222672368519768,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010927850579906613,
+      "loss": 1.0998,
+      "step": 6073
+    },
+    {
+      "epoch": 1.8225673343836748,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010926344328965206,
+      "loss": 1.0587,
+      "step": 6074
+    },
+    {
+      "epoch": 1.8228674319153724,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.000109248380780238,
+      "loss": 1.0083,
+      "step": 6075
+    },
+    {
+      "epoch": 1.8231675294470704,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010923331827082392,
+      "loss": 0.968,
+      "step": 6076
+    },
+    {
+      "epoch": 1.823467626978768,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010921825576140985,
+      "loss": 1.0037,
+      "step": 6077
+    },
+    {
+      "epoch": 1.823767724510466,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00010920319325199579,
+      "loss": 1.1684,
+      "step": 6078
+    },
+    {
+      "epoch": 1.8240678220421636,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010918813074258172,
+      "loss": 1.0941,
+      "step": 6079
+    },
+    {
+      "epoch": 1.8243679195738616,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010917306823316764,
+      "loss": 1.0483,
+      "step": 6080
+    },
+    {
+      "epoch": 1.8246680171055591,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010915800572375358,
+      "loss": 0.9545,
+      "step": 6081
+    },
+    {
+      "epoch": 1.8249681146372572,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010914294321433951,
+      "loss": 1.0834,
+      "step": 6082
+    },
+    {
+      "epoch": 1.825268212168955,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010912788070492545,
+      "loss": 1.0004,
+      "step": 6083
+    },
+    {
+      "epoch": 1.8255683097006528,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010911281819551138,
+      "loss": 0.9983,
+      "step": 6084
+    },
+    {
+      "epoch": 1.8258684072323506,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001090977556860973,
+      "loss": 1.0427,
+      "step": 6085
+    },
+    {
+      "epoch": 1.8261685047640484,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010908269317668325,
+      "loss": 1.0206,
+      "step": 6086
+    },
+    {
+      "epoch": 1.8264686022957461,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010906763066726917,
+      "loss": 1.211,
+      "step": 6087
+    },
+    {
+      "epoch": 1.826768699827444,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001090525681578551,
+      "loss": 1.0513,
+      "step": 6088
+    },
+    {
+      "epoch": 1.8270687973591417,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010903750564844104,
+      "loss": 0.8665,
+      "step": 6089
+    },
+    {
+      "epoch": 1.8273688948908395,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00010902244313902697,
+      "loss": 0.908,
+      "step": 6090
+    },
+    {
+      "epoch": 1.8276689924225373,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001090073806296129,
+      "loss": 0.9063,
+      "step": 6091
+    },
+    {
+      "epoch": 1.8279690899542351,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00010899231812019883,
+      "loss": 1.1907,
+      "step": 6092
+    },
+    {
+      "epoch": 1.828269187485933,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010897725561078476,
+      "loss": 1.0124,
+      "step": 6093
+    },
+    {
+      "epoch": 1.8285692850176307,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001089621931013707,
+      "loss": 1.0522,
+      "step": 6094
+    },
+    {
+      "epoch": 1.8288693825493285,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010894713059195663,
+      "loss": 0.9985,
+      "step": 6095
+    },
+    {
+      "epoch": 1.8291694800810263,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010893206808254255,
+      "loss": 1.0089,
+      "step": 6096
+    },
+    {
+      "epoch": 1.829469577612724,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001089170055731285,
+      "loss": 1.1578,
+      "step": 6097
+    },
+    {
+      "epoch": 1.829769675144422,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010890194306371442,
+      "loss": 0.9567,
+      "step": 6098
+    },
+    {
+      "epoch": 1.8300697726761197,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010888688055430035,
+      "loss": 0.9772,
+      "step": 6099
+    },
+    {
+      "epoch": 1.8303698702078175,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010887181804488629,
+      "loss": 1.0721,
+      "step": 6100
+    },
+    {
+      "epoch": 1.8306699677395153,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010885675553547221,
+      "loss": 1.0455,
+      "step": 6101
+    },
+    {
+      "epoch": 1.830970065271213,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010884169302605815,
+      "loss": 0.9991,
+      "step": 6102
+    },
+    {
+      "epoch": 1.831270162802911,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010882663051664408,
+      "loss": 1.0956,
+      "step": 6103
+    },
+    {
+      "epoch": 1.8315702603346087,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010881156800723001,
+      "loss": 1.108,
+      "step": 6104
+    },
+    {
+      "epoch": 1.8318703578663067,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010879650549781595,
+      "loss": 0.9699,
+      "step": 6105
+    },
+    {
+      "epoch": 1.8321704553980043,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010878144298840188,
+      "loss": 0.9674,
+      "step": 6106
+    },
+    {
+      "epoch": 1.8324705529297023,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010876638047898779,
+      "loss": 0.9369,
+      "step": 6107
+    },
+    {
+      "epoch": 1.8327706504613999,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010875131796957374,
+      "loss": 0.9853,
+      "step": 6108
+    },
+    {
+      "epoch": 1.8330707479930979,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010873625546015967,
+      "loss": 0.9458,
+      "step": 6109
+    },
+    {
+      "epoch": 1.8333708455247955,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010872119295074561,
+      "loss": 1.0722,
+      "step": 6110
+    },
+    {
+      "epoch": 1.8336709430564935,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010870613044133154,
+      "loss": 1.0098,
+      "step": 6111
+    },
+    {
+      "epoch": 1.833971040588191,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00010869106793191746,
+      "loss": 1.3263,
+      "step": 6112
+    },
+    {
+      "epoch": 1.834271138119889,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001086760054225034,
+      "loss": 1.1402,
+      "step": 6113
+    },
+    {
+      "epoch": 1.8345712356515866,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010866094291308933,
+      "loss": 1.1151,
+      "step": 6114
+    },
+    {
+      "epoch": 1.8348713331832847,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.00010864588040367524,
+      "loss": 1.026,
+      "step": 6115
+    },
+    {
+      "epoch": 1.8351714307149822,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001086308178942612,
+      "loss": 1.028,
+      "step": 6116
+    },
+    {
+      "epoch": 1.8354715282466803,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00010861575538484712,
+      "loss": 1.0082,
+      "step": 6117
+    },
+    {
+      "epoch": 1.8357716257783778,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010860069287543306,
+      "loss": 1.0675,
+      "step": 6118
+    },
+    {
+      "epoch": 1.8360717233100758,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010858563036601899,
+      "loss": 1.0728,
+      "step": 6119
+    },
+    {
+      "epoch": 1.8363718208417734,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001085705678566049,
+      "loss": 0.9314,
+      "step": 6120
+    },
+    {
+      "epoch": 1.8366719183734714,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010855550534719086,
+      "loss": 0.9078,
+      "step": 6121
+    },
+    {
+      "epoch": 1.8369720159051692,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010854044283777678,
+      "loss": 1.1173,
+      "step": 6122
+    },
+    {
+      "epoch": 1.837272113436867,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001085253803283627,
+      "loss": 1.0666,
+      "step": 6123
+    },
+    {
+      "epoch": 1.8375722109685648,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00010851031781894865,
+      "loss": 1.0746,
+      "step": 6124
+    },
+    {
+      "epoch": 1.8378723085002626,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010849525530953456,
+      "loss": 0.9654,
+      "step": 6125
+    },
+    {
+      "epoch": 1.8381724060319604,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010848019280012049,
+      "loss": 0.9954,
+      "step": 6126
+    },
+    {
+      "epoch": 1.8384725035636582,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00010846513029070645,
+      "loss": 0.92,
+      "step": 6127
+    },
+    {
+      "epoch": 1.838772601095356,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010845006778129236,
+      "loss": 1.0606,
+      "step": 6128
+    },
+    {
+      "epoch": 1.8390726986270538,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010843500527187831,
+      "loss": 0.9052,
+      "step": 6129
+    },
+    {
+      "epoch": 1.8393727961587516,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00010841994276246424,
+      "loss": 1.014,
+      "step": 6130
+    },
+    {
+      "epoch": 1.8396728936904494,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010840488025305015,
+      "loss": 1.0594,
+      "step": 6131
+    },
+    {
+      "epoch": 1.8399729912221472,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.0001083898177436361,
+      "loss": 0.9538,
+      "step": 6132
+    },
+    {
+      "epoch": 1.840273088753845,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010837475523422202,
+      "loss": 0.8939,
+      "step": 6133
+    },
+    {
+      "epoch": 1.8405731862855428,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010835969272480795,
+      "loss": 1.032,
+      "step": 6134
+    },
+    {
+      "epoch": 1.8408732838172406,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001083446302153939,
+      "loss": 1.0176,
+      "step": 6135
+    },
+    {
+      "epoch": 1.8411733813489384,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00010832956770597981,
+      "loss": 1.0188,
+      "step": 6136
+    },
+    {
+      "epoch": 1.8414734788806362,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010831450519656577,
+      "loss": 0.8746,
+      "step": 6137
+    },
+    {
+      "epoch": 1.841773576412334,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010829944268715168,
+      "loss": 1.048,
+      "step": 6138
+    },
+    {
+      "epoch": 1.8420736739440318,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010828438017773761,
+      "loss": 1.0363,
+      "step": 6139
+    },
+    {
+      "epoch": 1.8423737714757296,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010826931766832356,
+      "loss": 1.0051,
+      "step": 6140
+    },
+    {
+      "epoch": 1.8426738690074274,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010825425515890947,
+      "loss": 0.9448,
+      "step": 6141
+    },
+    {
+      "epoch": 1.8429739665391254,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001082391926494954,
+      "loss": 1.0818,
+      "step": 6142
+    },
+    {
+      "epoch": 1.843274064070823,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00010822413014008134,
+      "loss": 0.9325,
+      "step": 6143
+    },
+    {
+      "epoch": 1.843574161602521,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010820906763066727,
+      "loss": 0.9916,
+      "step": 6144
+    },
+    {
+      "epoch": 1.8438742591342185,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00010819400512125322,
+      "loss": 0.9814,
+      "step": 6145
+    },
+    {
+      "epoch": 1.8441743566659166,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010817894261183914,
+      "loss": 1.0539,
+      "step": 6146
+    },
+    {
+      "epoch": 1.8444744541976141,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010816388010242506,
+      "loss": 1.1002,
+      "step": 6147
+    },
+    {
+      "epoch": 1.8447745517293122,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010814881759301102,
+      "loss": 0.9165,
+      "step": 6148
+    },
+    {
+      "epoch": 1.8450746492610097,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010813375508359693,
+      "loss": 0.8476,
+      "step": 6149
+    },
+    {
+      "epoch": 1.8453747467927077,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00010811869257418286,
+      "loss": 1.0951,
+      "step": 6150
+    },
+    {
+      "epoch": 1.8456748443244053,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.0001081036300647688,
+      "loss": 0.8372,
+      "step": 6151
+    },
+    {
+      "epoch": 1.8459749418561033,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010808856755535472,
+      "loss": 0.9173,
+      "step": 6152
+    },
+    {
+      "epoch": 1.846275039387801,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010807350504594065,
+      "loss": 1.0772,
+      "step": 6153
+    },
+    {
+      "epoch": 1.846575136919499,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010805844253652659,
+      "loss": 1.0156,
+      "step": 6154
+    },
+    {
+      "epoch": 1.8468752344511965,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010804338002711252,
+      "loss": 1.0969,
+      "step": 6155
+    },
+    {
+      "epoch": 1.8471753319828945,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010802831751769846,
+      "loss": 0.9285,
+      "step": 6156
+    },
+    {
+      "epoch": 1.847475429514592,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00010801325500828438,
+      "loss": 1.0568,
+      "step": 6157
+    },
+    {
+      "epoch": 1.8477755270462901,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010799819249887031,
+      "loss": 1.045,
+      "step": 6158
+    },
+    {
+      "epoch": 1.8480756245779877,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010798312998945625,
+      "loss": 0.9683,
+      "step": 6159
+    },
+    {
+      "epoch": 1.8483757221096857,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010796806748004218,
+      "loss": 1.0089,
+      "step": 6160
+    },
+    {
+      "epoch": 1.8486758196413835,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001079530049706281,
+      "loss": 1.0747,
+      "step": 6161
+    },
+    {
+      "epoch": 1.8489759171730813,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00010793794246121404,
+      "loss": 0.9475,
+      "step": 6162
+    },
+    {
+      "epoch": 1.849276014704779,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010792287995179997,
+      "loss": 0.9594,
+      "step": 6163
+    },
+    {
+      "epoch": 1.8495761122364769,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010790781744238591,
+      "loss": 0.9847,
+      "step": 6164
+    },
+    {
+      "epoch": 1.8498762097681747,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010789275493297184,
+      "loss": 0.9617,
+      "step": 6165
+    },
+    {
+      "epoch": 1.8501763072998725,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010787769242355777,
+      "loss": 0.9898,
+      "step": 6166
+    },
+    {
+      "epoch": 1.8504764048315703,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001078626299141437,
+      "loss": 0.9681,
+      "step": 6167
+    },
+    {
+      "epoch": 1.850776502363268,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010784756740472963,
+      "loss": 1.0796,
+      "step": 6168
+    },
+    {
+      "epoch": 1.8510765998949659,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010783250489531556,
+      "loss": 1.1019,
+      "step": 6169
+    },
+    {
+      "epoch": 1.8513766974266637,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001078174423859015,
+      "loss": 0.9031,
+      "step": 6170
+    },
+    {
+      "epoch": 1.8516767949583615,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010780237987648743,
+      "loss": 0.9343,
+      "step": 6171
+    },
+    {
+      "epoch": 1.8519768924900593,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010778731736707337,
+      "loss": 0.9771,
+      "step": 6172
+    },
+    {
+      "epoch": 1.852276990021757,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00010777225485765929,
+      "loss": 1.0623,
+      "step": 6173
+    },
+    {
+      "epoch": 1.8525770875534548,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010775719234824522,
+      "loss": 1.0284,
+      "step": 6174
+    },
+    {
+      "epoch": 1.8528771850851526,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010774212983883116,
+      "loss": 0.8961,
+      "step": 6175
+    },
+    {
+      "epoch": 1.8531772826168504,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00010772706732941709,
+      "loss": 1.042,
+      "step": 6176
+    },
+    {
+      "epoch": 1.8534773801485482,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010771200482000301,
+      "loss": 1.0847,
+      "step": 6177
+    },
+    {
+      "epoch": 1.853777477680246,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010769694231058895,
+      "loss": 0.9485,
+      "step": 6178
+    },
+    {
+      "epoch": 1.8540775752119438,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.00010768187980117488,
+      "loss": 0.9505,
+      "step": 6179
+    },
+    {
+      "epoch": 1.8543776727436416,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010766681729176081,
+      "loss": 0.9869,
+      "step": 6180
+    },
+    {
+      "epoch": 1.8546777702753396,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010765175478234675,
+      "loss": 0.9772,
+      "step": 6181
+    },
+    {
+      "epoch": 1.8549778678070372,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010763669227293267,
+      "loss": 1.0594,
+      "step": 6182
+    },
+    {
+      "epoch": 1.8552779653387352,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010762162976351861,
+      "loss": 1.0066,
+      "step": 6183
+    },
+    {
+      "epoch": 1.8555780628704328,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010760656725410454,
+      "loss": 1.1014,
+      "step": 6184
+    },
+    {
+      "epoch": 1.8558781604021308,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010759150474469047,
+      "loss": 1.0971,
+      "step": 6185
+    },
+    {
+      "epoch": 1.8561782579338284,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00010757644223527641,
+      "loss": 0.9901,
+      "step": 6186
+    },
+    {
+      "epoch": 1.8564783554655264,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00010756137972586234,
+      "loss": 0.8656,
+      "step": 6187
+    },
+    {
+      "epoch": 1.856778452997224,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010754631721644826,
+      "loss": 0.9985,
+      "step": 6188
+    },
+    {
+      "epoch": 1.857078550528922,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001075312547070342,
+      "loss": 0.9943,
+      "step": 6189
+    },
+    {
+      "epoch": 1.8573786480606196,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010751619219762013,
+      "loss": 0.9115,
+      "step": 6190
+    },
+    {
+      "epoch": 1.8576787455923176,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010750112968820607,
+      "loss": 0.9517,
+      "step": 6191
+    },
+    {
+      "epoch": 1.8579788431240152,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.000107486067178792,
+      "loss": 0.9503,
+      "step": 6192
+    },
+    {
+      "epoch": 1.8582789406557132,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010747100466937792,
+      "loss": 0.9211,
+      "step": 6193
+    },
+    {
+      "epoch": 1.8585790381874108,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00010745594215996386,
+      "loss": 0.9556,
+      "step": 6194
+    },
+    {
+      "epoch": 1.8588791357191088,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00010744087965054979,
+      "loss": 0.8907,
+      "step": 6195
+    },
+    {
+      "epoch": 1.8591792332508064,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001074258171411357,
+      "loss": 1.0917,
+      "step": 6196
+    },
+    {
+      "epoch": 1.8594793307825044,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010741075463172166,
+      "loss": 0.9318,
+      "step": 6197
+    },
+    {
+      "epoch": 1.859779428314202,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010739569212230758,
+      "loss": 1.0325,
+      "step": 6198
+    },
+    {
+      "epoch": 1.8600795258459,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010738062961289352,
+      "loss": 1.0569,
+      "step": 6199
+    },
+    {
+      "epoch": 1.8603796233775978,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010736556710347945,
+      "loss": 1.0366,
+      "step": 6200
+    },
+    {
+      "epoch": 1.8606797209092956,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010735050459406536,
+      "loss": 1.0472,
+      "step": 6201
+    },
+    {
+      "epoch": 1.8609798184409934,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00010733544208465132,
+      "loss": 1.047,
+      "step": 6202
+    },
+    {
+      "epoch": 1.8612799159726912,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010732037957523724,
+      "loss": 0.9607,
+      "step": 6203
+    },
+    {
+      "epoch": 1.861580013504389,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010730531706582316,
+      "loss": 0.9066,
+      "step": 6204
+    },
+    {
+      "epoch": 1.8618801110360867,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010729025455640911,
+      "loss": 0.9204,
+      "step": 6205
+    },
+    {
+      "epoch": 1.8621802085677845,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010727519204699504,
+      "loss": 1.0145,
+      "step": 6206
+    },
+    {
+      "epoch": 1.8624803060994823,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00010726012953758095,
+      "loss": 1.148,
+      "step": 6207
+    },
+    {
+      "epoch": 1.8627804036311801,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001072450670281669,
+      "loss": 1.2362,
+      "step": 6208
+    },
+    {
+      "epoch": 1.863080501162878,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010723000451875282,
+      "loss": 1.0734,
+      "step": 6209
+    },
+    {
+      "epoch": 1.8633805986945757,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010721494200933877,
+      "loss": 0.8696,
+      "step": 6210
+    },
+    {
+      "epoch": 1.8636806962262735,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001071998794999247,
+      "loss": 1.1699,
+      "step": 6211
+    },
+    {
+      "epoch": 1.8639807937579713,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010718481699051061,
+      "loss": 0.978,
+      "step": 6212
+    },
+    {
+      "epoch": 1.8642808912896691,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010716975448109657,
+      "loss": 1.0311,
+      "step": 6213
+    },
+    {
+      "epoch": 1.864580988821367,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010715469197168248,
+      "loss": 1.1857,
+      "step": 6214
+    },
+    {
+      "epoch": 1.8648810863530647,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001071396294622684,
+      "loss": 1.0655,
+      "step": 6215
+    },
+    {
+      "epoch": 1.8651811838847625,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010712456695285436,
+      "loss": 0.952,
+      "step": 6216
+    },
+    {
+      "epoch": 1.8654812814164603,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010710950444344027,
+      "loss": 1.0364,
+      "step": 6217
+    },
+    {
+      "epoch": 1.865781378948158,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00010709444193402623,
+      "loss": 1.0529,
+      "step": 6218
+    },
+    {
+      "epoch": 1.866081476479856,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010707937942461215,
+      "loss": 0.9968,
+      "step": 6219
+    },
+    {
+      "epoch": 1.866381574011554,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010706431691519807,
+      "loss": 1.0301,
+      "step": 6220
+    },
+    {
+      "epoch": 1.8666816715432515,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010704925440578402,
+      "loss": 1.1259,
+      "step": 6221
+    },
+    {
+      "epoch": 1.8669817690749495,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010703419189636993,
+      "loss": 0.9762,
+      "step": 6222
+    },
+    {
+      "epoch": 1.867281866606647,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010701912938695586,
+      "loss": 1.0251,
+      "step": 6223
+    },
+    {
+      "epoch": 1.867581964138345,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010700406687754181,
+      "loss": 1.0478,
+      "step": 6224
+    },
+    {
+      "epoch": 1.8678820616700427,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010698900436812773,
+      "loss": 0.9624,
+      "step": 6225
+    },
+    {
+      "epoch": 1.8681821592017407,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010697394185871368,
+      "loss": 1.0564,
+      "step": 6226
+    },
+    {
+      "epoch": 1.8684822567334383,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001069588793492996,
+      "loss": 0.88,
+      "step": 6227
+    },
+    {
+      "epoch": 1.8687823542651363,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010694381683988552,
+      "loss": 0.7695,
+      "step": 6228
+    },
+    {
+      "epoch": 1.8690824517968339,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010692875433047148,
+      "loss": 0.9427,
+      "step": 6229
+    },
+    {
+      "epoch": 1.8693825493285319,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010691369182105739,
+      "loss": 0.9448,
+      "step": 6230
+    },
+    {
+      "epoch": 1.8696826468602294,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00010689862931164332,
+      "loss": 0.9868,
+      "step": 6231
+    },
+    {
+      "epoch": 1.8699827443919275,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010688356680222926,
+      "loss": 1.0393,
+      "step": 6232
+    },
+    {
+      "epoch": 1.870282841923625,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010686850429281518,
+      "loss": 0.98,
+      "step": 6233
+    },
+    {
+      "epoch": 1.870582939455323,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010685344178340111,
+      "loss": 0.9917,
+      "step": 6234
+    },
+    {
+      "epoch": 1.8708830369870206,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010683837927398705,
+      "loss": 1.0117,
+      "step": 6235
+    },
+    {
+      "epoch": 1.8711831345187186,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010682331676457298,
+      "loss": 0.9887,
+      "step": 6236
+    },
+    {
+      "epoch": 1.8714832320504162,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010680825425515893,
+      "loss": 0.9309,
+      "step": 6237
+    },
+    {
+      "epoch": 1.8717833295821142,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010679319174574484,
+      "loss": 0.9014,
+      "step": 6238
+    },
+    {
+      "epoch": 1.872083427113812,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010677812923633077,
+      "loss": 0.9669,
+      "step": 6239
+    },
+    {
+      "epoch": 1.8723835246455098,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010676306672691671,
+      "loss": 1.1659,
+      "step": 6240
+    },
+    {
+      "epoch": 1.8726836221772076,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00010674800421750264,
+      "loss": 0.9706,
+      "step": 6241
+    },
+    {
+      "epoch": 1.8729837197089054,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010673294170808856,
+      "loss": 0.8591,
+      "step": 6242
+    },
+    {
+      "epoch": 1.8732838172406032,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001067178791986745,
+      "loss": 0.9951,
+      "step": 6243
+    },
+    {
+      "epoch": 1.873583914772301,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010670281668926043,
+      "loss": 0.877,
+      "step": 6244
+    },
+    {
+      "epoch": 1.8738840123039988,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010668775417984637,
+      "loss": 1.032,
+      "step": 6245
+    },
+    {
+      "epoch": 1.8741841098356966,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001066726916704323,
+      "loss": 0.9372,
+      "step": 6246
+    },
+    {
+      "epoch": 1.8744842073673944,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00010665762916101823,
+      "loss": 1.0334,
+      "step": 6247
+    },
+    {
+      "epoch": 1.8747843048990922,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00010664256665160417,
+      "loss": 1.0507,
+      "step": 6248
+    },
+    {
+      "epoch": 1.87508440243079,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010662750414219009,
+      "loss": 0.8998,
+      "step": 6249
+    },
+    {
+      "epoch": 1.8753844999624878,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00010661244163277602,
+      "loss": 0.8992,
+      "step": 6250
+    },
+    {
+      "epoch": 1.8756845974941856,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010659737912336196,
+      "loss": 0.9471,
+      "step": 6251
+    },
+    {
+      "epoch": 1.8759846950258834,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010658231661394789,
+      "loss": 0.9473,
+      "step": 6252
+    },
+    {
+      "epoch": 1.8762847925575812,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010656725410453383,
+      "loss": 0.9274,
+      "step": 6253
+    },
+    {
+      "epoch": 1.876584890089279,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00010655219159511975,
+      "loss": 1.1102,
+      "step": 6254
+    },
+    {
+      "epoch": 1.8768849876209768,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010653712908570568,
+      "loss": 0.9779,
+      "step": 6255
+    },
+    {
+      "epoch": 1.8771850851526746,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010652206657629162,
+      "loss": 1.0675,
+      "step": 6256
+    },
+    {
+      "epoch": 1.8774851826843724,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010650700406687755,
+      "loss": 0.9922,
+      "step": 6257
+    },
+    {
+      "epoch": 1.8777852802160702,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010649194155746347,
+      "loss": 1.1151,
+      "step": 6258
+    },
+    {
+      "epoch": 1.8780853777477682,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010647687904804941,
+      "loss": 1.0337,
+      "step": 6259
+    },
+    {
+      "epoch": 1.8783854752794658,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010646181653863534,
+      "loss": 1.001,
+      "step": 6260
+    },
+    {
+      "epoch": 1.8786855728111638,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010644675402922127,
+      "loss": 0.9337,
+      "step": 6261
+    },
+    {
+      "epoch": 1.8789856703428613,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010643169151980721,
+      "loss": 0.9977,
+      "step": 6262
+    },
+    {
+      "epoch": 1.8792857678745594,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010641662901039313,
+      "loss": 0.9799,
+      "step": 6263
+    },
+    {
+      "epoch": 1.879585865406257,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010640156650097907,
+      "loss": 1.0624,
+      "step": 6264
+    },
+    {
+      "epoch": 1.879885962937955,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.000106386503991565,
+      "loss": 0.8608,
+      "step": 6265
+    },
+    {
+      "epoch": 1.8801860604696525,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010637144148215093,
+      "loss": 0.9719,
+      "step": 6266
+    },
+    {
+      "epoch": 1.8804861580013505,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010635637897273687,
+      "loss": 1.1354,
+      "step": 6267
+    },
+    {
+      "epoch": 1.8807862555330481,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001063413164633228,
+      "loss": 1.0487,
+      "step": 6268
+    },
+    {
+      "epoch": 1.8810863530647461,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010632625395390872,
+      "loss": 1.0073,
+      "step": 6269
+    },
+    {
+      "epoch": 1.8813864505964437,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010631119144449466,
+      "loss": 1.1122,
+      "step": 6270
+    },
+    {
+      "epoch": 1.8816865481281417,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010629612893508059,
+      "loss": 0.9322,
+      "step": 6271
+    },
+    {
+      "epoch": 1.8819866456598393,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010628106642566653,
+      "loss": 1.0931,
+      "step": 6272
+    },
+    {
+      "epoch": 1.8822867431915373,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010626600391625246,
+      "loss": 0.8368,
+      "step": 6273
+    },
+    {
+      "epoch": 1.882586840723235,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010625094140683838,
+      "loss": 0.964,
+      "step": 6274
+    },
+    {
+      "epoch": 1.882886938254933,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010623587889742432,
+      "loss": 0.9495,
+      "step": 6275
+    },
+    {
+      "epoch": 1.8831870357866305,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010622081638801025,
+      "loss": 1.0005,
+      "step": 6276
+    },
+    {
+      "epoch": 1.8834871333183285,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010620575387859618,
+      "loss": 0.966,
+      "step": 6277
+    },
+    {
+      "epoch": 1.8837872308500263,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010619069136918212,
+      "loss": 1.0777,
+      "step": 6278
+    },
+    {
+      "epoch": 1.884087328381724,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010617562885976804,
+      "loss": 0.9965,
+      "step": 6279
+    },
+    {
+      "epoch": 1.884387425913422,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010616056635035398,
+      "loss": 0.9628,
+      "step": 6280
+    },
+    {
+      "epoch": 1.8846875234451197,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010614550384093991,
+      "loss": 0.9627,
+      "step": 6281
+    },
+    {
+      "epoch": 1.8849876209768175,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010613044133152584,
+      "loss": 0.9605,
+      "step": 6282
+    },
+    {
+      "epoch": 1.8852877185085153,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010611537882211178,
+      "loss": 1.0052,
+      "step": 6283
+    },
+    {
+      "epoch": 1.885587816040213,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001061003163126977,
+      "loss": 0.9527,
+      "step": 6284
+    },
+    {
+      "epoch": 1.8858879135719109,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010608525380328362,
+      "loss": 1.0226,
+      "step": 6285
+    },
+    {
+      "epoch": 1.8861880111036087,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00010607019129386957,
+      "loss": 1.0103,
+      "step": 6286
+    },
+    {
+      "epoch": 1.8864881086353065,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001060551287844555,
+      "loss": 0.9302,
+      "step": 6287
+    },
+    {
+      "epoch": 1.8867882061670043,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00010604006627504141,
+      "loss": 1.1528,
+      "step": 6288
+    },
+    {
+      "epoch": 1.887088303698702,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00010602500376562737,
+      "loss": 0.9632,
+      "step": 6289
+    },
+    {
+      "epoch": 1.8873884012303999,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010600994125621328,
+      "loss": 0.9662,
+      "step": 6290
+    },
+    {
+      "epoch": 1.8876884987620977,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010599487874679923,
+      "loss": 1.0757,
+      "step": 6291
+    },
+    {
+      "epoch": 1.8879885962937955,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010597981623738516,
+      "loss": 1.112,
+      "step": 6292
+    },
+    {
+      "epoch": 1.8882886938254932,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00010596475372797107,
+      "loss": 1.0219,
+      "step": 6293
+    },
+    {
+      "epoch": 1.888588791357191,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010594969121855703,
+      "loss": 1.0567,
+      "step": 6294
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00010593462870914295,
+      "loss": 0.9499,
+      "step": 6295
+    },
+    {
+      "epoch": 1.8891889864205866,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010591956619972887,
+      "loss": 0.999,
+      "step": 6296
+    },
+    {
+      "epoch": 1.8894890839522844,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010590450369031482,
+      "loss": 1.0039,
+      "step": 6297
+    },
+    {
+      "epoch": 1.8897891814839824,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010588944118090073,
+      "loss": 1.0398,
+      "step": 6298
+    },
+    {
+      "epoch": 1.89008927901568,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010587437867148669,
+      "loss": 1.093,
+      "step": 6299
+    },
+    {
+      "epoch": 1.890389376547378,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010585931616207261,
+      "loss": 0.9305,
+      "step": 6300
+    },
+    {
+      "epoch": 1.8906894740790756,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00010584425365265853,
+      "loss": 1.1072,
+      "step": 6301
+    },
+    {
+      "epoch": 1.8909895716107736,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010582919114324448,
+      "loss": 0.9859,
+      "step": 6302
+    },
+    {
+      "epoch": 1.8912896691424712,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001058141286338304,
+      "loss": 1.0918,
+      "step": 6303
+    },
+    {
+      "epoch": 1.8915897666741692,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010579906612441632,
+      "loss": 0.9386,
+      "step": 6304
+    },
+    {
+      "epoch": 1.8918898642058668,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010578400361500227,
+      "loss": 0.9979,
+      "step": 6305
+    },
+    {
+      "epoch": 1.8921899617375648,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010576894110558819,
+      "loss": 1.0043,
+      "step": 6306
+    },
+    {
+      "epoch": 1.8924900592692624,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010575387859617414,
+      "loss": 0.9916,
+      "step": 6307
+    },
+    {
+      "epoch": 1.8927901568009604,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00010573881608676006,
+      "loss": 1.2092,
+      "step": 6308
+    },
+    {
+      "epoch": 1.893090254332658,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010572375357734598,
+      "loss": 1.064,
+      "step": 6309
+    },
+    {
+      "epoch": 1.893390351864356,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00010570869106793194,
+      "loss": 0.9169,
+      "step": 6310
+    },
+    {
+      "epoch": 1.8936904493960536,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00010569362855851785,
+      "loss": 1.0032,
+      "step": 6311
+    },
+    {
+      "epoch": 1.8939905469277516,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010567856604910378,
+      "loss": 0.9008,
+      "step": 6312
+    },
+    {
+      "epoch": 1.8942906444594492,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010566350353968973,
+      "loss": 1.011,
+      "step": 6313
+    },
+    {
+      "epoch": 1.8945907419911472,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010564844103027564,
+      "loss": 0.7991,
+      "step": 6314
+    },
+    {
+      "epoch": 1.8948908395228448,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010563337852086157,
+      "loss": 1.0601,
+      "step": 6315
+    },
+    {
+      "epoch": 1.8951909370545428,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010561831601144751,
+      "loss": 0.9982,
+      "step": 6316
+    },
+    {
+      "epoch": 1.8954910345862406,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010560325350203344,
+      "loss": 1.0822,
+      "step": 6317
+    },
+    {
+      "epoch": 1.8957911321179384,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00010558819099261939,
+      "loss": 0.8935,
+      "step": 6318
+    },
+    {
+      "epoch": 1.8960912296496362,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.0001055731284832053,
+      "loss": 1.063,
+      "step": 6319
+    },
+    {
+      "epoch": 1.896391327181334,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010555806597379123,
+      "loss": 0.9116,
+      "step": 6320
+    },
+    {
+      "epoch": 1.8966914247130318,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010554300346437717,
+      "loss": 1.0838,
+      "step": 6321
+    },
+    {
+      "epoch": 1.8969915222447296,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001055279409549631,
+      "loss": 1.0025,
+      "step": 6322
+    },
+    {
+      "epoch": 1.8972916197764274,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010551287844554902,
+      "loss": 0.9691,
+      "step": 6323
+    },
+    {
+      "epoch": 1.8975917173081251,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010549781593613496,
+      "loss": 1.0001,
+      "step": 6324
+    },
+    {
+      "epoch": 1.897891814839823,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00010548275342672089,
+      "loss": 0.8441,
+      "step": 6325
+    },
+    {
+      "epoch": 1.8981919123715207,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00010546769091730683,
+      "loss": 1.0289,
+      "step": 6326
+    },
+    {
+      "epoch": 1.8984920099032185,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010545262840789276,
+      "loss": 1.0508,
+      "step": 6327
+    },
+    {
+      "epoch": 1.8987921074349163,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010543756589847869,
+      "loss": 0.9324,
+      "step": 6328
+    },
+    {
+      "epoch": 1.8990922049666141,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010542250338906463,
+      "loss": 0.9489,
+      "step": 6329
+    },
+    {
+      "epoch": 1.899392302498312,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00010540744087965055,
+      "loss": 1.0303,
+      "step": 6330
+    },
+    {
+      "epoch": 1.8996924000300097,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010539237837023648,
+      "loss": 0.8721,
+      "step": 6331
+    },
+    {
+      "epoch": 1.8999924975617075,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010537731586082242,
+      "loss": 1.1659,
+      "step": 6332
+    },
+    {
+      "epoch": 1.9002925950934053,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010536225335140835,
+      "loss": 0.9525,
+      "step": 6333
+    },
+    {
+      "epoch": 1.900592692625103,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010534719084199429,
+      "loss": 0.9298,
+      "step": 6334
+    },
+    {
+      "epoch": 1.9008927901568011,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010533212833258021,
+      "loss": 1.0536,
+      "step": 6335
+    },
+    {
+      "epoch": 1.9011928876884987,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00010531706582316614,
+      "loss": 0.8854,
+      "step": 6336
+    },
+    {
+      "epoch": 1.9014929852201967,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00010530200331375208,
+      "loss": 1.1016,
+      "step": 6337
+    },
+    {
+      "epoch": 1.9017930827518943,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010528694080433801,
+      "loss": 0.9456,
+      "step": 6338
+    },
+    {
+      "epoch": 1.9020931802835923,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010527187829492393,
+      "loss": 0.9938,
+      "step": 6339
+    },
+    {
+      "epoch": 1.9023932778152899,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010525681578550987,
+      "loss": 0.879,
+      "step": 6340
+    },
+    {
+      "epoch": 1.902693375346988,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001052417532760958,
+      "loss": 1.0631,
+      "step": 6341
+    },
+    {
+      "epoch": 1.9029934728786855,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010522669076668173,
+      "loss": 1.0045,
+      "step": 6342
+    },
+    {
+      "epoch": 1.9032935704103835,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010521162825726767,
+      "loss": 0.9897,
+      "step": 6343
+    },
+    {
+      "epoch": 1.903593667942081,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001051965657478536,
+      "loss": 0.999,
+      "step": 6344
+    },
+    {
+      "epoch": 1.903893765473779,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010518150323843953,
+      "loss": 1.0013,
+      "step": 6345
+    },
+    {
+      "epoch": 1.9041938630054767,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00010516644072902546,
+      "loss": 0.9752,
+      "step": 6346
+    },
+    {
+      "epoch": 1.9044939605371747,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010515137821961139,
+      "loss": 1.1374,
+      "step": 6347
+    },
+    {
+      "epoch": 1.9047940580688723,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010513631571019733,
+      "loss": 0.9708,
+      "step": 6348
+    },
+    {
+      "epoch": 1.9050941556005703,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010512125320078326,
+      "loss": 0.9576,
+      "step": 6349
+    },
+    {
+      "epoch": 1.9053942531322678,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010510619069136918,
+      "loss": 0.85,
+      "step": 6350
+    },
+    {
+      "epoch": 1.9056943506639659,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010509112818195512,
+      "loss": 0.9433,
+      "step": 6351
+    },
+    {
+      "epoch": 1.9059944481956634,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010507606567254105,
+      "loss": 1.0328,
+      "step": 6352
+    },
+    {
+      "epoch": 1.9062945457273615,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010506100316312699,
+      "loss": 0.8746,
+      "step": 6353
+    },
+    {
+      "epoch": 1.906594643259059,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00010504594065371292,
+      "loss": 1.1503,
+      "step": 6354
+    },
+    {
+      "epoch": 1.906894740790757,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010503087814429884,
+      "loss": 0.9227,
+      "step": 6355
+    },
+    {
+      "epoch": 1.9071948383224548,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00010501581563488478,
+      "loss": 0.8705,
+      "step": 6356
+    },
+    {
+      "epoch": 1.9074949358541526,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010500075312547071,
+      "loss": 0.8819,
+      "step": 6357
+    },
+    {
+      "epoch": 1.9077950333858504,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010498569061605664,
+      "loss": 0.9936,
+      "step": 6358
+    },
+    {
+      "epoch": 1.9080951309175482,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010497062810664258,
+      "loss": 1.0449,
+      "step": 6359
+    },
+    {
+      "epoch": 1.908395228449246,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001049555655972285,
+      "loss": 0.9457,
+      "step": 6360
+    },
+    {
+      "epoch": 1.9086953259809438,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010494050308781444,
+      "loss": 0.9653,
+      "step": 6361
+    },
+    {
+      "epoch": 1.9089954235126416,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010492544057840037,
+      "loss": 0.9875,
+      "step": 6362
+    },
+    {
+      "epoch": 1.9092955210443394,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001049103780689863,
+      "loss": 0.9529,
+      "step": 6363
+    },
+    {
+      "epoch": 1.9095956185760372,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010489531555957224,
+      "loss": 0.9532,
+      "step": 6364
+    },
+    {
+      "epoch": 1.909895716107735,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010488025305015816,
+      "loss": 1.0438,
+      "step": 6365
+    },
+    {
+      "epoch": 1.9101958136394328,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010486519054074409,
+      "loss": 0.9882,
+      "step": 6366
+    },
+    {
+      "epoch": 1.9104959111711306,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00010485012803133003,
+      "loss": 0.903,
+      "step": 6367
+    },
+    {
+      "epoch": 1.9107960087028284,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010483506552191596,
+      "loss": 1.127,
+      "step": 6368
+    },
+    {
+      "epoch": 1.9110961062345262,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010482000301250187,
+      "loss": 0.8914,
+      "step": 6369
+    },
+    {
+      "epoch": 1.911396203766224,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00010480494050308783,
+      "loss": 0.9192,
+      "step": 6370
+    },
+    {
+      "epoch": 1.9116963012979218,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010478987799367375,
+      "loss": 0.9644,
+      "step": 6371
+    },
+    {
+      "epoch": 1.9119963988296196,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010477481548425969,
+      "loss": 0.9825,
+      "step": 6372
+    },
+    {
+      "epoch": 1.9122964963613174,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010475975297484562,
+      "loss": 1.0716,
+      "step": 6373
+    },
+    {
+      "epoch": 1.9125965938930154,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00010474469046543153,
+      "loss": 1.0381,
+      "step": 6374
+    },
+    {
+      "epoch": 1.912896691424713,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010472962795601749,
+      "loss": 0.923,
+      "step": 6375
+    },
+    {
+      "epoch": 1.913196788956411,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010471456544660341,
+      "loss": 0.8346,
+      "step": 6376
+    },
+    {
+      "epoch": 1.9134968864881086,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010469950293718933,
+      "loss": 0.9053,
+      "step": 6377
+    },
+    {
+      "epoch": 1.9137969840198066,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010468444042777528,
+      "loss": 0.9927,
+      "step": 6378
+    },
+    {
+      "epoch": 1.9140970815515042,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001046693779183612,
+      "loss": 1.0214,
+      "step": 6379
+    },
+    {
+      "epoch": 1.9143971790832022,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010465431540894715,
+      "loss": 1.1164,
+      "step": 6380
+    },
+    {
+      "epoch": 1.9146972766148997,
+      "grad_norm": 0.86328125,
+      "learning_rate": 0.00010463925289953307,
+      "loss": 0.9625,
+      "step": 6381
+    },
+    {
+      "epoch": 1.9149973741465978,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010462419039011899,
+      "loss": 0.9468,
+      "step": 6382
+    },
+    {
+      "epoch": 1.9152974716782953,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010460912788070494,
+      "loss": 0.9181,
+      "step": 6383
+    },
+    {
+      "epoch": 1.9155975692099934,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010459406537129087,
+      "loss": 1.1199,
+      "step": 6384
+    },
+    {
+      "epoch": 1.915897666741691,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010457900286187678,
+      "loss": 1.072,
+      "step": 6385
+    },
+    {
+      "epoch": 1.916197764273389,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00010456394035246273,
+      "loss": 1.0282,
+      "step": 6386
+    },
+    {
+      "epoch": 1.9164978618050865,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010454887784304865,
+      "loss": 0.943,
+      "step": 6387
+    },
+    {
+      "epoch": 1.9167979593367845,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001045338153336346,
+      "loss": 1.0134,
+      "step": 6388
+    },
+    {
+      "epoch": 1.9170980568684821,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010451875282422053,
+      "loss": 0.9052,
+      "step": 6389
+    },
+    {
+      "epoch": 1.9173981544001801,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00010450369031480644,
+      "loss": 0.9539,
+      "step": 6390
+    },
+    {
+      "epoch": 1.9176982519318777,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001044886278053924,
+      "loss": 0.997,
+      "step": 6391
+    },
+    {
+      "epoch": 1.9179983494635757,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010447356529597831,
+      "loss": 1.0536,
+      "step": 6392
+    },
+    {
+      "epoch": 1.9182984469952733,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010445850278656424,
+      "loss": 1.0233,
+      "step": 6393
+    },
+    {
+      "epoch": 1.9185985445269713,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010444344027715019,
+      "loss": 1.2047,
+      "step": 6394
+    },
+    {
+      "epoch": 1.9188986420586691,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.0001044283777677361,
+      "loss": 0.9992,
+      "step": 6395
+    },
+    {
+      "epoch": 1.919198739590367,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010441331525832203,
+      "loss": 0.8861,
+      "step": 6396
+    },
+    {
+      "epoch": 1.9194988371220647,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010439825274890797,
+      "loss": 1.0248,
+      "step": 6397
+    },
+    {
+      "epoch": 1.9197989346537625,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.0001043831902394939,
+      "loss": 0.9909,
+      "step": 6398
+    },
+    {
+      "epoch": 1.9200990321854603,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00010436812773007985,
+      "loss": 0.9976,
+      "step": 6399
+    },
+    {
+      "epoch": 1.920399129717158,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010435306522066576,
+      "loss": 0.9328,
+      "step": 6400
+    },
+    {
+      "epoch": 1.920699227248856,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010433800271125169,
+      "loss": 0.9823,
+      "step": 6401
+    },
+    {
+      "epoch": 1.9209993247805537,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010432294020183764,
+      "loss": 0.8846,
+      "step": 6402
+    },
+    {
+      "epoch": 1.9212994223122515,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010430787769242356,
+      "loss": 0.9147,
+      "step": 6403
+    },
+    {
+      "epoch": 1.9215995198439493,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010429281518300948,
+      "loss": 1.0622,
+      "step": 6404
+    },
+    {
+      "epoch": 1.921899617375647,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010427775267359542,
+      "loss": 0.9373,
+      "step": 6405
+    },
+    {
+      "epoch": 1.9221997149073449,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010426269016418135,
+      "loss": 0.9425,
+      "step": 6406
+    },
+    {
+      "epoch": 1.9224998124390427,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001042476276547673,
+      "loss": 0.9747,
+      "step": 6407
+    },
+    {
+      "epoch": 1.9227999099707405,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010423256514535322,
+      "loss": 0.9927,
+      "step": 6408
+    },
+    {
+      "epoch": 1.9231000075024383,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010421750263593915,
+      "loss": 1.0198,
+      "step": 6409
+    },
+    {
+      "epoch": 1.923400105034136,
+      "grad_norm": 0.98828125,
+      "learning_rate": 0.00010420244012652509,
+      "loss": 0.9927,
+      "step": 6410
+    },
+    {
+      "epoch": 1.9237002025658338,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010418737761711101,
+      "loss": 0.9669,
+      "step": 6411
+    },
+    {
+      "epoch": 1.9240003000975316,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00010417231510769694,
+      "loss": 1.1559,
+      "step": 6412
+    },
+    {
+      "epoch": 1.9243003976292297,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010415725259828288,
+      "loss": 0.9787,
+      "step": 6413
+    },
+    {
+      "epoch": 1.9246004951609272,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001041421900888688,
+      "loss": 1.1175,
+      "step": 6414
+    },
+    {
+      "epoch": 1.9249005926926253,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010412712757945475,
+      "loss": 1.1262,
+      "step": 6415
+    },
+    {
+      "epoch": 1.9252006902243228,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010411206507004067,
+      "loss": 0.8996,
+      "step": 6416
+    },
+    {
+      "epoch": 1.9255007877560208,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.0001040970025606266,
+      "loss": 0.9806,
+      "step": 6417
+    },
+    {
+      "epoch": 1.9258008852877184,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010408194005121254,
+      "loss": 1.0138,
+      "step": 6418
+    },
+    {
+      "epoch": 1.9261009828194164,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010406687754179847,
+      "loss": 1.0571,
+      "step": 6419
+    },
+    {
+      "epoch": 1.926401080351114,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.0001040518150323844,
+      "loss": 0.8105,
+      "step": 6420
+    },
+    {
+      "epoch": 1.926701177882812,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010403675252297033,
+      "loss": 0.8318,
+      "step": 6421
+    },
+    {
+      "epoch": 1.9270012754145096,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010402169001355626,
+      "loss": 0.9407,
+      "step": 6422
+    },
+    {
+      "epoch": 1.9273013729462076,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00010400662750414219,
+      "loss": 0.941,
+      "step": 6423
+    },
+    {
+      "epoch": 1.9276014704779052,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010399156499472813,
+      "loss": 1.0017,
+      "step": 6424
+    },
+    {
+      "epoch": 1.9279015680096032,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010397650248531405,
+      "loss": 0.996,
+      "step": 6425
+    },
+    {
+      "epoch": 1.9282016655413008,
+      "grad_norm": 0.87109375,
+      "learning_rate": 0.0001039614399759,
+      "loss": 0.8407,
+      "step": 6426
+    },
+    {
+      "epoch": 1.9285017630729988,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00010394637746648592,
+      "loss": 1.0338,
+      "step": 6427
+    },
+    {
+      "epoch": 1.9288018606046964,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00010393131495707185,
+      "loss": 0.9634,
+      "step": 6428
+    },
+    {
+      "epoch": 1.9291019581363944,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010391625244765779,
+      "loss": 1.0736,
+      "step": 6429
+    },
+    {
+      "epoch": 1.929402055668092,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00010390118993824372,
+      "loss": 0.9149,
+      "step": 6430
+    },
+    {
+      "epoch": 1.92970215319979,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010388612742882964,
+      "loss": 1.1327,
+      "step": 6431
+    },
+    {
+      "epoch": 1.9300022507314876,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010387106491941558,
+      "loss": 1.0125,
+      "step": 6432
+    },
+    {
+      "epoch": 1.9303023482631856,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010385600241000151,
+      "loss": 1.034,
+      "step": 6433
+    },
+    {
+      "epoch": 1.9306024457948834,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010384093990058745,
+      "loss": 0.9829,
+      "step": 6434
+    },
+    {
+      "epoch": 1.9309025433265812,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00010382587739117338,
+      "loss": 1.0408,
+      "step": 6435
+    },
+    {
+      "epoch": 1.931202640858279,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001038108148817593,
+      "loss": 1.1435,
+      "step": 6436
+    },
+    {
+      "epoch": 1.9315027383899768,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010379575237234524,
+      "loss": 0.9865,
+      "step": 6437
+    },
+    {
+      "epoch": 1.9318028359216746,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010378068986293117,
+      "loss": 1.0695,
+      "step": 6438
+    },
+    {
+      "epoch": 1.9321029334533724,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.0001037656273535171,
+      "loss": 0.8975,
+      "step": 6439
+    },
+    {
+      "epoch": 1.9324030309850702,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010375056484410304,
+      "loss": 0.8864,
+      "step": 6440
+    },
+    {
+      "epoch": 1.932703128516768,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010373550233468896,
+      "loss": 0.9884,
+      "step": 6441
+    },
+    {
+      "epoch": 1.9330032260484658,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001037204398252749,
+      "loss": 0.9573,
+      "step": 6442
+    },
+    {
+      "epoch": 1.9333033235801635,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00010370537731586083,
+      "loss": 0.9657,
+      "step": 6443
+    },
+    {
+      "epoch": 1.9336034211118613,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010369031480644676,
+      "loss": 0.9164,
+      "step": 6444
+    },
+    {
+      "epoch": 1.9339035186435591,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001036752522970327,
+      "loss": 0.9466,
+      "step": 6445
+    },
+    {
+      "epoch": 1.934203616175257,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00010366018978761862,
+      "loss": 0.9493,
+      "step": 6446
+    },
+    {
+      "epoch": 1.9345037137069547,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010364512727820455,
+      "loss": 0.9477,
+      "step": 6447
+    },
+    {
+      "epoch": 1.9348038112386525,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00010363006476879049,
+      "loss": 1.022,
+      "step": 6448
+    },
+    {
+      "epoch": 1.9351039087703503,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010361500225937642,
+      "loss": 1.0788,
+      "step": 6449
+    },
+    {
+      "epoch": 1.9354040063020481,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00010359993974996233,
+      "loss": 1.0739,
+      "step": 6450
+    },
+    {
+      "epoch": 1.935704103833746,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010358487724054829,
+      "loss": 0.9702,
+      "step": 6451
+    },
+    {
+      "epoch": 1.936004201365444,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010356981473113421,
+      "loss": 1.0931,
+      "step": 6452
+    },
+    {
+      "epoch": 1.9363042988971415,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010355475222172015,
+      "loss": 1.0571,
+      "step": 6453
+    },
+    {
+      "epoch": 1.9366043964288395,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010353968971230608,
+      "loss": 1.0218,
+      "step": 6454
+    },
+    {
+      "epoch": 1.936904493960537,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.000103524627202892,
+      "loss": 1.0337,
+      "step": 6455
+    },
+    {
+      "epoch": 1.9372045914922351,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010350956469347795,
+      "loss": 1.0809,
+      "step": 6456
+    },
+    {
+      "epoch": 1.9375046890239327,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010349450218406387,
+      "loss": 1.1126,
+      "step": 6457
+    },
+    {
+      "epoch": 1.9378047865556307,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00010347943967464979,
+      "loss": 1.0193,
+      "step": 6458
+    },
+    {
+      "epoch": 1.9381048840873283,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00010346437716523574,
+      "loss": 0.9742,
+      "step": 6459
+    },
+    {
+      "epoch": 1.9384049816190263,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00010344931465582167,
+      "loss": 0.9467,
+      "step": 6460
+    },
+    {
+      "epoch": 1.9387050791507239,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010343425214640761,
+      "loss": 1.0162,
+      "step": 6461
+    },
+    {
+      "epoch": 1.939005176682422,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010341918963699353,
+      "loss": 0.9774,
+      "step": 6462
+    },
+    {
+      "epoch": 1.9393052742141195,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010340412712757945,
+      "loss": 0.9726,
+      "step": 6463
+    },
+    {
+      "epoch": 1.9396053717458175,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.0001033890646181654,
+      "loss": 0.9277,
+      "step": 6464
+    },
+    {
+      "epoch": 1.939905469277515,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010337400210875133,
+      "loss": 0.9229,
+      "step": 6465
+    },
+    {
+      "epoch": 1.940205566809213,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010335893959933724,
+      "loss": 1.0128,
+      "step": 6466
+    },
+    {
+      "epoch": 1.9405056643409107,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001033438770899232,
+      "loss": 0.9645,
+      "step": 6467
+    },
+    {
+      "epoch": 1.9408057618726087,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010332881458050911,
+      "loss": 0.8655,
+      "step": 6468
+    },
+    {
+      "epoch": 1.9411058594043062,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010331375207109506,
+      "loss": 1.06,
+      "step": 6469
+    },
+    {
+      "epoch": 1.9414059569360043,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010329868956168099,
+      "loss": 1.064,
+      "step": 6470
+    },
+    {
+      "epoch": 1.941706054467702,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001032836270522669,
+      "loss": 0.9625,
+      "step": 6471
+    },
+    {
+      "epoch": 1.9420061519993999,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010326856454285286,
+      "loss": 1.0543,
+      "step": 6472
+    },
+    {
+      "epoch": 1.9423062495310977,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010325350203343878,
+      "loss": 1.0248,
+      "step": 6473
+    },
+    {
+      "epoch": 1.9426063470627954,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.0001032384395240247,
+      "loss": 1.0152,
+      "step": 6474
+    },
+    {
+      "epoch": 1.9429064445944932,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010322337701461065,
+      "loss": 0.9998,
+      "step": 6475
+    },
+    {
+      "epoch": 1.943206542126191,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00010320831450519656,
+      "loss": 1.0415,
+      "step": 6476
+    },
+    {
+      "epoch": 1.9435066396578888,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010319325199578249,
+      "loss": 0.9537,
+      "step": 6477
+    },
+    {
+      "epoch": 1.9438067371895866,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010317818948636844,
+      "loss": 1.0185,
+      "step": 6478
+    },
+    {
+      "epoch": 1.9441068347212844,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010316312697695436,
+      "loss": 1.0559,
+      "step": 6479
+    },
+    {
+      "epoch": 1.9444069322529822,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010314806446754031,
+      "loss": 0.9051,
+      "step": 6480
+    },
+    {
+      "epoch": 1.94470702978468,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00010313300195812622,
+      "loss": 1.0669,
+      "step": 6481
+    },
+    {
+      "epoch": 1.9450071273163778,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010311793944871215,
+      "loss": 1.0479,
+      "step": 6482
+    },
+    {
+      "epoch": 1.9453072248480756,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.0001031028769392981,
+      "loss": 0.8906,
+      "step": 6483
+    },
+    {
+      "epoch": 1.9456073223797734,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010308781442988402,
+      "loss": 0.9074,
+      "step": 6484
+    },
+    {
+      "epoch": 1.9459074199114712,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010307275192046994,
+      "loss": 0.975,
+      "step": 6485
+    },
+    {
+      "epoch": 1.946207517443169,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010305768941105588,
+      "loss": 0.9904,
+      "step": 6486
+    },
+    {
+      "epoch": 1.9465076149748668,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00010304262690164181,
+      "loss": 0.9518,
+      "step": 6487
+    },
+    {
+      "epoch": 1.9468077125065646,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010302756439222777,
+      "loss": 1.0357,
+      "step": 6488
+    },
+    {
+      "epoch": 1.9471078100382624,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010301250188281368,
+      "loss": 0.8796,
+      "step": 6489
+    },
+    {
+      "epoch": 1.9474079075699602,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.0001029974393733996,
+      "loss": 0.9174,
+      "step": 6490
+    },
+    {
+      "epoch": 1.9477080051016582,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010298237686398556,
+      "loss": 0.9356,
+      "step": 6491
+    },
+    {
+      "epoch": 1.9480081026333558,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010296731435457147,
+      "loss": 0.9679,
+      "step": 6492
+    },
+    {
+      "epoch": 1.9483082001650538,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001029522518451574,
+      "loss": 1.1511,
+      "step": 6493
+    },
+    {
+      "epoch": 1.9486082976967514,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00010293718933574334,
+      "loss": 1.1627,
+      "step": 6494
+    },
+    {
+      "epoch": 1.9489083952284494,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010292212682632927,
+      "loss": 0.9581,
+      "step": 6495
+    },
+    {
+      "epoch": 1.949208492760147,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010290706431691522,
+      "loss": 1.0163,
+      "step": 6496
+    },
+    {
+      "epoch": 1.949508590291845,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010289200180750113,
+      "loss": 1.018,
+      "step": 6497
+    },
+    {
+      "epoch": 1.9498086878235426,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010287693929808706,
+      "loss": 1.1095,
+      "step": 6498
+    },
+    {
+      "epoch": 1.9501087853552406,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.000102861876788673,
+      "loss": 0.9695,
+      "step": 6499
+    },
+    {
+      "epoch": 1.9504088828869381,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010284681427925893,
+      "loss": 1.0028,
+      "step": 6500
+    },
+    {
+      "epoch": 1.9507089804186362,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010283175176984485,
+      "loss": 1.0018,
+      "step": 6501
+    },
+    {
+      "epoch": 1.9510090779503337,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.0001028166892604308,
+      "loss": 0.9274,
+      "step": 6502
+    },
+    {
+      "epoch": 1.9513091754820318,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010280162675101672,
+      "loss": 1.0011,
+      "step": 6503
+    },
+    {
+      "epoch": 1.9516092730137293,
+      "grad_norm": 0.96875,
+      "learning_rate": 0.00010278656424160265,
+      "loss": 0.9671,
+      "step": 6504
+    },
+    {
+      "epoch": 1.9519093705454273,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010277150173218859,
+      "loss": 0.953,
+      "step": 6505
+    },
+    {
+      "epoch": 1.952209468077125,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010275643922277451,
+      "loss": 1.0508,
+      "step": 6506
+    },
+    {
+      "epoch": 1.952509565608823,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010274137671336045,
+      "loss": 0.9578,
+      "step": 6507
+    },
+    {
+      "epoch": 1.9528096631405205,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010272631420394638,
+      "loss": 0.9305,
+      "step": 6508
+    },
+    {
+      "epoch": 1.9531097606722185,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010271125169453231,
+      "loss": 0.8183,
+      "step": 6509
+    },
+    {
+      "epoch": 1.9534098582039163,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010269618918511825,
+      "loss": 0.9385,
+      "step": 6510
+    },
+    {
+      "epoch": 1.9537099557356141,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010268112667570418,
+      "loss": 1.0295,
+      "step": 6511
+    },
+    {
+      "epoch": 1.954010053267312,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001026660641662901,
+      "loss": 1.0075,
+      "step": 6512
+    },
+    {
+      "epoch": 1.9543101507990097,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010265100165687604,
+      "loss": 1.0872,
+      "step": 6513
+    },
+    {
+      "epoch": 1.9546102483307075,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010263593914746197,
+      "loss": 0.8987,
+      "step": 6514
+    },
+    {
+      "epoch": 1.9549103458624053,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010262087663804791,
+      "loss": 1.094,
+      "step": 6515
+    },
+    {
+      "epoch": 1.955210443394103,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00010260581412863384,
+      "loss": 0.8628,
+      "step": 6516
+    },
+    {
+      "epoch": 1.955510540925801,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010259075161921976,
+      "loss": 0.9357,
+      "step": 6517
+    },
+    {
+      "epoch": 1.9558106384574987,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001025756891098057,
+      "loss": 1.0412,
+      "step": 6518
+    },
+    {
+      "epoch": 1.9561107359891965,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010256062660039163,
+      "loss": 1.01,
+      "step": 6519
+    },
+    {
+      "epoch": 1.9564108335208943,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010254556409097756,
+      "loss": 1.2504,
+      "step": 6520
+    },
+    {
+      "epoch": 1.956710931052592,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001025305015815635,
+      "loss": 1.0304,
+      "step": 6521
+    },
+    {
+      "epoch": 1.9570110285842899,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00010251543907214942,
+      "loss": 1.0211,
+      "step": 6522
+    },
+    {
+      "epoch": 1.9573111261159877,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010250037656273536,
+      "loss": 0.8713,
+      "step": 6523
+    },
+    {
+      "epoch": 1.9576112236476855,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010248531405332129,
+      "loss": 0.9766,
+      "step": 6524
+    },
+    {
+      "epoch": 1.9579113211793833,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.00010247025154390722,
+      "loss": 1.1396,
+      "step": 6525
+    },
+    {
+      "epoch": 1.958211418711081,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010245518903449316,
+      "loss": 0.9019,
+      "step": 6526
+    },
+    {
+      "epoch": 1.9585115162427789,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010244012652507908,
+      "loss": 0.9308,
+      "step": 6527
+    },
+    {
+      "epoch": 1.9588116137744767,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010242506401566501,
+      "loss": 0.9672,
+      "step": 6528
+    },
+    {
+      "epoch": 1.9591117113061745,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010241000150625095,
+      "loss": 0.9369,
+      "step": 6529
+    },
+    {
+      "epoch": 1.9594118088378725,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010239493899683688,
+      "loss": 0.9887,
+      "step": 6530
+    },
+    {
+      "epoch": 1.95971190636957,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001023798764874228,
+      "loss": 1.0523,
+      "step": 6531
+    },
+    {
+      "epoch": 1.960012003901268,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010236481397800875,
+      "loss": 0.886,
+      "step": 6532
+    },
+    {
+      "epoch": 1.9603121014329656,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010234975146859467,
+      "loss": 0.9649,
+      "step": 6533
+    },
+    {
+      "epoch": 1.9606121989646637,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010233468895918061,
+      "loss": 1.0452,
+      "step": 6534
+    },
+    {
+      "epoch": 1.9609122964963612,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010231962644976654,
+      "loss": 1.0367,
+      "step": 6535
+    },
+    {
+      "epoch": 1.9612123940280592,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010230456394035247,
+      "loss": 0.9434,
+      "step": 6536
+    },
+    {
+      "epoch": 1.9615124915597568,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001022895014309384,
+      "loss": 0.9839,
+      "step": 6537
+    },
+    {
+      "epoch": 1.9618125890914548,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00010227443892152433,
+      "loss": 0.9823,
+      "step": 6538
+    },
+    {
+      "epoch": 1.9621126866231524,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010225937641211025,
+      "loss": 0.9332,
+      "step": 6539
+    },
+    {
+      "epoch": 1.9624127841548504,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001022443139026962,
+      "loss": 0.9889,
+      "step": 6540
+    },
+    {
+      "epoch": 1.962712881686548,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00010222925139328213,
+      "loss": 0.9559,
+      "step": 6541
+    },
+    {
+      "epoch": 1.963012979218246,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010221418888386807,
+      "loss": 0.822,
+      "step": 6542
+    },
+    {
+      "epoch": 1.9633130767499436,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.000102199126374454,
+      "loss": 0.9535,
+      "step": 6543
+    },
+    {
+      "epoch": 1.9636131742816416,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010218406386503991,
+      "loss": 0.8684,
+      "step": 6544
+    },
+    {
+      "epoch": 1.9639132718133392,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010216900135562586,
+      "loss": 1.1138,
+      "step": 6545
+    },
+    {
+      "epoch": 1.9642133693450372,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010215393884621179,
+      "loss": 1.0078,
+      "step": 6546
+    },
+    {
+      "epoch": 1.9645134668767348,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.0001021388763367977,
+      "loss": 1.0197,
+      "step": 6547
+    },
+    {
+      "epoch": 1.9648135644084328,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010212381382738365,
+      "loss": 1.039,
+      "step": 6548
+    },
+    {
+      "epoch": 1.9651136619401306,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010210875131796958,
+      "loss": 1.0656,
+      "step": 6549
+    },
+    {
+      "epoch": 1.9654137594718284,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010209368880855552,
+      "loss": 0.9502,
+      "step": 6550
+    },
+    {
+      "epoch": 1.9657138570035262,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010207862629914145,
+      "loss": 0.923,
+      "step": 6551
+    },
+    {
+      "epoch": 1.966013954535224,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010206356378972736,
+      "loss": 0.9479,
+      "step": 6552
+    },
+    {
+      "epoch": 1.9663140520669218,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010204850128031332,
+      "loss": 0.8754,
+      "step": 6553
+    },
+    {
+      "epoch": 1.9666141495986196,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010203343877089924,
+      "loss": 0.961,
+      "step": 6554
+    },
+    {
+      "epoch": 1.9669142471303174,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010201837626148516,
+      "loss": 0.9371,
+      "step": 6555
+    },
+    {
+      "epoch": 1.9672143446620152,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010200331375207111,
+      "loss": 0.9073,
+      "step": 6556
+    },
+    {
+      "epoch": 1.967514442193713,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010198825124265702,
+      "loss": 1.0242,
+      "step": 6557
+    },
+    {
+      "epoch": 1.9678145397254108,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010197318873324295,
+      "loss": 0.961,
+      "step": 6558
+    },
+    {
+      "epoch": 1.9681146372571086,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.0001019581262238289,
+      "loss": 1.0004,
+      "step": 6559
+    },
+    {
+      "epoch": 1.9684147347888064,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010194306371441482,
+      "loss": 0.8086,
+      "step": 6560
+    },
+    {
+      "epoch": 1.9687148323205041,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010192800120500077,
+      "loss": 0.9333,
+      "step": 6561
+    },
+    {
+      "epoch": 1.969014929852202,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010191293869558668,
+      "loss": 0.9772,
+      "step": 6562
+    },
+    {
+      "epoch": 1.9693150273838997,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010189787618617261,
+      "loss": 0.9959,
+      "step": 6563
+    },
+    {
+      "epoch": 1.9696151249155975,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00010188281367675856,
+      "loss": 0.9828,
+      "step": 6564
+    },
+    {
+      "epoch": 1.9699152224472953,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00010186775116734448,
+      "loss": 1.0263,
+      "step": 6565
+    },
+    {
+      "epoch": 1.9702153199789931,
+      "grad_norm": 0.99609375,
+      "learning_rate": 0.0001018526886579304,
+      "loss": 1.1725,
+      "step": 6566
+    },
+    {
+      "epoch": 1.970515417510691,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010183762614851636,
+      "loss": 0.787,
+      "step": 6567
+    },
+    {
+      "epoch": 1.9708155150423887,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010182256363910227,
+      "loss": 0.9796,
+      "step": 6568
+    },
+    {
+      "epoch": 1.9711156125740867,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010180750112968823,
+      "loss": 1.0046,
+      "step": 6569
+    },
+    {
+      "epoch": 1.9714157101057843,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010179243862027414,
+      "loss": 0.9164,
+      "step": 6570
+    },
+    {
+      "epoch": 1.9717158076374823,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010177737611086007,
+      "loss": 0.9909,
+      "step": 6571
+    },
+    {
+      "epoch": 1.97201590516918,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00010176231360144602,
+      "loss": 0.9987,
+      "step": 6572
+    },
+    {
+      "epoch": 1.972316002700878,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010174725109203193,
+      "loss": 1.1027,
+      "step": 6573
+    },
+    {
+      "epoch": 1.9726161002325755,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010173218858261786,
+      "loss": 0.8751,
+      "step": 6574
+    },
+    {
+      "epoch": 1.9729161977642735,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.0001017171260732038,
+      "loss": 0.9783,
+      "step": 6575
+    },
+    {
+      "epoch": 1.973216295295971,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00010170206356378973,
+      "loss": 1.0557,
+      "step": 6576
+    },
+    {
+      "epoch": 1.973516392827669,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010168700105437568,
+      "loss": 0.9852,
+      "step": 6577
+    },
+    {
+      "epoch": 1.9738164903593667,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010167193854496159,
+      "loss": 0.9574,
+      "step": 6578
+    },
+    {
+      "epoch": 1.9741165878910647,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00010165687603554752,
+      "loss": 1.0934,
+      "step": 6579
+    },
+    {
+      "epoch": 1.9744166854227623,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00010164181352613347,
+      "loss": 1.1087,
+      "step": 6580
+    },
+    {
+      "epoch": 1.9747167829544603,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00010162675101671939,
+      "loss": 0.9417,
+      "step": 6581
+    },
+    {
+      "epoch": 1.9750168804861579,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00010161168850730531,
+      "loss": 0.8939,
+      "step": 6582
+    },
+    {
+      "epoch": 1.9753169780178559,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010159662599789125,
+      "loss": 0.8787,
+      "step": 6583
+    },
+    {
+      "epoch": 1.9756170755495535,
+      "grad_norm": 0.94140625,
+      "learning_rate": 0.00010158156348847718,
+      "loss": 0.9561,
+      "step": 6584
+    },
+    {
+      "epoch": 1.9759171730812515,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010156650097906311,
+      "loss": 0.999,
+      "step": 6585
+    },
+    {
+      "epoch": 1.976217270612949,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010155143846964905,
+      "loss": 0.9186,
+      "step": 6586
+    },
+    {
+      "epoch": 1.976517368144647,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010153637596023497,
+      "loss": 1.0361,
+      "step": 6587
+    },
+    {
+      "epoch": 1.9768174656763449,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010152131345082091,
+      "loss": 1.0396,
+      "step": 6588
+    },
+    {
+      "epoch": 1.9771175632080427,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010150625094140684,
+      "loss": 0.9427,
+      "step": 6589
+    },
+    {
+      "epoch": 1.9774176607397405,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010149118843199277,
+      "loss": 1.119,
+      "step": 6590
+    },
+    {
+      "epoch": 1.9777177582714383,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010147612592257871,
+      "loss": 1.0111,
+      "step": 6591
+    },
+    {
+      "epoch": 1.978017855803136,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00010146106341316464,
+      "loss": 0.9943,
+      "step": 6592
+    },
+    {
+      "epoch": 1.9783179533348338,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010144600090375056,
+      "loss": 1.0096,
+      "step": 6593
+    },
+    {
+      "epoch": 1.9786180508665316,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001014309383943365,
+      "loss": 1.0362,
+      "step": 6594
+    },
+    {
+      "epoch": 1.9789181483982294,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010141587588492243,
+      "loss": 1.1049,
+      "step": 6595
+    },
+    {
+      "epoch": 1.9792182459299272,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010140081337550837,
+      "loss": 0.9622,
+      "step": 6596
+    },
+    {
+      "epoch": 1.979518343461625,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.0001013857508660943,
+      "loss": 0.9976,
+      "step": 6597
+    },
+    {
+      "epoch": 1.9798184409933228,
+      "grad_norm": 0.90625,
+      "learning_rate": 0.00010137068835668022,
+      "loss": 0.9078,
+      "step": 6598
+    },
+    {
+      "epoch": 1.9801185385250206,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010135562584726616,
+      "loss": 0.9706,
+      "step": 6599
+    },
+    {
+      "epoch": 1.9804186360567184,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010134056333785209,
+      "loss": 1.0162,
+      "step": 6600
+    },
+    {
+      "epoch": 1.9807187335884162,
+      "grad_norm": 0.90234375,
+      "learning_rate": 0.00010132550082843802,
+      "loss": 1.0584,
+      "step": 6601
+    },
+    {
+      "epoch": 1.981018831120114,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010131043831902396,
+      "loss": 1.0245,
+      "step": 6602
+    },
+    {
+      "epoch": 1.9813189286518118,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010129537580960988,
+      "loss": 1.0877,
+      "step": 6603
+    },
+    {
+      "epoch": 1.9816190261835096,
+      "grad_norm": 0.8984375,
+      "learning_rate": 0.00010128031330019582,
+      "loss": 0.8722,
+      "step": 6604
+    },
+    {
+      "epoch": 1.9819191237152074,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010126525079078175,
+      "loss": 1.0286,
+      "step": 6605
+    },
+    {
+      "epoch": 1.9822192212469052,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010125018828136768,
+      "loss": 0.9021,
+      "step": 6606
+    },
+    {
+      "epoch": 1.982519318778603,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010123512577195362,
+      "loss": 0.9946,
+      "step": 6607
+    },
+    {
+      "epoch": 1.982819416310301,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00010122006326253954,
+      "loss": 1.0394,
+      "step": 6608
+    },
+    {
+      "epoch": 1.9831195138419986,
+      "grad_norm": 0.94921875,
+      "learning_rate": 0.00010120500075312547,
+      "loss": 0.9365,
+      "step": 6609
+    },
+    {
+      "epoch": 1.9834196113736966,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010118993824371141,
+      "loss": 1.0331,
+      "step": 6610
+    },
+    {
+      "epoch": 1.9837197089053942,
+      "grad_norm": 0.89453125,
+      "learning_rate": 0.00010117487573429734,
+      "loss": 0.9425,
+      "step": 6611
+    },
+    {
+      "epoch": 1.9840198064370922,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010115981322488327,
+      "loss": 0.8651,
+      "step": 6612
+    },
+    {
+      "epoch": 1.9843199039687898,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001011447507154692,
+      "loss": 1.0657,
+      "step": 6613
+    },
+    {
+      "epoch": 1.9846200015004878,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00010112968820605513,
+      "loss": 1.0542,
+      "step": 6614
+    },
+    {
+      "epoch": 1.9849200990321854,
+      "grad_norm": 0.88671875,
+      "learning_rate": 0.00010111462569664107,
+      "loss": 0.9684,
+      "step": 6615
+    },
+    {
+      "epoch": 1.9852201965638834,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.000101099563187227,
+      "loss": 1.0599,
+      "step": 6616
+    },
+    {
+      "epoch": 1.985520294095581,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00010108450067781293,
+      "loss": 0.9861,
+      "step": 6617
+    },
+    {
+      "epoch": 1.985820391627279,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010106943816839887,
+      "loss": 1.064,
+      "step": 6618
+    },
+    {
+      "epoch": 1.9861204891589765,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001010543756589848,
+      "loss": 0.9888,
+      "step": 6619
+    },
+    {
+      "epoch": 1.9864205866906746,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010103931314957072,
+      "loss": 1.035,
+      "step": 6620
+    },
+    {
+      "epoch": 1.9867206842223721,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00010102425064015666,
+      "loss": 0.9545,
+      "step": 6621
+    },
+    {
+      "epoch": 1.9870207817540702,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010100918813074259,
+      "loss": 1.0408,
+      "step": 6622
+    },
+    {
+      "epoch": 1.9873208792857677,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010099412562132853,
+      "loss": 1.0247,
+      "step": 6623
+    },
+    {
+      "epoch": 1.9876209768174657,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00010097906311191445,
+      "loss": 1.0001,
+      "step": 6624
+    },
+    {
+      "epoch": 1.9879210743491633,
+      "grad_norm": 0.87890625,
+      "learning_rate": 0.00010096400060250038,
+      "loss": 0.8493,
+      "step": 6625
+    },
+    {
+      "epoch": 1.9882211718808613,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00010094893809308632,
+      "loss": 0.9554,
+      "step": 6626
+    },
+    {
+      "epoch": 1.9885212694125591,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010093387558367225,
+      "loss": 0.9177,
+      "step": 6627
+    },
+    {
+      "epoch": 1.988821366944257,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010091881307425816,
+      "loss": 0.8738,
+      "step": 6628
+    },
+    {
+      "epoch": 1.9891214644759547,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00010090375056484412,
+      "loss": 1.1083,
+      "step": 6629
+    },
+    {
+      "epoch": 1.9894215620076525,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010088868805543004,
+      "loss": 0.8981,
+      "step": 6630
+    },
+    {
+      "epoch": 1.9897216595393503,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00010087362554601598,
+      "loss": 0.8221,
+      "step": 6631
+    },
+    {
+      "epoch": 1.9900217570710481,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010085856303660191,
+      "loss": 0.9971,
+      "step": 6632
+    },
+    {
+      "epoch": 1.990321854602746,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010084350052718782,
+      "loss": 0.9143,
+      "step": 6633
+    },
+    {
+      "epoch": 1.9906219521344437,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010082843801777378,
+      "loss": 0.946,
+      "step": 6634
+    },
+    {
+      "epoch": 1.9909220496661415,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.0001008133755083597,
+      "loss": 0.9722,
+      "step": 6635
+    },
+    {
+      "epoch": 1.9912221471978393,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00010079831299894562,
+      "loss": 0.9569,
+      "step": 6636
+    },
+    {
+      "epoch": 1.991522244729537,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010078325048953157,
+      "loss": 0.8724,
+      "step": 6637
+    },
+    {
+      "epoch": 1.991822342261235,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.0001007681879801175,
+      "loss": 0.9618,
+      "step": 6638
+    },
+    {
+      "epoch": 1.9921224397929327,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010075312547070341,
+      "loss": 0.9884,
+      "step": 6639
+    },
+    {
+      "epoch": 1.9924225373246305,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00010073806296128936,
+      "loss": 1.0999,
+      "step": 6640
+    },
+    {
+      "epoch": 1.9927226348563283,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010072300045187528,
+      "loss": 1.0748,
+      "step": 6641
+    },
+    {
+      "epoch": 1.993022732388026,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.00010070793794246123,
+      "loss": 0.9107,
+      "step": 6642
+    },
+    {
+      "epoch": 1.9933228299197239,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00010069287543304716,
+      "loss": 1.122,
+      "step": 6643
+    },
+    {
+      "epoch": 1.9936229274514217,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010067781292363307,
+      "loss": 1.0083,
+      "step": 6644
+    },
+    {
+      "epoch": 1.9939230249831195,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010066275041421902,
+      "loss": 1.0631,
+      "step": 6645
+    },
+    {
+      "epoch": 1.9942231225148173,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00010064768790480494,
+      "loss": 0.9963,
+      "step": 6646
+    },
+    {
+      "epoch": 1.9945232200465153,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010063262539539086,
+      "loss": 1.0052,
+      "step": 6647
+    },
+    {
+      "epoch": 1.9948233175782129,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00010061756288597682,
+      "loss": 0.9221,
+      "step": 6648
+    },
+    {
+      "epoch": 1.9951234151099109,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00010060250037656273,
+      "loss": 0.9707,
+      "step": 6649
+    },
+    {
+      "epoch": 1.9954235126416084,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00010058743786714869,
+      "loss": 1.1067,
+      "step": 6650
+    },
+    {
+      "epoch": 1.9957236101733065,
+      "grad_norm": 0.890625,
+      "learning_rate": 0.0001005723753577346,
+      "loss": 1.0249,
+      "step": 6651
+    },
+    {
+      "epoch": 1.996023707705004,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010055731284832053,
+      "loss": 0.9825,
+      "step": 6652
+    },
+    {
+      "epoch": 1.996323805236702,
+      "grad_norm": 0.8828125,
+      "learning_rate": 0.00010054225033890648,
+      "loss": 0.8915,
+      "step": 6653
+    },
+    {
+      "epoch": 1.9966239027683996,
+      "grad_norm": 0.9453125,
+      "learning_rate": 0.00010052718782949239,
+      "loss": 0.9883,
+      "step": 6654
+    },
+    {
+      "epoch": 1.9969240003000976,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010051212532007832,
+      "loss": 0.9897,
+      "step": 6655
+    },
+    {
+      "epoch": 1.9972240978317952,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010049706281066427,
+      "loss": 0.9971,
+      "step": 6656
+    },
+    {
+      "epoch": 1.9975241953634932,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00010048200030125019,
+      "loss": 0.8587,
+      "step": 6657
+    },
+    {
+      "epoch": 1.9978242928951908,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010046693779183614,
+      "loss": 0.9948,
+      "step": 6658
+    },
+    {
+      "epoch": 1.9981243904268888,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.00010045187528242205,
+      "loss": 0.9249,
+      "step": 6659
+    },
+    {
+      "epoch": 1.9984244879585864,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00010043681277300798,
+      "loss": 1.0826,
+      "step": 6660
+    },
+    {
+      "epoch": 1.9987245854902844,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.00010042175026359393,
+      "loss": 0.8099,
+      "step": 6661
+    },
+    {
+      "epoch": 1.999024683021982,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00010040668775417985,
+      "loss": 1.035,
+      "step": 6662
+    },
+    {
+      "epoch": 1.99932478055368,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00010039162524476577,
+      "loss": 0.9938,
+      "step": 6663
+    },
+    {
+      "epoch": 1.9996248780853776,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00010037656273535171,
+      "loss": 0.9379,
+      "step": 6664
+    },
+    {
+      "epoch": 1.9999249756170756,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00010036150022593764,
+      "loss": 0.9951,
+      "step": 6665
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00010034643771652357,
+      "loss": 0.7289,
+      "step": 6666
+    },
+    {
+      "epoch": 2.000300097531698,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00010033137520710951,
+      "loss": 0.7171,
+      "step": 6667
+    },
+    {
+      "epoch": 2.0006001950633956,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.00010031631269769543,
+      "loss": 0.7144,
+      "step": 6668
+    },
+    {
+      "epoch": 2.0009002925950936,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00010030125018828137,
+      "loss": 0.712,
+      "step": 6669
+    },
+    {
+      "epoch": 2.001200390126791,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001002861876788673,
+      "loss": 0.634,
+      "step": 6670
+    },
+    {
+      "epoch": 2.001500487658489,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00010027112516945323,
+      "loss": 0.5498,
+      "step": 6671
+    },
+    {
+      "epoch": 2.0018005851901868,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00010025606266003917,
+      "loss": 0.6628,
+      "step": 6672
+    },
+    {
+      "epoch": 2.002100682721885,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001002410001506251,
+      "loss": 0.6642,
+      "step": 6673
+    },
+    {
+      "epoch": 2.0024007802535824,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00010022593764121102,
+      "loss": 0.532,
+      "step": 6674
+    },
+    {
+      "epoch": 2.0027008777852804,
+      "grad_norm": 0.93359375,
+      "learning_rate": 0.00010021087513179696,
+      "loss": 0.5854,
+      "step": 6675
+    },
+    {
+      "epoch": 2.003000975316978,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.00010019581262238289,
+      "loss": 0.5534,
+      "step": 6676
+    },
+    {
+      "epoch": 2.003301072848676,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010018075011296883,
+      "loss": 0.5995,
+      "step": 6677
+    },
+    {
+      "epoch": 2.0036011703803736,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00010016568760355476,
+      "loss": 0.6571,
+      "step": 6678
+    },
+    {
+      "epoch": 2.0039012679120716,
+      "grad_norm": 0.8359375,
+      "learning_rate": 0.00010015062509414068,
+      "loss": 0.618,
+      "step": 6679
+    },
+    {
+      "epoch": 2.004201365443769,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00010013556258472662,
+      "loss": 0.6858,
+      "step": 6680
+    },
+    {
+      "epoch": 2.004501462975467,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00010012050007531255,
+      "loss": 0.58,
+      "step": 6681
+    },
+    {
+      "epoch": 2.0048015605071647,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00010010543756589848,
+      "loss": 0.5865,
+      "step": 6682
+    },
+    {
+      "epoch": 2.0051016580388628,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00010009037505648442,
+      "loss": 0.6914,
+      "step": 6683
+    },
+    {
+      "epoch": 2.0054017555705603,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00010007531254707034,
+      "loss": 0.6825,
+      "step": 6684
+    },
+    {
+      "epoch": 2.0057018531022583,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.00010006025003765628,
+      "loss": 0.7016,
+      "step": 6685
+    },
+    {
+      "epoch": 2.006001950633956,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00010004518752824221,
+      "loss": 0.4122,
+      "step": 6686
+    },
+    {
+      "epoch": 2.006302048165654,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00010003012501882814,
+      "loss": 0.5614,
+      "step": 6687
+    },
+    {
+      "epoch": 2.0066021456973515,
+      "grad_norm": 1.0234375,
+      "learning_rate": 0.00010001506250941408,
+      "loss": 0.7738,
+      "step": 6688
+    },
+    {
+      "epoch": 2.0069022432290495,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001,
+      "loss": 0.6826,
+      "step": 6689
+    },
+    {
+      "epoch": 2.007202340760747,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.998493749058595e-05,
+      "loss": 0.5685,
+      "step": 6690
+    },
+    {
+      "epoch": 2.007502438292445,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.996987498117186e-05,
+      "loss": 0.6282,
+      "step": 6691
+    },
+    {
+      "epoch": 2.0078025358241427,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.99548124717578e-05,
+      "loss": 0.5576,
+      "step": 6692
+    },
+    {
+      "epoch": 2.0081026333558407,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.993974996234373e-05,
+      "loss": 0.6175,
+      "step": 6693
+    },
+    {
+      "epoch": 2.0084027308875383,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.992468745292967e-05,
+      "loss": 0.6216,
+      "step": 6694
+    },
+    {
+      "epoch": 2.0087028284192363,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.990962494351559e-05,
+      "loss": 0.4999,
+      "step": 6695
+    },
+    {
+      "epoch": 2.009002925950934,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.989456243410152e-05,
+      "loss": 0.6181,
+      "step": 6696
+    },
+    {
+      "epoch": 2.009303023482632,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.987949992468746e-05,
+      "loss": 0.5356,
+      "step": 6697
+    },
+    {
+      "epoch": 2.0096031210143295,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.98644374152734e-05,
+      "loss": 0.5218,
+      "step": 6698
+    },
+    {
+      "epoch": 2.0099032185460275,
+      "grad_norm": 1.0,
+      "learning_rate": 9.984937490585931e-05,
+      "loss": 0.5243,
+      "step": 6699
+    },
+    {
+      "epoch": 2.010203316077725,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.983431239644525e-05,
+      "loss": 0.6915,
+      "step": 6700
+    },
+    {
+      "epoch": 2.010503413609423,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.981924988703118e-05,
+      "loss": 0.6032,
+      "step": 6701
+    },
+    {
+      "epoch": 2.0108035111411207,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.980418737761712e-05,
+      "loss": 0.567,
+      "step": 6702
+    },
+    {
+      "epoch": 2.0111036086728187,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.978912486820305e-05,
+      "loss": 0.7281,
+      "step": 6703
+    },
+    {
+      "epoch": 2.0114037062045163,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.977406235878897e-05,
+      "loss": 0.6087,
+      "step": 6704
+    },
+    {
+      "epoch": 2.0117038037362143,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.975899984937491e-05,
+      "loss": 0.6137,
+      "step": 6705
+    },
+    {
+      "epoch": 2.0120039012679123,
+      "grad_norm": 1.125,
+      "learning_rate": 9.974393733996084e-05,
+      "loss": 0.7059,
+      "step": 6706
+    },
+    {
+      "epoch": 2.01230399879961,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.972887483054677e-05,
+      "loss": 0.5535,
+      "step": 6707
+    },
+    {
+      "epoch": 2.012604096331308,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.971381232113271e-05,
+      "loss": 0.6406,
+      "step": 6708
+    },
+    {
+      "epoch": 2.0129041938630055,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.969874981171863e-05,
+      "loss": 0.577,
+      "step": 6709
+    },
+    {
+      "epoch": 2.0132042913947035,
+      "grad_norm": 0.875,
+      "learning_rate": 9.968368730230458e-05,
+      "loss": 0.6124,
+      "step": 6710
+    },
+    {
+      "epoch": 2.013504388926401,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.96686247928905e-05,
+      "loss": 0.5548,
+      "step": 6711
+    },
+    {
+      "epoch": 2.013804486458099,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.965356228347643e-05,
+      "loss": 0.5482,
+      "step": 6712
+    },
+    {
+      "epoch": 2.0141045839897966,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.963849977406237e-05,
+      "loss": 0.6466,
+      "step": 6713
+    },
+    {
+      "epoch": 2.0144046815214947,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.96234372646483e-05,
+      "loss": 0.672,
+      "step": 6714
+    },
+    {
+      "epoch": 2.0147047790531922,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.960837475523422e-05,
+      "loss": 0.5592,
+      "step": 6715
+    },
+    {
+      "epoch": 2.0150048765848902,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.959331224582016e-05,
+      "loss": 0.5784,
+      "step": 6716
+    },
+    {
+      "epoch": 2.015304974116588,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.957824973640609e-05,
+      "loss": 0.5164,
+      "step": 6717
+    },
+    {
+      "epoch": 2.015605071648286,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.956318722699202e-05,
+      "loss": 0.5556,
+      "step": 6718
+    },
+    {
+      "epoch": 2.0159051691799834,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.954812471757796e-05,
+      "loss": 0.5763,
+      "step": 6719
+    },
+    {
+      "epoch": 2.0162052667116814,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.953306220816388e-05,
+      "loss": 0.5467,
+      "step": 6720
+    },
+    {
+      "epoch": 2.016505364243379,
+      "grad_norm": 1.125,
+      "learning_rate": 9.951799969874982e-05,
+      "loss": 0.6786,
+      "step": 6721
+    },
+    {
+      "epoch": 2.016805461775077,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.950293718933574e-05,
+      "loss": 0.6294,
+      "step": 6722
+    },
+    {
+      "epoch": 2.0171055593067746,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.948787467992168e-05,
+      "loss": 0.5959,
+      "step": 6723
+    },
+    {
+      "epoch": 2.0174056568384726,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.947281217050762e-05,
+      "loss": 0.6029,
+      "step": 6724
+    },
+    {
+      "epoch": 2.01770575437017,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.945774966109354e-05,
+      "loss": 0.5786,
+      "step": 6725
+    },
+    {
+      "epoch": 2.018005851901868,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.944268715167947e-05,
+      "loss": 0.6487,
+      "step": 6726
+    },
+    {
+      "epoch": 2.018305949433566,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.942762464226541e-05,
+      "loss": 0.5917,
+      "step": 6727
+    },
+    {
+      "epoch": 2.018606046965264,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.941256213285134e-05,
+      "loss": 0.7317,
+      "step": 6728
+    },
+    {
+      "epoch": 2.0189061444969614,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.939749962343728e-05,
+      "loss": 0.5388,
+      "step": 6729
+    },
+    {
+      "epoch": 2.0192062420286594,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.938243711402319e-05,
+      "loss": 0.5958,
+      "step": 6730
+    },
+    {
+      "epoch": 2.019506339560357,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.936737460460913e-05,
+      "loss": 0.5142,
+      "step": 6731
+    },
+    {
+      "epoch": 2.019806437092055,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.935231209519507e-05,
+      "loss": 0.7185,
+      "step": 6732
+    },
+    {
+      "epoch": 2.0201065346237526,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.9337249585781e-05,
+      "loss": 0.5338,
+      "step": 6733
+    },
+    {
+      "epoch": 2.0204066321554506,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.932218707636693e-05,
+      "loss": 0.5863,
+      "step": 6734
+    },
+    {
+      "epoch": 2.020706729687148,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.930712456695285e-05,
+      "loss": 0.5892,
+      "step": 6735
+    },
+    {
+      "epoch": 2.021006827218846,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.929206205753879e-05,
+      "loss": 0.6232,
+      "step": 6736
+    },
+    {
+      "epoch": 2.0213069247505437,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.927699954812473e-05,
+      "loss": 0.6596,
+      "step": 6737
+    },
+    {
+      "epoch": 2.0216070222822418,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.926193703871065e-05,
+      "loss": 0.6239,
+      "step": 6738
+    },
+    {
+      "epoch": 2.0219071198139393,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.924687452929659e-05,
+      "loss": 0.6401,
+      "step": 6739
+    },
+    {
+      "epoch": 2.0222072173456374,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.923181201988251e-05,
+      "loss": 0.5744,
+      "step": 6740
+    },
+    {
+      "epoch": 2.022507314877335,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.921674951046845e-05,
+      "loss": 0.6213,
+      "step": 6741
+    },
+    {
+      "epoch": 2.022807412409033,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.920168700105438e-05,
+      "loss": 0.6936,
+      "step": 6742
+    },
+    {
+      "epoch": 2.0231075099407305,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.918662449164031e-05,
+      "loss": 0.5203,
+      "step": 6743
+    },
+    {
+      "epoch": 2.0234076074724285,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.917156198222625e-05,
+      "loss": 0.5814,
+      "step": 6744
+    },
+    {
+      "epoch": 2.0237077050041266,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.915649947281217e-05,
+      "loss": 0.5813,
+      "step": 6745
+    },
+    {
+      "epoch": 2.024007802535824,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.91414369633981e-05,
+      "loss": 0.5346,
+      "step": 6746
+    },
+    {
+      "epoch": 2.024307900067522,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.912637445398404e-05,
+      "loss": 0.5549,
+      "step": 6747
+    },
+    {
+      "epoch": 2.0246079975992197,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.911131194456997e-05,
+      "loss": 0.6252,
+      "step": 6748
+    },
+    {
+      "epoch": 2.0249080951309177,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.90962494351559e-05,
+      "loss": 0.4835,
+      "step": 6749
+    },
+    {
+      "epoch": 2.0252081926626153,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.908118692574183e-05,
+      "loss": 0.5883,
+      "step": 6750
+    },
+    {
+      "epoch": 2.0255082901943133,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.906612441632776e-05,
+      "loss": 0.6276,
+      "step": 6751
+    },
+    {
+      "epoch": 2.025808387726011,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.90510619069137e-05,
+      "loss": 0.7278,
+      "step": 6752
+    },
+    {
+      "epoch": 2.026108485257709,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.903599939749963e-05,
+      "loss": 0.6009,
+      "step": 6753
+    },
+    {
+      "epoch": 2.0264085827894065,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.902093688808556e-05,
+      "loss": 0.4843,
+      "step": 6754
+    },
+    {
+      "epoch": 2.0267086803211045,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.90058743786715e-05,
+      "loss": 0.6047,
+      "step": 6755
+    },
+    {
+      "epoch": 2.027008777852802,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.899081186925742e-05,
+      "loss": 0.5111,
+      "step": 6756
+    },
+    {
+      "epoch": 2.0273088753845,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.897574935984335e-05,
+      "loss": 0.5767,
+      "step": 6757
+    },
+    {
+      "epoch": 2.0276089729161977,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.896068685042929e-05,
+      "loss": 0.6274,
+      "step": 6758
+    },
+    {
+      "epoch": 2.0279090704478957,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.894562434101522e-05,
+      "loss": 0.5883,
+      "step": 6759
+    },
+    {
+      "epoch": 2.0282091679795933,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.893056183160116e-05,
+      "loss": 0.4776,
+      "step": 6760
+    },
+    {
+      "epoch": 2.0285092655112913,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.891549932218708e-05,
+      "loss": 0.5967,
+      "step": 6761
+    },
+    {
+      "epoch": 2.028809363042989,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.890043681277301e-05,
+      "loss": 0.573,
+      "step": 6762
+    },
+    {
+      "epoch": 2.029109460574687,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.888537430335895e-05,
+      "loss": 0.6042,
+      "step": 6763
+    },
+    {
+      "epoch": 2.0294095581063845,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.887031179394488e-05,
+      "loss": 0.5431,
+      "step": 6764
+    },
+    {
+      "epoch": 2.0297096556380825,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.88552492845308e-05,
+      "loss": 0.4754,
+      "step": 6765
+    },
+    {
+      "epoch": 2.03000975316978,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.884018677511674e-05,
+      "loss": 0.6577,
+      "step": 6766
+    },
+    {
+      "epoch": 2.030309850701478,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.882512426570267e-05,
+      "loss": 0.652,
+      "step": 6767
+    },
+    {
+      "epoch": 2.0306099482331756,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.881006175628861e-05,
+      "loss": 0.6513,
+      "step": 6768
+    },
+    {
+      "epoch": 2.0309100457648737,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.879499924687452e-05,
+      "loss": 0.6786,
+      "step": 6769
+    },
+    {
+      "epoch": 2.0312101432965712,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.877993673746046e-05,
+      "loss": 0.5528,
+      "step": 6770
+    },
+    {
+      "epoch": 2.0315102408282693,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.87648742280464e-05,
+      "loss": 0.6643,
+      "step": 6771
+    },
+    {
+      "epoch": 2.031810338359967,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.874981171863232e-05,
+      "loss": 0.5659,
+      "step": 6772
+    },
+    {
+      "epoch": 2.032110435891665,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.873474920921826e-05,
+      "loss": 0.5412,
+      "step": 6773
+    },
+    {
+      "epoch": 2.0324105334233624,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.87196866998042e-05,
+      "loss": 0.5905,
+      "step": 6774
+    },
+    {
+      "epoch": 2.0327106309550604,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.870462419039013e-05,
+      "loss": 0.5353,
+      "step": 6775
+    },
+    {
+      "epoch": 2.033010728486758,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.868956168097605e-05,
+      "loss": 0.6191,
+      "step": 6776
+    },
+    {
+      "epoch": 2.033310826018456,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.867449917156198e-05,
+      "loss": 0.5421,
+      "step": 6777
+    },
+    {
+      "epoch": 2.0336109235501536,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.865943666214792e-05,
+      "loss": 0.5664,
+      "step": 6778
+    },
+    {
+      "epoch": 2.0339110210818516,
+      "grad_norm": 1.0,
+      "learning_rate": 9.864437415273386e-05,
+      "loss": 0.5211,
+      "step": 6779
+    },
+    {
+      "epoch": 2.034211118613549,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.862931164331977e-05,
+      "loss": 0.6461,
+      "step": 6780
+    },
+    {
+      "epoch": 2.034511216145247,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.861424913390571e-05,
+      "loss": 0.7192,
+      "step": 6781
+    },
+    {
+      "epoch": 2.0348113136769452,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.859918662449164e-05,
+      "loss": 0.483,
+      "step": 6782
+    },
+    {
+      "epoch": 2.035111411208643,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.858412411507758e-05,
+      "loss": 0.5673,
+      "step": 6783
+    },
+    {
+      "epoch": 2.035411508740341,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.856906160566351e-05,
+      "loss": 0.5191,
+      "step": 6784
+    },
+    {
+      "epoch": 2.0357116062720384,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.855399909624943e-05,
+      "loss": 0.5277,
+      "step": 6785
+    },
+    {
+      "epoch": 2.0360117038037364,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.853893658683537e-05,
+      "loss": 0.5467,
+      "step": 6786
+    },
+    {
+      "epoch": 2.036311801335434,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.85238740774213e-05,
+      "loss": 0.6202,
+      "step": 6787
+    },
+    {
+      "epoch": 2.036611898867132,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.850881156800723e-05,
+      "loss": 0.5431,
+      "step": 6788
+    },
+    {
+      "epoch": 2.0369119963988296,
+      "grad_norm": 1.125,
+      "learning_rate": 9.849374905859317e-05,
+      "loss": 0.6548,
+      "step": 6789
+    },
+    {
+      "epoch": 2.0372120939305276,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.84786865491791e-05,
+      "loss": 0.505,
+      "step": 6790
+    },
+    {
+      "epoch": 2.037512191462225,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.846362403976504e-05,
+      "loss": 0.4846,
+      "step": 6791
+    },
+    {
+      "epoch": 2.037812288993923,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.844856153035096e-05,
+      "loss": 0.6105,
+      "step": 6792
+    },
+    {
+      "epoch": 2.0381123865256208,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.843349902093689e-05,
+      "loss": 0.6392,
+      "step": 6793
+    },
+    {
+      "epoch": 2.038412484057319,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.841843651152283e-05,
+      "loss": 0.596,
+      "step": 6794
+    },
+    {
+      "epoch": 2.0387125815890164,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.840337400210876e-05,
+      "loss": 0.5574,
+      "step": 6795
+    },
+    {
+      "epoch": 2.0390126791207144,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.838831149269468e-05,
+      "loss": 0.5652,
+      "step": 6796
+    },
+    {
+      "epoch": 2.039312776652412,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.837324898328062e-05,
+      "loss": 0.7139,
+      "step": 6797
+    },
+    {
+      "epoch": 2.03961287418411,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.835818647386655e-05,
+      "loss": 0.5807,
+      "step": 6798
+    },
+    {
+      "epoch": 2.0399129717158075,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.834312396445248e-05,
+      "loss": 0.6377,
+      "step": 6799
+    },
+    {
+      "epoch": 2.0402130692475056,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.832806145503842e-05,
+      "loss": 0.4835,
+      "step": 6800
+    },
+    {
+      "epoch": 2.040513166779203,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.831299894562434e-05,
+      "loss": 0.5582,
+      "step": 6801
+    },
+    {
+      "epoch": 2.040813264310901,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.829793643621028e-05,
+      "loss": 0.6779,
+      "step": 6802
+    },
+    {
+      "epoch": 2.0411133618425987,
+      "grad_norm": 1.125,
+      "learning_rate": 9.828287392679621e-05,
+      "loss": 0.6928,
+      "step": 6803
+    },
+    {
+      "epoch": 2.0414134593742967,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.826781141738214e-05,
+      "loss": 0.6577,
+      "step": 6804
+    },
+    {
+      "epoch": 2.0417135569059943,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.825274890796808e-05,
+      "loss": 0.6115,
+      "step": 6805
+    },
+    {
+      "epoch": 2.0420136544376923,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.8237686398554e-05,
+      "loss": 0.526,
+      "step": 6806
+    },
+    {
+      "epoch": 2.04231375196939,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.822262388913993e-05,
+      "loss": 0.6418,
+      "step": 6807
+    },
+    {
+      "epoch": 2.042613849501088,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.820756137972587e-05,
+      "loss": 0.5139,
+      "step": 6808
+    },
+    {
+      "epoch": 2.0429139470327855,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.81924988703118e-05,
+      "loss": 0.6292,
+      "step": 6809
+    },
+    {
+      "epoch": 2.0432140445644835,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.817743636089774e-05,
+      "loss": 0.532,
+      "step": 6810
+    },
+    {
+      "epoch": 2.043514142096181,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.816237385148365e-05,
+      "loss": 0.5865,
+      "step": 6811
+    },
+    {
+      "epoch": 2.043814239627879,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.814731134206959e-05,
+      "loss": 0.5676,
+      "step": 6812
+    },
+    {
+      "epoch": 2.0441143371595767,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.813224883265553e-05,
+      "loss": 0.5588,
+      "step": 6813
+    },
+    {
+      "epoch": 2.0444144346912747,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.811718632324146e-05,
+      "loss": 0.6391,
+      "step": 6814
+    },
+    {
+      "epoch": 2.0447145322229723,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.810212381382739e-05,
+      "loss": 0.702,
+      "step": 6815
+    },
+    {
+      "epoch": 2.0450146297546703,
+      "grad_norm": 0.875,
+      "learning_rate": 9.808706130441333e-05,
+      "loss": 0.54,
+      "step": 6816
+    },
+    {
+      "epoch": 2.045314727286368,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.807199879499925e-05,
+      "loss": 0.6696,
+      "step": 6817
+    },
+    {
+      "epoch": 2.045614824818066,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.805693628558519e-05,
+      "loss": 0.5428,
+      "step": 6818
+    },
+    {
+      "epoch": 2.0459149223497635,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.80418737761711e-05,
+      "loss": 0.5538,
+      "step": 6819
+    },
+    {
+      "epoch": 2.0462150198814615,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.802681126675705e-05,
+      "loss": 0.4848,
+      "step": 6820
+    },
+    {
+      "epoch": 2.0465151174131595,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.801174875734299e-05,
+      "loss": 0.4823,
+      "step": 6821
+    },
+    {
+      "epoch": 2.046815214944857,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.799668624792891e-05,
+      "loss": 0.5074,
+      "step": 6822
+    },
+    {
+      "epoch": 2.047115312476555,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.798162373851484e-05,
+      "loss": 0.494,
+      "step": 6823
+    },
+    {
+      "epoch": 2.0474154100082527,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.796656122910077e-05,
+      "loss": 0.6122,
+      "step": 6824
+    },
+    {
+      "epoch": 2.0477155075399507,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.795149871968671e-05,
+      "loss": 0.6273,
+      "step": 6825
+    },
+    {
+      "epoch": 2.0480156050716483,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.793643621027263e-05,
+      "loss": 0.5925,
+      "step": 6826
+    },
+    {
+      "epoch": 2.0483157026033463,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.792137370085856e-05,
+      "loss": 0.5193,
+      "step": 6827
+    },
+    {
+      "epoch": 2.048615800135044,
+      "grad_norm": 0.90625,
+      "learning_rate": 9.79063111914445e-05,
+      "loss": 0.5981,
+      "step": 6828
+    },
+    {
+      "epoch": 2.048915897666742,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.789124868203043e-05,
+      "loss": 0.5735,
+      "step": 6829
+    },
+    {
+      "epoch": 2.0492159951984394,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.787618617261635e-05,
+      "loss": 0.5964,
+      "step": 6830
+    },
+    {
+      "epoch": 2.0495160927301375,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.78611236632023e-05,
+      "loss": 0.5168,
+      "step": 6831
+    },
+    {
+      "epoch": 2.049816190261835,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.784606115378822e-05,
+      "loss": 0.6591,
+      "step": 6832
+    },
+    {
+      "epoch": 2.050116287793533,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.783099864437416e-05,
+      "loss": 0.4954,
+      "step": 6833
+    },
+    {
+      "epoch": 2.0504163853252306,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.781593613496009e-05,
+      "loss": 0.508,
+      "step": 6834
+    },
+    {
+      "epoch": 2.0507164828569286,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.780087362554602e-05,
+      "loss": 0.6849,
+      "step": 6835
+    },
+    {
+      "epoch": 2.051016580388626,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.778581111613196e-05,
+      "loss": 0.5225,
+      "step": 6836
+    },
+    {
+      "epoch": 2.0513166779203242,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.777074860671788e-05,
+      "loss": 0.648,
+      "step": 6837
+    },
+    {
+      "epoch": 2.051616775452022,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.775568609730381e-05,
+      "loss": 0.6929,
+      "step": 6838
+    },
+    {
+      "epoch": 2.05191687298372,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.774062358788975e-05,
+      "loss": 0.5471,
+      "step": 6839
+    },
+    {
+      "epoch": 2.0522169705154174,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.772556107847568e-05,
+      "loss": 0.6303,
+      "step": 6840
+    },
+    {
+      "epoch": 2.0525170680471154,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.771049856906162e-05,
+      "loss": 0.4457,
+      "step": 6841
+    },
+    {
+      "epoch": 2.052817165578813,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.769543605964754e-05,
+      "loss": 0.593,
+      "step": 6842
+    },
+    {
+      "epoch": 2.053117263110511,
+      "grad_norm": 1.0,
+      "learning_rate": 9.768037355023347e-05,
+      "loss": 0.6567,
+      "step": 6843
+    },
+    {
+      "epoch": 2.0534173606422086,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.766531104081941e-05,
+      "loss": 0.6885,
+      "step": 6844
+    },
+    {
+      "epoch": 2.0537174581739066,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.765024853140534e-05,
+      "loss": 0.5881,
+      "step": 6845
+    },
+    {
+      "epoch": 2.054017555705604,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.763518602199126e-05,
+      "loss": 0.6009,
+      "step": 6846
+    },
+    {
+      "epoch": 2.054317653237302,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.76201235125772e-05,
+      "loss": 0.5676,
+      "step": 6847
+    },
+    {
+      "epoch": 2.0546177507689998,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.760506100316313e-05,
+      "loss": 0.4887,
+      "step": 6848
+    },
+    {
+      "epoch": 2.054917848300698,
+      "grad_norm": 1.0,
+      "learning_rate": 9.758999849374907e-05,
+      "loss": 0.6433,
+      "step": 6849
+    },
+    {
+      "epoch": 2.0552179458323954,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.7574935984335e-05,
+      "loss": 0.5717,
+      "step": 6850
+    },
+    {
+      "epoch": 2.0555180433640934,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.755987347492092e-05,
+      "loss": 0.609,
+      "step": 6851
+    },
+    {
+      "epoch": 2.055818140895791,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.754481096550687e-05,
+      "loss": 0.5438,
+      "step": 6852
+    },
+    {
+      "epoch": 2.056118238427489,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.752974845609278e-05,
+      "loss": 0.5926,
+      "step": 6853
+    },
+    {
+      "epoch": 2.0564183359591865,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.751468594667872e-05,
+      "loss": 0.5905,
+      "step": 6854
+    },
+    {
+      "epoch": 2.0567184334908846,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.749962343726466e-05,
+      "loss": 0.5359,
+      "step": 6855
+    },
+    {
+      "epoch": 2.057018531022582,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.748456092785059e-05,
+      "loss": 0.4849,
+      "step": 6856
+    },
+    {
+      "epoch": 2.05731862855428,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.746949841843651e-05,
+      "loss": 0.5883,
+      "step": 6857
+    },
+    {
+      "epoch": 2.0576187260859777,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.745443590902244e-05,
+      "loss": 0.5385,
+      "step": 6858
+    },
+    {
+      "epoch": 2.0579188236176758,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.743937339960838e-05,
+      "loss": 0.5743,
+      "step": 6859
+    },
+    {
+      "epoch": 2.0582189211493738,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.742431089019432e-05,
+      "loss": 0.5637,
+      "step": 6860
+    },
+    {
+      "epoch": 2.0585190186810713,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.740924838078023e-05,
+      "loss": 0.6401,
+      "step": 6861
+    },
+    {
+      "epoch": 2.0588191162127694,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.739418587136617e-05,
+      "loss": 0.6188,
+      "step": 6862
+    },
+    {
+      "epoch": 2.059119213744467,
+      "grad_norm": 0.875,
+      "learning_rate": 9.737912336195211e-05,
+      "loss": 0.4801,
+      "step": 6863
+    },
+    {
+      "epoch": 2.059419311276165,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.736406085253804e-05,
+      "loss": 0.6317,
+      "step": 6864
+    },
+    {
+      "epoch": 2.0597194088078625,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.734899834312397e-05,
+      "loss": 0.5696,
+      "step": 6865
+    },
+    {
+      "epoch": 2.0600195063395605,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.73339358337099e-05,
+      "loss": 0.5673,
+      "step": 6866
+    },
+    {
+      "epoch": 2.060319603871258,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.731887332429583e-05,
+      "loss": 0.4641,
+      "step": 6867
+    },
+    {
+      "epoch": 2.060619701402956,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.730381081488177e-05,
+      "loss": 0.6512,
+      "step": 6868
+    },
+    {
+      "epoch": 2.0609197989346537,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.728874830546769e-05,
+      "loss": 0.5633,
+      "step": 6869
+    },
+    {
+      "epoch": 2.0612198964663517,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.727368579605363e-05,
+      "loss": 0.5694,
+      "step": 6870
+    },
+    {
+      "epoch": 2.0615199939980493,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.725862328663955e-05,
+      "loss": 0.5409,
+      "step": 6871
+    },
+    {
+      "epoch": 2.0618200915297473,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.72435607772255e-05,
+      "loss": 0.5137,
+      "step": 6872
+    },
+    {
+      "epoch": 2.062120189061445,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.722849826781142e-05,
+      "loss": 0.4897,
+      "step": 6873
+    },
+    {
+      "epoch": 2.062420286593143,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.721343575839735e-05,
+      "loss": 0.625,
+      "step": 6874
+    },
+    {
+      "epoch": 2.0627203841248405,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.719837324898329e-05,
+      "loss": 0.4602,
+      "step": 6875
+    },
+    {
+      "epoch": 2.0630204816565385,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.718331073956922e-05,
+      "loss": 0.5586,
+      "step": 6876
+    },
+    {
+      "epoch": 2.063320579188236,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.716824823015514e-05,
+      "loss": 0.5419,
+      "step": 6877
+    },
+    {
+      "epoch": 2.063620676719934,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.715318572074108e-05,
+      "loss": 0.5294,
+      "step": 6878
+    },
+    {
+      "epoch": 2.0639207742516317,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.713812321132701e-05,
+      "loss": 0.64,
+      "step": 6879
+    },
+    {
+      "epoch": 2.0642208717833297,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.712306070191294e-05,
+      "loss": 0.7181,
+      "step": 6880
+    },
+    {
+      "epoch": 2.0645209693150273,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.710799819249888e-05,
+      "loss": 0.5812,
+      "step": 6881
+    },
+    {
+      "epoch": 2.0648210668467253,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.70929356830848e-05,
+      "loss": 0.5857,
+      "step": 6882
+    },
+    {
+      "epoch": 2.065121164378423,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.707787317367074e-05,
+      "loss": 0.6157,
+      "step": 6883
+    },
+    {
+      "epoch": 2.065421261910121,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.706281066425667e-05,
+      "loss": 0.4609,
+      "step": 6884
+    },
+    {
+      "epoch": 2.0657213594418184,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.70477481548426e-05,
+      "loss": 0.4902,
+      "step": 6885
+    },
+    {
+      "epoch": 2.0660214569735165,
+      "grad_norm": 1.125,
+      "learning_rate": 9.703268564542854e-05,
+      "loss": 0.6479,
+      "step": 6886
+    },
+    {
+      "epoch": 2.066321554505214,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.701762313601446e-05,
+      "loss": 0.6545,
+      "step": 6887
+    },
+    {
+      "epoch": 2.066621652036912,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.700256062660039e-05,
+      "loss": 0.5093,
+      "step": 6888
+    },
+    {
+      "epoch": 2.0669217495686096,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.698749811718633e-05,
+      "loss": 0.5728,
+      "step": 6889
+    },
+    {
+      "epoch": 2.0672218471003077,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.697243560777226e-05,
+      "loss": 0.5684,
+      "step": 6890
+    },
+    {
+      "epoch": 2.0675219446320052,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.69573730983582e-05,
+      "loss": 0.549,
+      "step": 6891
+    },
+    {
+      "epoch": 2.0678220421637032,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.694231058894412e-05,
+      "loss": 0.584,
+      "step": 6892
+    },
+    {
+      "epoch": 2.068122139695401,
+      "grad_norm": 1.0,
+      "learning_rate": 9.692724807953005e-05,
+      "loss": 0.7322,
+      "step": 6893
+    },
+    {
+      "epoch": 2.068422237227099,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.691218557011599e-05,
+      "loss": 0.6386,
+      "step": 6894
+    },
+    {
+      "epoch": 2.0687223347587964,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.689712306070192e-05,
+      "loss": 0.5324,
+      "step": 6895
+    },
+    {
+      "epoch": 2.0690224322904944,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.688206055128785e-05,
+      "loss": 0.5968,
+      "step": 6896
+    },
+    {
+      "epoch": 2.069322529822192,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.686699804187379e-05,
+      "loss": 0.5789,
+      "step": 6897
+    },
+    {
+      "epoch": 2.06962262735389,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.685193553245971e-05,
+      "loss": 0.6162,
+      "step": 6898
+    },
+    {
+      "epoch": 2.069922724885588,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.683687302304565e-05,
+      "loss": 0.5023,
+      "step": 6899
+    },
+    {
+      "epoch": 2.0702228224172856,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.682181051363157e-05,
+      "loss": 0.6409,
+      "step": 6900
+    },
+    {
+      "epoch": 2.0705229199489836,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.68067480042175e-05,
+      "loss": 0.6002,
+      "step": 6901
+    },
+    {
+      "epoch": 2.070823017480681,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.679168549480345e-05,
+      "loss": 0.6309,
+      "step": 6902
+    },
+    {
+      "epoch": 2.0711231150123792,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.677662298538937e-05,
+      "loss": 0.5907,
+      "step": 6903
+    },
+    {
+      "epoch": 2.071423212544077,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.67615604759753e-05,
+      "loss": 0.566,
+      "step": 6904
+    },
+    {
+      "epoch": 2.071723310075775,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.674649796656123e-05,
+      "loss": 0.5996,
+      "step": 6905
+    },
+    {
+      "epoch": 2.0720234076074724,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.673143545714717e-05,
+      "loss": 0.5264,
+      "step": 6906
+    },
+    {
+      "epoch": 2.0723235051391704,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.67163729477331e-05,
+      "loss": 0.617,
+      "step": 6907
+    },
+    {
+      "epoch": 2.072623602670868,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.670131043831902e-05,
+      "loss": 0.5587,
+      "step": 6908
+    },
+    {
+      "epoch": 2.072923700202566,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.668624792890496e-05,
+      "loss": 0.6587,
+      "step": 6909
+    },
+    {
+      "epoch": 2.0732237977342636,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.66711854194909e-05,
+      "loss": 0.517,
+      "step": 6910
+    },
+    {
+      "epoch": 2.0735238952659616,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.665612291007681e-05,
+      "loss": 0.658,
+      "step": 6911
+    },
+    {
+      "epoch": 2.073823992797659,
+      "grad_norm": 0.875,
+      "learning_rate": 9.664106040066275e-05,
+      "loss": 0.5538,
+      "step": 6912
+    },
+    {
+      "epoch": 2.074124090329357,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.662599789124868e-05,
+      "loss": 0.5464,
+      "step": 6913
+    },
+    {
+      "epoch": 2.0744241878610548,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.661093538183462e-05,
+      "loss": 0.6264,
+      "step": 6914
+    },
+    {
+      "epoch": 2.0747242853927528,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.659587287242055e-05,
+      "loss": 0.5917,
+      "step": 6915
+    },
+    {
+      "epoch": 2.0750243829244504,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.658081036300648e-05,
+      "loss": 0.5844,
+      "step": 6916
+    },
+    {
+      "epoch": 2.0753244804561484,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.656574785359242e-05,
+      "loss": 0.567,
+      "step": 6917
+    },
+    {
+      "epoch": 2.075624577987846,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.655068534417834e-05,
+      "loss": 0.544,
+      "step": 6918
+    },
+    {
+      "epoch": 2.075924675519544,
+      "grad_norm": 0.85546875,
+      "learning_rate": 9.653562283476427e-05,
+      "loss": 0.5574,
+      "step": 6919
+    },
+    {
+      "epoch": 2.0762247730512415,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.652056032535021e-05,
+      "loss": 0.6041,
+      "step": 6920
+    },
+    {
+      "epoch": 2.0765248705829396,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.650549781593614e-05,
+      "loss": 0.5944,
+      "step": 6921
+    },
+    {
+      "epoch": 2.076824968114637,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.649043530652208e-05,
+      "loss": 0.6212,
+      "step": 6922
+    },
+    {
+      "epoch": 2.077125065646335,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.6475372797108e-05,
+      "loss": 0.6115,
+      "step": 6923
+    },
+    {
+      "epoch": 2.0774251631780327,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.646031028769393e-05,
+      "loss": 0.5558,
+      "step": 6924
+    },
+    {
+      "epoch": 2.0777252607097307,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.644524777827987e-05,
+      "loss": 0.599,
+      "step": 6925
+    },
+    {
+      "epoch": 2.0780253582414283,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.64301852688658e-05,
+      "loss": 0.5907,
+      "step": 6926
+    },
+    {
+      "epoch": 2.0783254557731263,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.641512275945172e-05,
+      "loss": 0.5682,
+      "step": 6927
+    },
+    {
+      "epoch": 2.078625553304824,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.640006025003766e-05,
+      "loss": 0.524,
+      "step": 6928
+    },
+    {
+      "epoch": 2.078925650836522,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.638499774062359e-05,
+      "loss": 0.6119,
+      "step": 6929
+    },
+    {
+      "epoch": 2.0792257483682195,
+      "grad_norm": 1.3984375,
+      "learning_rate": 9.636993523120953e-05,
+      "loss": 0.6434,
+      "step": 6930
+    },
+    {
+      "epoch": 2.0795258458999175,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.635487272179546e-05,
+      "loss": 0.557,
+      "step": 6931
+    },
+    {
+      "epoch": 2.079825943431615,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.633981021238138e-05,
+      "loss": 0.6081,
+      "step": 6932
+    },
+    {
+      "epoch": 2.080126040963313,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.632474770296733e-05,
+      "loss": 0.5777,
+      "step": 6933
+    },
+    {
+      "epoch": 2.0804261384950107,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.630968519355325e-05,
+      "loss": 0.5116,
+      "step": 6934
+    },
+    {
+      "epoch": 2.0807262360267087,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.629462268413918e-05,
+      "loss": 0.5742,
+      "step": 6935
+    },
+    {
+      "epoch": 2.0810263335584063,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.627956017472512e-05,
+      "loss": 0.6006,
+      "step": 6936
+    },
+    {
+      "epoch": 2.0813264310901043,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.626449766531105e-05,
+      "loss": 0.5591,
+      "step": 6937
+    },
+    {
+      "epoch": 2.0816265286218023,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.624943515589697e-05,
+      "loss": 0.5798,
+      "step": 6938
+    },
+    {
+      "epoch": 2.0819266261535,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.623437264648291e-05,
+      "loss": 0.6301,
+      "step": 6939
+    },
+    {
+      "epoch": 2.082226723685198,
+      "grad_norm": 1.0,
+      "learning_rate": 9.621931013706884e-05,
+      "loss": 0.614,
+      "step": 6940
+    },
+    {
+      "epoch": 2.0825268212168955,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.620424762765478e-05,
+      "loss": 0.598,
+      "step": 6941
+    },
+    {
+      "epoch": 2.0828269187485935,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.618918511824069e-05,
+      "loss": 0.6404,
+      "step": 6942
+    },
+    {
+      "epoch": 2.083127016280291,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.617412260882663e-05,
+      "loss": 0.4947,
+      "step": 6943
+    },
+    {
+      "epoch": 2.083427113811989,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.615906009941257e-05,
+      "loss": 0.5754,
+      "step": 6944
+    },
+    {
+      "epoch": 2.0837272113436867,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.61439975899985e-05,
+      "loss": 0.6292,
+      "step": 6945
+    },
+    {
+      "epoch": 2.0840273088753847,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.612893508058443e-05,
+      "loss": 0.5501,
+      "step": 6946
+    },
+    {
+      "epoch": 2.0843274064070823,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.611387257117035e-05,
+      "loss": 0.6202,
+      "step": 6947
+    },
+    {
+      "epoch": 2.0846275039387803,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.60988100617563e-05,
+      "loss": 0.6409,
+      "step": 6948
+    },
+    {
+      "epoch": 2.084927601470478,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.608374755234223e-05,
+      "loss": 0.5221,
+      "step": 6949
+    },
+    {
+      "epoch": 2.085227699002176,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.606868504292815e-05,
+      "loss": 0.6227,
+      "step": 6950
+    },
+    {
+      "epoch": 2.0855277965338734,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.605362253351409e-05,
+      "loss": 0.5898,
+      "step": 6951
+    },
+    {
+      "epoch": 2.0858278940655715,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.603856002410003e-05,
+      "loss": 0.5088,
+      "step": 6952
+    },
+    {
+      "epoch": 2.086127991597269,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.602349751468596e-05,
+      "loss": 0.6073,
+      "step": 6953
+    },
+    {
+      "epoch": 2.086428089128967,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.600843500527188e-05,
+      "loss": 0.6701,
+      "step": 6954
+    },
+    {
+      "epoch": 2.0867281866606646,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.599337249585781e-05,
+      "loss": 0.6788,
+      "step": 6955
+    },
+    {
+      "epoch": 2.0870282841923626,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.597830998644375e-05,
+      "loss": 0.5758,
+      "step": 6956
+    },
+    {
+      "epoch": 2.08732838172406,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.596324747702969e-05,
+      "loss": 0.6852,
+      "step": 6957
+    },
+    {
+      "epoch": 2.0876284792557582,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.59481849676156e-05,
+      "loss": 0.6531,
+      "step": 6958
+    },
+    {
+      "epoch": 2.087928576787456,
+      "grad_norm": 1.3046875,
+      "learning_rate": 9.593312245820154e-05,
+      "loss": 0.6563,
+      "step": 6959
+    },
+    {
+      "epoch": 2.088228674319154,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.591805994878747e-05,
+      "loss": 0.5893,
+      "step": 6960
+    },
+    {
+      "epoch": 2.0885287718508514,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.59029974393734e-05,
+      "loss": 0.4789,
+      "step": 6961
+    },
+    {
+      "epoch": 2.0888288693825494,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.588793492995934e-05,
+      "loss": 0.5238,
+      "step": 6962
+    },
+    {
+      "epoch": 2.089128966914247,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.587287242054526e-05,
+      "loss": 0.6123,
+      "step": 6963
+    },
+    {
+      "epoch": 2.089429064445945,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.58578099111312e-05,
+      "loss": 0.6544,
+      "step": 6964
+    },
+    {
+      "epoch": 2.0897291619776426,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.584274740171713e-05,
+      "loss": 0.6226,
+      "step": 6965
+    },
+    {
+      "epoch": 2.0900292595093406,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.582768489230306e-05,
+      "loss": 0.5867,
+      "step": 6966
+    },
+    {
+      "epoch": 2.090329357041038,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.5812622382889e-05,
+      "loss": 0.5641,
+      "step": 6967
+    },
+    {
+      "epoch": 2.090629454572736,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.579755987347492e-05,
+      "loss": 0.5938,
+      "step": 6968
+    },
+    {
+      "epoch": 2.0909295521044338,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.578249736406085e-05,
+      "loss": 0.6829,
+      "step": 6969
+    },
+    {
+      "epoch": 2.091229649636132,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.576743485464679e-05,
+      "loss": 0.6187,
+      "step": 6970
+    },
+    {
+      "epoch": 2.0915297471678294,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.575237234523272e-05,
+      "loss": 0.5641,
+      "step": 6971
+    },
+    {
+      "epoch": 2.0918298446995274,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.573730983581866e-05,
+      "loss": 0.6217,
+      "step": 6972
+    },
+    {
+      "epoch": 2.092129942231225,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.572224732640458e-05,
+      "loss": 0.5175,
+      "step": 6973
+    },
+    {
+      "epoch": 2.092430039762923,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.570718481699051e-05,
+      "loss": 0.5448,
+      "step": 6974
+    },
+    {
+      "epoch": 2.0927301372946205,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.569212230757645e-05,
+      "loss": 0.62,
+      "step": 6975
+    },
+    {
+      "epoch": 2.0930302348263186,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.567705979816238e-05,
+      "loss": 0.7301,
+      "step": 6976
+    },
+    {
+      "epoch": 2.0933303323580166,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.56619972887483e-05,
+      "loss": 0.5154,
+      "step": 6977
+    },
+    {
+      "epoch": 2.093630429889714,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.564693477933425e-05,
+      "loss": 0.5464,
+      "step": 6978
+    },
+    {
+      "epoch": 2.093930527421412,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.563187226992017e-05,
+      "loss": 0.5611,
+      "step": 6979
+    },
+    {
+      "epoch": 2.0942306249531097,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.561680976050611e-05,
+      "loss": 0.5936,
+      "step": 6980
+    },
+    {
+      "epoch": 2.0945307224848078,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.560174725109204e-05,
+      "loss": 0.537,
+      "step": 6981
+    },
+    {
+      "epoch": 2.0948308200165053,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.558668474167797e-05,
+      "loss": 0.4879,
+      "step": 6982
+    },
+    {
+      "epoch": 2.0951309175482034,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.55716222322639e-05,
+      "loss": 0.5737,
+      "step": 6983
+    },
+    {
+      "epoch": 2.095431015079901,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.555655972284983e-05,
+      "loss": 0.5802,
+      "step": 6984
+    },
+    {
+      "epoch": 2.095731112611599,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.554149721343576e-05,
+      "loss": 0.6661,
+      "step": 6985
+    },
+    {
+      "epoch": 2.0960312101432965,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.55264347040217e-05,
+      "loss": 0.6199,
+      "step": 6986
+    },
+    {
+      "epoch": 2.0963313076749945,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.551137219460763e-05,
+      "loss": 0.6256,
+      "step": 6987
+    },
+    {
+      "epoch": 2.096631405206692,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.549630968519355e-05,
+      "loss": 0.5336,
+      "step": 6988
+    },
+    {
+      "epoch": 2.09693150273839,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.548124717577948e-05,
+      "loss": 0.6831,
+      "step": 6989
+    },
+    {
+      "epoch": 2.0972316002700877,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.546618466636542e-05,
+      "loss": 0.587,
+      "step": 6990
+    },
+    {
+      "epoch": 2.0975316978017857,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.545112215695136e-05,
+      "loss": 0.6512,
+      "step": 6991
+    },
+    {
+      "epoch": 2.0978317953334833,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.543605964753727e-05,
+      "loss": 0.6205,
+      "step": 6992
+    },
+    {
+      "epoch": 2.0981318928651813,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.542099713812321e-05,
+      "loss": 0.4964,
+      "step": 6993
+    },
+    {
+      "epoch": 2.098431990396879,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.540593462870914e-05,
+      "loss": 0.5901,
+      "step": 6994
+    },
+    {
+      "epoch": 2.098732087928577,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.539087211929508e-05,
+      "loss": 0.5088,
+      "step": 6995
+    },
+    {
+      "epoch": 2.0990321854602745,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.537580960988101e-05,
+      "loss": 0.5227,
+      "step": 6996
+    },
+    {
+      "epoch": 2.0993322829919725,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.536074710046694e-05,
+      "loss": 0.5752,
+      "step": 6997
+    },
+    {
+      "epoch": 2.09963238052367,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.534568459105288e-05,
+      "loss": 0.5706,
+      "step": 6998
+    },
+    {
+      "epoch": 2.099932478055368,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.533062208163882e-05,
+      "loss": 0.5169,
+      "step": 6999
+    },
+    {
+      "epoch": 2.1002325755870657,
+      "grad_norm": 1.171875,
+      "learning_rate": 9.531555957222473e-05,
+      "loss": 0.5565,
+      "step": 7000
+    },
+    {
+      "epoch": 2.1005326731187637,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.530049706281067e-05,
+      "loss": 0.6293,
+      "step": 7001
+    },
+    {
+      "epoch": 2.1008327706504613,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.52854345533966e-05,
+      "loss": 0.5436,
+      "step": 7002
+    },
+    {
+      "epoch": 2.1011328681821593,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.527037204398254e-05,
+      "loss": 0.6303,
+      "step": 7003
+    },
+    {
+      "epoch": 2.101432965713857,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.525530953456846e-05,
+      "loss": 0.5352,
+      "step": 7004
+    },
+    {
+      "epoch": 2.101733063245555,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.524024702515439e-05,
+      "loss": 0.6207,
+      "step": 7005
+    },
+    {
+      "epoch": 2.1020331607772524,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.522518451574033e-05,
+      "loss": 0.5919,
+      "step": 7006
+    },
+    {
+      "epoch": 2.1023332583089505,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.521012200632626e-05,
+      "loss": 0.5844,
+      "step": 7007
+    },
+    {
+      "epoch": 2.102633355840648,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.519505949691218e-05,
+      "loss": 0.5969,
+      "step": 7008
+    },
+    {
+      "epoch": 2.102933453372346,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.517999698749812e-05,
+      "loss": 0.7185,
+      "step": 7009
+    },
+    {
+      "epoch": 2.1032335509040436,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.516493447808405e-05,
+      "loss": 0.571,
+      "step": 7010
+    },
+    {
+      "epoch": 2.1035336484357416,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.514987196866999e-05,
+      "loss": 0.6199,
+      "step": 7011
+    },
+    {
+      "epoch": 2.103833745967439,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.513480945925592e-05,
+      "loss": 0.6018,
+      "step": 7012
+    },
+    {
+      "epoch": 2.1041338434991372,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.511974694984184e-05,
+      "loss": 0.52,
+      "step": 7013
+    },
+    {
+      "epoch": 2.104433941030835,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.510468444042779e-05,
+      "loss": 0.6906,
+      "step": 7014
+    },
+    {
+      "epoch": 2.104734038562533,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.508962193101371e-05,
+      "loss": 0.5936,
+      "step": 7015
+    },
+    {
+      "epoch": 2.105034136094231,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.507455942159964e-05,
+      "loss": 0.5655,
+      "step": 7016
+    },
+    {
+      "epoch": 2.1053342336259284,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.505949691218558e-05,
+      "loss": 0.4866,
+      "step": 7017
+    },
+    {
+      "epoch": 2.1056343311576264,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.50444344027715e-05,
+      "loss": 0.7705,
+      "step": 7018
+    },
+    {
+      "epoch": 2.105934428689324,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.502937189335743e-05,
+      "loss": 0.6346,
+      "step": 7019
+    },
+    {
+      "epoch": 2.106234526221022,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.501430938394337e-05,
+      "loss": 0.5994,
+      "step": 7020
+    },
+    {
+      "epoch": 2.1065346237527196,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.49992468745293e-05,
+      "loss": 0.5469,
+      "step": 7021
+    },
+    {
+      "epoch": 2.1068347212844176,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.498418436511524e-05,
+      "loss": 0.5371,
+      "step": 7022
+    },
+    {
+      "epoch": 2.107134818816115,
+      "grad_norm": 1.125,
+      "learning_rate": 9.496912185570115e-05,
+      "loss": 0.812,
+      "step": 7023
+    },
+    {
+      "epoch": 2.107434916347813,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.49540593462871e-05,
+      "loss": 0.4811,
+      "step": 7024
+    },
+    {
+      "epoch": 2.107735013879511,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.493899683687303e-05,
+      "loss": 0.5749,
+      "step": 7025
+    },
+    {
+      "epoch": 2.108035111411209,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.492393432745896e-05,
+      "loss": 0.6363,
+      "step": 7026
+    },
+    {
+      "epoch": 2.1083352089429064,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.490887181804489e-05,
+      "loss": 0.4825,
+      "step": 7027
+    },
+    {
+      "epoch": 2.1086353064746044,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.489380930863083e-05,
+      "loss": 0.5535,
+      "step": 7028
+    },
+    {
+      "epoch": 2.108935404006302,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.487874679921675e-05,
+      "loss": 0.6057,
+      "step": 7029
+    },
+    {
+      "epoch": 2.109235501538,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.48636842898027e-05,
+      "loss": 0.567,
+      "step": 7030
+    },
+    {
+      "epoch": 2.1095355990696976,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.484862178038861e-05,
+      "loss": 0.5444,
+      "step": 7031
+    },
+    {
+      "epoch": 2.1098356966013956,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.483355927097455e-05,
+      "loss": 0.6194,
+      "step": 7032
+    },
+    {
+      "epoch": 2.110135794133093,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.481849676156049e-05,
+      "loss": 0.49,
+      "step": 7033
+    },
+    {
+      "epoch": 2.110435891664791,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.480343425214642e-05,
+      "loss": 0.586,
+      "step": 7034
+    },
+    {
+      "epoch": 2.1107359891964887,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.478837174273234e-05,
+      "loss": 0.5869,
+      "step": 7035
+    },
+    {
+      "epoch": 2.1110360867281868,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.477330923331827e-05,
+      "loss": 0.6801,
+      "step": 7036
+    },
+    {
+      "epoch": 2.1113361842598843,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.475824672390421e-05,
+      "loss": 0.5457,
+      "step": 7037
+    },
+    {
+      "epoch": 2.1116362817915824,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.474318421449015e-05,
+      "loss": 0.5911,
+      "step": 7038
+    },
+    {
+      "epoch": 2.11193637932328,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.472812170507606e-05,
+      "loss": 0.6602,
+      "step": 7039
+    },
+    {
+      "epoch": 2.112236476854978,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.4713059195662e-05,
+      "loss": 0.6252,
+      "step": 7040
+    },
+    {
+      "epoch": 2.1125365743866755,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.469799668624793e-05,
+      "loss": 0.5223,
+      "step": 7041
+    },
+    {
+      "epoch": 2.1128366719183735,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.468293417683386e-05,
+      "loss": 0.6451,
+      "step": 7042
+    },
+    {
+      "epoch": 2.113136769450071,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.46678716674198e-05,
+      "loss": 0.6045,
+      "step": 7043
+    },
+    {
+      "epoch": 2.113436866981769,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.465280915800572e-05,
+      "loss": 0.5291,
+      "step": 7044
+    },
+    {
+      "epoch": 2.1137369645134667,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.463774664859166e-05,
+      "loss": 0.6279,
+      "step": 7045
+    },
+    {
+      "epoch": 2.1140370620451647,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.462268413917759e-05,
+      "loss": 0.5464,
+      "step": 7046
+    },
+    {
+      "epoch": 2.1143371595768623,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.460762162976352e-05,
+      "loss": 0.4713,
+      "step": 7047
+    },
+    {
+      "epoch": 2.1146372571085603,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.459255912034946e-05,
+      "loss": 0.6195,
+      "step": 7048
+    },
+    {
+      "epoch": 2.114937354640258,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.457749661093538e-05,
+      "loss": 0.5428,
+      "step": 7049
+    },
+    {
+      "epoch": 2.115237452171956,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.456243410152131e-05,
+      "loss": 0.5956,
+      "step": 7050
+    },
+    {
+      "epoch": 2.1155375497036535,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.454737159210725e-05,
+      "loss": 0.5526,
+      "step": 7051
+    },
+    {
+      "epoch": 2.1158376472353515,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.453230908269318e-05,
+      "loss": 0.5515,
+      "step": 7052
+    },
+    {
+      "epoch": 2.116137744767049,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.451724657327912e-05,
+      "loss": 0.5945,
+      "step": 7053
+    },
+    {
+      "epoch": 2.116437842298747,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.450218406386505e-05,
+      "loss": 0.5878,
+      "step": 7054
+    },
+    {
+      "epoch": 2.116737939830445,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.448712155445097e-05,
+      "loss": 0.5737,
+      "step": 7055
+    },
+    {
+      "epoch": 2.1170380373621427,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.447205904503691e-05,
+      "loss": 0.5609,
+      "step": 7056
+    },
+    {
+      "epoch": 2.1173381348938407,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.445699653562284e-05,
+      "loss": 0.6389,
+      "step": 7057
+    },
+    {
+      "epoch": 2.1176382324255383,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.444193402620877e-05,
+      "loss": 0.5073,
+      "step": 7058
+    },
+    {
+      "epoch": 2.1179383299572363,
+      "grad_norm": 1.125,
+      "learning_rate": 9.44268715167947e-05,
+      "loss": 0.5772,
+      "step": 7059
+    },
+    {
+      "epoch": 2.118238427488934,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.441180900738063e-05,
+      "loss": 0.6223,
+      "step": 7060
+    },
+    {
+      "epoch": 2.118538525020632,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.439674649796657e-05,
+      "loss": 0.5513,
+      "step": 7061
+    },
+    {
+      "epoch": 2.1188386225523295,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.43816839885525e-05,
+      "loss": 0.4916,
+      "step": 7062
+    },
+    {
+      "epoch": 2.1191387200840275,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.436662147913843e-05,
+      "loss": 0.4904,
+      "step": 7063
+    },
+    {
+      "epoch": 2.119438817615725,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.435155896972437e-05,
+      "loss": 0.5214,
+      "step": 7064
+    },
+    {
+      "epoch": 2.119738915147423,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.43364964603103e-05,
+      "loss": 0.5868,
+      "step": 7065
+    },
+    {
+      "epoch": 2.1200390126791206,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.432143395089622e-05,
+      "loss": 0.5691,
+      "step": 7066
+    },
+    {
+      "epoch": 2.1203391102108187,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.430637144148216e-05,
+      "loss": 0.5986,
+      "step": 7067
+    },
+    {
+      "epoch": 2.1206392077425162,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.429130893206809e-05,
+      "loss": 0.59,
+      "step": 7068
+    },
+    {
+      "epoch": 2.1209393052742143,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.427624642265401e-05,
+      "loss": 0.6254,
+      "step": 7069
+    },
+    {
+      "epoch": 2.121239402805912,
+      "grad_norm": 0.91796875,
+      "learning_rate": 9.426118391323995e-05,
+      "loss": 0.5571,
+      "step": 7070
+    },
+    {
+      "epoch": 2.12153950033761,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.424612140382588e-05,
+      "loss": 0.6457,
+      "step": 7071
+    },
+    {
+      "epoch": 2.1218395978693074,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.423105889441182e-05,
+      "loss": 0.5564,
+      "step": 7072
+    },
+    {
+      "epoch": 2.1221396954010054,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.421599638499773e-05,
+      "loss": 0.5876,
+      "step": 7073
+    },
+    {
+      "epoch": 2.122439792932703,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.420093387558367e-05,
+      "loss": 0.6488,
+      "step": 7074
+    },
+    {
+      "epoch": 2.122739890464401,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.418587136616962e-05,
+      "loss": 0.6555,
+      "step": 7075
+    },
+    {
+      "epoch": 2.1230399879960986,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.417080885675554e-05,
+      "loss": 0.634,
+      "step": 7076
+    },
+    {
+      "epoch": 2.1233400855277966,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.415574634734147e-05,
+      "loss": 0.4896,
+      "step": 7077
+    },
+    {
+      "epoch": 2.123640183059494,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.41406838379274e-05,
+      "loss": 0.5653,
+      "step": 7078
+    },
+    {
+      "epoch": 2.123940280591192,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.412562132851334e-05,
+      "loss": 0.6147,
+      "step": 7079
+    },
+    {
+      "epoch": 2.12424037812289,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.411055881909928e-05,
+      "loss": 0.5373,
+      "step": 7080
+    },
+    {
+      "epoch": 2.124540475654588,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.409549630968519e-05,
+      "loss": 0.6169,
+      "step": 7081
+    },
+    {
+      "epoch": 2.1248405731862854,
+      "grad_norm": 1.125,
+      "learning_rate": 9.408043380027113e-05,
+      "loss": 0.5703,
+      "step": 7082
+    },
+    {
+      "epoch": 2.1251406707179834,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.406537129085706e-05,
+      "loss": 0.5762,
+      "step": 7083
+    },
+    {
+      "epoch": 2.125440768249681,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.4050308781443e-05,
+      "loss": 0.5905,
+      "step": 7084
+    },
+    {
+      "epoch": 2.125740865781379,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.403524627202892e-05,
+      "loss": 0.5685,
+      "step": 7085
+    },
+    {
+      "epoch": 2.1260409633130766,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.402018376261485e-05,
+      "loss": 0.6112,
+      "step": 7086
+    },
+    {
+      "epoch": 2.1263410608447746,
+      "grad_norm": 1.0,
+      "learning_rate": 9.400512125320079e-05,
+      "loss": 0.7094,
+      "step": 7087
+    },
+    {
+      "epoch": 2.126641158376472,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.399005874378673e-05,
+      "loss": 0.5063,
+      "step": 7088
+    },
+    {
+      "epoch": 2.12694125590817,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.397499623437264e-05,
+      "loss": 0.5584,
+      "step": 7089
+    },
+    {
+      "epoch": 2.1272413534398678,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.395993372495858e-05,
+      "loss": 0.5707,
+      "step": 7090
+    },
+    {
+      "epoch": 2.1275414509715658,
+      "grad_norm": 1.3515625,
+      "learning_rate": 9.394487121554451e-05,
+      "loss": 0.6446,
+      "step": 7091
+    },
+    {
+      "epoch": 2.1278415485032633,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.392980870613045e-05,
+      "loss": 0.5204,
+      "step": 7092
+    },
+    {
+      "epoch": 2.1281416460349614,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.391474619671638e-05,
+      "loss": 0.6388,
+      "step": 7093
+    },
+    {
+      "epoch": 2.1284417435666594,
+      "grad_norm": 1.125,
+      "learning_rate": 9.38996836873023e-05,
+      "loss": 0.6046,
+      "step": 7094
+    },
+    {
+      "epoch": 2.128741841098357,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.388462117788825e-05,
+      "loss": 0.6587,
+      "step": 7095
+    },
+    {
+      "epoch": 2.129041938630055,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.386955866847417e-05,
+      "loss": 0.6565,
+      "step": 7096
+    },
+    {
+      "epoch": 2.1293420361617525,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.38544961590601e-05,
+      "loss": 0.4833,
+      "step": 7097
+    },
+    {
+      "epoch": 2.1296421336934506,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.383943364964604e-05,
+      "loss": 0.7264,
+      "step": 7098
+    },
+    {
+      "epoch": 2.129942231225148,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.382437114023197e-05,
+      "loss": 0.6399,
+      "step": 7099
+    },
+    {
+      "epoch": 2.130242328756846,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.380930863081789e-05,
+      "loss": 0.5471,
+      "step": 7100
+    },
+    {
+      "epoch": 2.1305424262885437,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.379424612140383e-05,
+      "loss": 0.585,
+      "step": 7101
+    },
+    {
+      "epoch": 2.1308425238202418,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.377918361198976e-05,
+      "loss": 0.6062,
+      "step": 7102
+    },
+    {
+      "epoch": 2.1311426213519393,
+      "grad_norm": 0.87890625,
+      "learning_rate": 9.37641211025757e-05,
+      "loss": 0.483,
+      "step": 7103
+    },
+    {
+      "epoch": 2.1314427188836373,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.374905859316163e-05,
+      "loss": 0.5546,
+      "step": 7104
+    },
+    {
+      "epoch": 2.131742816415335,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.373399608374755e-05,
+      "loss": 0.6017,
+      "step": 7105
+    },
+    {
+      "epoch": 2.132042913947033,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.37189335743335e-05,
+      "loss": 0.5121,
+      "step": 7106
+    },
+    {
+      "epoch": 2.1323430114787305,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.370387106491942e-05,
+      "loss": 0.6007,
+      "step": 7107
+    },
+    {
+      "epoch": 2.1326431090104285,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.368880855550535e-05,
+      "loss": 0.5906,
+      "step": 7108
+    },
+    {
+      "epoch": 2.132943206542126,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.367374604609129e-05,
+      "loss": 0.7363,
+      "step": 7109
+    },
+    {
+      "epoch": 2.133243304073824,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.365868353667721e-05,
+      "loss": 0.5571,
+      "step": 7110
+    },
+    {
+      "epoch": 2.1335434016055217,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.364362102726315e-05,
+      "loss": 0.5648,
+      "step": 7111
+    },
+    {
+      "epoch": 2.1338434991372197,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.362855851784907e-05,
+      "loss": 0.6146,
+      "step": 7112
+    },
+    {
+      "epoch": 2.1341435966689173,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.361349600843501e-05,
+      "loss": 0.5514,
+      "step": 7113
+    },
+    {
+      "epoch": 2.1344436942006153,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.359843349902095e-05,
+      "loss": 0.6059,
+      "step": 7114
+    },
+    {
+      "epoch": 2.134743791732313,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.358337098960688e-05,
+      "loss": 0.548,
+      "step": 7115
+    },
+    {
+      "epoch": 2.135043889264011,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.35683084801928e-05,
+      "loss": 0.5784,
+      "step": 7116
+    },
+    {
+      "epoch": 2.1353439867957085,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.355324597077874e-05,
+      "loss": 0.6485,
+      "step": 7117
+    },
+    {
+      "epoch": 2.1356440843274065,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.353818346136467e-05,
+      "loss": 0.6091,
+      "step": 7118
+    },
+    {
+      "epoch": 2.135944181859104,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.352312095195061e-05,
+      "loss": 0.5746,
+      "step": 7119
+    },
+    {
+      "epoch": 2.136244279390802,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.350805844253652e-05,
+      "loss": 0.5873,
+      "step": 7120
+    },
+    {
+      "epoch": 2.1365443769224997,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.349299593312246e-05,
+      "loss": 0.676,
+      "step": 7121
+    },
+    {
+      "epoch": 2.1368444744541977,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.34779334237084e-05,
+      "loss": 0.5344,
+      "step": 7122
+    },
+    {
+      "epoch": 2.1371445719858952,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.346287091429432e-05,
+      "loss": 0.53,
+      "step": 7123
+    },
+    {
+      "epoch": 2.1374446695175933,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.344780840488026e-05,
+      "loss": 0.6104,
+      "step": 7124
+    },
+    {
+      "epoch": 2.137744767049291,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.343274589546618e-05,
+      "loss": 0.6054,
+      "step": 7125
+    },
+    {
+      "epoch": 2.138044864580989,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.341768338605212e-05,
+      "loss": 0.538,
+      "step": 7126
+    },
+    {
+      "epoch": 2.1383449621126864,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.340262087663805e-05,
+      "loss": 0.4907,
+      "step": 7127
+    },
+    {
+      "epoch": 2.1386450596443844,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.338755836722398e-05,
+      "loss": 0.5186,
+      "step": 7128
+    },
+    {
+      "epoch": 2.138945157176082,
+      "grad_norm": 1.234375,
+      "learning_rate": 9.337249585780992e-05,
+      "loss": 0.7085,
+      "step": 7129
+    },
+    {
+      "epoch": 2.13924525470778,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.335743334839584e-05,
+      "loss": 0.6032,
+      "step": 7130
+    },
+    {
+      "epoch": 2.1395453522394776,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.334237083898177e-05,
+      "loss": 0.5879,
+      "step": 7131
+    },
+    {
+      "epoch": 2.1398454497711756,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.332730832956771e-05,
+      "loss": 0.6265,
+      "step": 7132
+    },
+    {
+      "epoch": 2.1401455473028737,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.331224582015364e-05,
+      "loss": 0.6481,
+      "step": 7133
+    },
+    {
+      "epoch": 2.1404456448345712,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.329718331073958e-05,
+      "loss": 0.6301,
+      "step": 7134
+    },
+    {
+      "epoch": 2.1407457423662692,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.32821208013255e-05,
+      "loss": 0.6214,
+      "step": 7135
+    },
+    {
+      "epoch": 2.141045839897967,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.326705829191143e-05,
+      "loss": 0.551,
+      "step": 7136
+    },
+    {
+      "epoch": 2.141345937429665,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.325199578249737e-05,
+      "loss": 0.5992,
+      "step": 7137
+    },
+    {
+      "epoch": 2.1416460349613624,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.32369332730833e-05,
+      "loss": 0.6239,
+      "step": 7138
+    },
+    {
+      "epoch": 2.1419461324930604,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.322187076366923e-05,
+      "loss": 0.6237,
+      "step": 7139
+    },
+    {
+      "epoch": 2.142246230024758,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.320680825425517e-05,
+      "loss": 0.4927,
+      "step": 7140
+    },
+    {
+      "epoch": 2.142546327556456,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.319174574484109e-05,
+      "loss": 0.6796,
+      "step": 7141
+    },
+    {
+      "epoch": 2.1428464250881536,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.317668323542703e-05,
+      "loss": 0.5746,
+      "step": 7142
+    },
+    {
+      "epoch": 2.1431465226198516,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.316162072601296e-05,
+      "loss": 0.6824,
+      "step": 7143
+    },
+    {
+      "epoch": 2.143446620151549,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.314655821659889e-05,
+      "loss": 0.5259,
+      "step": 7144
+    },
+    {
+      "epoch": 2.143746717683247,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.313149570718483e-05,
+      "loss": 0.5349,
+      "step": 7145
+    },
+    {
+      "epoch": 2.144046815214945,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.311643319777075e-05,
+      "loss": 0.6011,
+      "step": 7146
+    },
+    {
+      "epoch": 2.144346912746643,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.310137068835668e-05,
+      "loss": 0.5999,
+      "step": 7147
+    },
+    {
+      "epoch": 2.1446470102783404,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.308630817894262e-05,
+      "loss": 0.5744,
+      "step": 7148
+    },
+    {
+      "epoch": 2.1449471078100384,
+      "grad_norm": 1.3125,
+      "learning_rate": 9.307124566952855e-05,
+      "loss": 0.589,
+      "step": 7149
+    },
+    {
+      "epoch": 2.145247205341736,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.305618316011447e-05,
+      "loss": 0.6162,
+      "step": 7150
+    },
+    {
+      "epoch": 2.145547302873434,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.304112065070041e-05,
+      "loss": 0.6247,
+      "step": 7151
+    },
+    {
+      "epoch": 2.1458474004051316,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.302605814128634e-05,
+      "loss": 0.6486,
+      "step": 7152
+    },
+    {
+      "epoch": 2.1461474979368296,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.301099563187228e-05,
+      "loss": 0.5906,
+      "step": 7153
+    },
+    {
+      "epoch": 2.146447595468527,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.29959331224582e-05,
+      "loss": 0.7043,
+      "step": 7154
+    },
+    {
+      "epoch": 2.146747693000225,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.298087061304413e-05,
+      "loss": 0.6433,
+      "step": 7155
+    },
+    {
+      "epoch": 2.1470477905319227,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.296580810363008e-05,
+      "loss": 0.5489,
+      "step": 7156
+    },
+    {
+      "epoch": 2.1473478880636208,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.2950745594216e-05,
+      "loss": 0.6255,
+      "step": 7157
+    },
+    {
+      "epoch": 2.1476479855953183,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.293568308480193e-05,
+      "loss": 0.5612,
+      "step": 7158
+    },
+    {
+      "epoch": 2.1479480831270163,
+      "grad_norm": 1.125,
+      "learning_rate": 9.292062057538786e-05,
+      "loss": 0.6813,
+      "step": 7159
+    },
+    {
+      "epoch": 2.148248180658714,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.29055580659738e-05,
+      "loss": 0.5677,
+      "step": 7160
+    },
+    {
+      "epoch": 2.148548278190412,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.289049555655974e-05,
+      "loss": 0.6761,
+      "step": 7161
+    },
+    {
+      "epoch": 2.1488483757221095,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.287543304714565e-05,
+      "loss": 0.6184,
+      "step": 7162
+    },
+    {
+      "epoch": 2.1491484732538075,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.286037053773159e-05,
+      "loss": 0.6242,
+      "step": 7163
+    },
+    {
+      "epoch": 2.149448570785505,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.284530802831753e-05,
+      "loss": 0.6788,
+      "step": 7164
+    },
+    {
+      "epoch": 2.149748668317203,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.283024551890346e-05,
+      "loss": 0.5876,
+      "step": 7165
+    },
+    {
+      "epoch": 2.1500487658489007,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.281518300948938e-05,
+      "loss": 0.5983,
+      "step": 7166
+    },
+    {
+      "epoch": 2.1503488633805987,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.280012050007531e-05,
+      "loss": 0.5917,
+      "step": 7167
+    },
+    {
+      "epoch": 2.1506489609122963,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.278505799066125e-05,
+      "loss": 0.5505,
+      "step": 7168
+    },
+    {
+      "epoch": 2.1509490584439943,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.276999548124719e-05,
+      "loss": 0.5187,
+      "step": 7169
+    },
+    {
+      "epoch": 2.151249155975692,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.27549329718331e-05,
+      "loss": 0.6129,
+      "step": 7170
+    },
+    {
+      "epoch": 2.15154925350739,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.273987046241904e-05,
+      "loss": 0.6025,
+      "step": 7171
+    },
+    {
+      "epoch": 2.151849351039088,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.272480795300497e-05,
+      "loss": 0.5371,
+      "step": 7172
+    },
+    {
+      "epoch": 2.1521494485707855,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.270974544359091e-05,
+      "loss": 0.5921,
+      "step": 7173
+    },
+    {
+      "epoch": 2.1524495461024835,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.269468293417684e-05,
+      "loss": 0.5681,
+      "step": 7174
+    },
+    {
+      "epoch": 2.152749643634181,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.267962042476276e-05,
+      "loss": 0.6342,
+      "step": 7175
+    },
+    {
+      "epoch": 2.153049741165879,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.26645579153487e-05,
+      "loss": 0.5482,
+      "step": 7176
+    },
+    {
+      "epoch": 2.1533498386975767,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.264949540593463e-05,
+      "loss": 0.6008,
+      "step": 7177
+    },
+    {
+      "epoch": 2.1536499362292747,
+      "grad_norm": 1.21875,
+      "learning_rate": 9.263443289652056e-05,
+      "loss": 0.5369,
+      "step": 7178
+    },
+    {
+      "epoch": 2.1539500337609723,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.26193703871065e-05,
+      "loss": 0.65,
+      "step": 7179
+    },
+    {
+      "epoch": 2.1542501312926703,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.260430787769243e-05,
+      "loss": 0.6807,
+      "step": 7180
+    },
+    {
+      "epoch": 2.154550228824368,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.258924536827835e-05,
+      "loss": 0.5573,
+      "step": 7181
+    },
+    {
+      "epoch": 2.154850326356066,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.257418285886429e-05,
+      "loss": 0.6462,
+      "step": 7182
+    },
+    {
+      "epoch": 2.1551504238877635,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.255912034945022e-05,
+      "loss": 0.6058,
+      "step": 7183
+    },
+    {
+      "epoch": 2.1554505214194615,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.254405784003616e-05,
+      "loss": 0.5001,
+      "step": 7184
+    },
+    {
+      "epoch": 2.155750618951159,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.252899533062209e-05,
+      "loss": 0.5062,
+      "step": 7185
+    },
+    {
+      "epoch": 2.156050716482857,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.251393282120801e-05,
+      "loss": 0.6677,
+      "step": 7186
+    },
+    {
+      "epoch": 2.1563508140145546,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.249887031179395e-05,
+      "loss": 0.6471,
+      "step": 7187
+    },
+    {
+      "epoch": 2.1566509115462527,
+      "grad_norm": 1.0,
+      "learning_rate": 9.248380780237988e-05,
+      "loss": 0.5512,
+      "step": 7188
+    },
+    {
+      "epoch": 2.1569510090779502,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.246874529296581e-05,
+      "loss": 0.5599,
+      "step": 7189
+    },
+    {
+      "epoch": 2.1572511066096483,
+      "grad_norm": 0.8984375,
+      "learning_rate": 9.245368278355175e-05,
+      "loss": 0.5232,
+      "step": 7190
+    },
+    {
+      "epoch": 2.157551204141346,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.243862027413767e-05,
+      "loss": 0.4739,
+      "step": 7191
+    },
+    {
+      "epoch": 2.157851301673044,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.242355776472361e-05,
+      "loss": 0.5248,
+      "step": 7192
+    },
+    {
+      "epoch": 2.1581513992047414,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.240849525530954e-05,
+      "loss": 0.505,
+      "step": 7193
+    },
+    {
+      "epoch": 2.1584514967364394,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.239343274589547e-05,
+      "loss": 0.5375,
+      "step": 7194
+    },
+    {
+      "epoch": 2.158751594268137,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.237837023648141e-05,
+      "loss": 0.6328,
+      "step": 7195
+    },
+    {
+      "epoch": 2.159051691799835,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.236330772706734e-05,
+      "loss": 0.641,
+      "step": 7196
+    },
+    {
+      "epoch": 2.1593517893315326,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.234824521765326e-05,
+      "loss": 0.5961,
+      "step": 7197
+    },
+    {
+      "epoch": 2.1596518868632306,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.23331827082392e-05,
+      "loss": 0.5774,
+      "step": 7198
+    },
+    {
+      "epoch": 2.159951984394928,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.231812019882513e-05,
+      "loss": 0.5813,
+      "step": 7199
+    },
+    {
+      "epoch": 2.160252081926626,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.230305768941107e-05,
+      "loss": 0.6342,
+      "step": 7200
+    },
+    {
+      "epoch": 2.160552179458324,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.228799517999698e-05,
+      "loss": 0.5837,
+      "step": 7201
+    },
+    {
+      "epoch": 2.160852276990022,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.227293267058292e-05,
+      "loss": 0.5712,
+      "step": 7202
+    },
+    {
+      "epoch": 2.1611523745217194,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.225787016116886e-05,
+      "loss": 0.5336,
+      "step": 7203
+    },
+    {
+      "epoch": 2.1614524720534174,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.224280765175478e-05,
+      "loss": 0.6269,
+      "step": 7204
+    },
+    {
+      "epoch": 2.161752569585115,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.222774514234072e-05,
+      "loss": 0.6367,
+      "step": 7205
+    },
+    {
+      "epoch": 2.162052667116813,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.221268263292666e-05,
+      "loss": 0.6396,
+      "step": 7206
+    },
+    {
+      "epoch": 2.1623527646485106,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.219762012351258e-05,
+      "loss": 0.5783,
+      "step": 7207
+    },
+    {
+      "epoch": 2.1626528621802086,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.218255761409851e-05,
+      "loss": 0.5845,
+      "step": 7208
+    },
+    {
+      "epoch": 2.162952959711906,
+      "grad_norm": 1.265625,
+      "learning_rate": 9.216749510468444e-05,
+      "loss": 0.6632,
+      "step": 7209
+    },
+    {
+      "epoch": 2.163253057243604,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.215243259527038e-05,
+      "loss": 0.5395,
+      "step": 7210
+    },
+    {
+      "epoch": 2.163553154775302,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.213737008585632e-05,
+      "loss": 0.5519,
+      "step": 7211
+    },
+    {
+      "epoch": 2.1638532523069998,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.212230757644223e-05,
+      "loss": 0.5112,
+      "step": 7212
+    },
+    {
+      "epoch": 2.164153349838698,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.210724506702817e-05,
+      "loss": 0.5382,
+      "step": 7213
+    },
+    {
+      "epoch": 2.1644534473703954,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.20921825576141e-05,
+      "loss": 0.543,
+      "step": 7214
+    },
+    {
+      "epoch": 2.1647535449020934,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.207712004820004e-05,
+      "loss": 0.5766,
+      "step": 7215
+    },
+    {
+      "epoch": 2.165053642433791,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.206205753878597e-05,
+      "loss": 0.6404,
+      "step": 7216
+    },
+    {
+      "epoch": 2.165353739965489,
+      "grad_norm": 0.921875,
+      "learning_rate": 9.204699502937189e-05,
+      "loss": 0.5036,
+      "step": 7217
+    },
+    {
+      "epoch": 2.1656538374971865,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.203193251995783e-05,
+      "loss": 0.6456,
+      "step": 7218
+    },
+    {
+      "epoch": 2.1659539350288846,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.201687001054376e-05,
+      "loss": 0.6247,
+      "step": 7219
+    },
+    {
+      "epoch": 2.166254032560582,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.200180750112969e-05,
+      "loss": 0.5464,
+      "step": 7220
+    },
+    {
+      "epoch": 2.16655413009228,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.198674499171563e-05,
+      "loss": 0.4721,
+      "step": 7221
+    },
+    {
+      "epoch": 2.1668542276239777,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.197168248230155e-05,
+      "loss": 0.6397,
+      "step": 7222
+    },
+    {
+      "epoch": 2.1671543251556757,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.195661997288749e-05,
+      "loss": 0.5103,
+      "step": 7223
+    },
+    {
+      "epoch": 2.1674544226873733,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.194155746347342e-05,
+      "loss": 0.5572,
+      "step": 7224
+    },
+    {
+      "epoch": 2.1677545202190713,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.192649495405935e-05,
+      "loss": 0.588,
+      "step": 7225
+    },
+    {
+      "epoch": 2.168054617750769,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.191143244464529e-05,
+      "loss": 0.7097,
+      "step": 7226
+    },
+    {
+      "epoch": 2.168354715282467,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.189636993523121e-05,
+      "loss": 0.6118,
+      "step": 7227
+    },
+    {
+      "epoch": 2.1686548128141645,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.188130742581714e-05,
+      "loss": 0.6231,
+      "step": 7228
+    },
+    {
+      "epoch": 2.1689549103458625,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.186624491640308e-05,
+      "loss": 0.6117,
+      "step": 7229
+    },
+    {
+      "epoch": 2.16925500787756,
+      "grad_norm": 1.1875,
+      "learning_rate": 9.185118240698901e-05,
+      "loss": 0.6122,
+      "step": 7230
+    },
+    {
+      "epoch": 2.169555105409258,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.183611989757493e-05,
+      "loss": 0.553,
+      "step": 7231
+    },
+    {
+      "epoch": 2.1698552029409557,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.182105738816087e-05,
+      "loss": 0.6124,
+      "step": 7232
+    },
+    {
+      "epoch": 2.1701553004726537,
+      "grad_norm": 0.984375,
+      "learning_rate": 9.18059948787468e-05,
+      "loss": 0.568,
+      "step": 7233
+    },
+    {
+      "epoch": 2.1704553980043513,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.179093236933274e-05,
+      "loss": 0.6222,
+      "step": 7234
+    },
+    {
+      "epoch": 2.1707554955360493,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.177586985991867e-05,
+      "loss": 0.6498,
+      "step": 7235
+    },
+    {
+      "epoch": 2.171055593067747,
+      "grad_norm": 1.125,
+      "learning_rate": 9.17608073505046e-05,
+      "loss": 0.5837,
+      "step": 7236
+    },
+    {
+      "epoch": 2.171355690599445,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.174574484109054e-05,
+      "loss": 0.5916,
+      "step": 7237
+    },
+    {
+      "epoch": 2.1716557881311425,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.173068233167646e-05,
+      "loss": 0.5973,
+      "step": 7238
+    },
+    {
+      "epoch": 2.1719558856628405,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.171561982226239e-05,
+      "loss": 0.5985,
+      "step": 7239
+    },
+    {
+      "epoch": 2.172255983194538,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.170055731284833e-05,
+      "loss": 0.6271,
+      "step": 7240
+    },
+    {
+      "epoch": 2.172556080726236,
+      "grad_norm": 1.0,
+      "learning_rate": 9.168549480343426e-05,
+      "loss": 0.616,
+      "step": 7241
+    },
+    {
+      "epoch": 2.1728561782579336,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.16704322940202e-05,
+      "loss": 0.5887,
+      "step": 7242
+    },
+    {
+      "epoch": 2.1731562757896317,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.165536978460611e-05,
+      "loss": 0.4982,
+      "step": 7243
+    },
+    {
+      "epoch": 2.1734563733213292,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.164030727519205e-05,
+      "loss": 0.5969,
+      "step": 7244
+    },
+    {
+      "epoch": 2.1737564708530273,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.162524476577799e-05,
+      "loss": 0.5781,
+      "step": 7245
+    },
+    {
+      "epoch": 2.174056568384725,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.161018225636392e-05,
+      "loss": 0.6472,
+      "step": 7246
+    },
+    {
+      "epoch": 2.174356665916423,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.159511974694984e-05,
+      "loss": 0.5961,
+      "step": 7247
+    },
+    {
+      "epoch": 2.1746567634481204,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.158005723753577e-05,
+      "loss": 0.5873,
+      "step": 7248
+    },
+    {
+      "epoch": 2.1749568609798184,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.156499472812171e-05,
+      "loss": 0.5835,
+      "step": 7249
+    },
+    {
+      "epoch": 2.1752569585115165,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.154993221870765e-05,
+      "loss": 0.6318,
+      "step": 7250
+    },
+    {
+      "epoch": 2.175557056043214,
+      "grad_norm": 1.5234375,
+      "learning_rate": 9.153486970929356e-05,
+      "loss": 0.5965,
+      "step": 7251
+    },
+    {
+      "epoch": 2.175857153574912,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.15198071998795e-05,
+      "loss": 0.5886,
+      "step": 7252
+    },
+    {
+      "epoch": 2.1761572511066096,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.150474469046544e-05,
+      "loss": 0.5248,
+      "step": 7253
+    },
+    {
+      "epoch": 2.1764573486383076,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.148968218105137e-05,
+      "loss": 0.6649,
+      "step": 7254
+    },
+    {
+      "epoch": 2.176757446170005,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.14746196716373e-05,
+      "loss": 0.5863,
+      "step": 7255
+    },
+    {
+      "epoch": 2.1770575437017032,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.145955716222322e-05,
+      "loss": 0.5116,
+      "step": 7256
+    },
+    {
+      "epoch": 2.177357641233401,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.144449465280917e-05,
+      "loss": 0.5172,
+      "step": 7257
+    },
+    {
+      "epoch": 2.177657738765099,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.14294321433951e-05,
+      "loss": 0.6163,
+      "step": 7258
+    },
+    {
+      "epoch": 2.1779578362967964,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.141436963398102e-05,
+      "loss": 0.7092,
+      "step": 7259
+    },
+    {
+      "epoch": 2.1782579338284944,
+      "grad_norm": 1.0,
+      "learning_rate": 9.139930712456696e-05,
+      "loss": 0.5661,
+      "step": 7260
+    },
+    {
+      "epoch": 2.178558031360192,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.138424461515289e-05,
+      "loss": 0.6028,
+      "step": 7261
+    },
+    {
+      "epoch": 2.17885812889189,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.136918210573881e-05,
+      "loss": 0.5354,
+      "step": 7262
+    },
+    {
+      "epoch": 2.1791582264235876,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.135411959632475e-05,
+      "loss": 0.5113,
+      "step": 7263
+    },
+    {
+      "epoch": 2.1794583239552856,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.133905708691068e-05,
+      "loss": 0.6097,
+      "step": 7264
+    },
+    {
+      "epoch": 2.179758421486983,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.132399457749662e-05,
+      "loss": 0.522,
+      "step": 7265
+    },
+    {
+      "epoch": 2.180058519018681,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.130893206808255e-05,
+      "loss": 0.5901,
+      "step": 7266
+    },
+    {
+      "epoch": 2.1803586165503788,
+      "grad_norm": 1.0,
+      "learning_rate": 9.129386955866847e-05,
+      "loss": 0.4346,
+      "step": 7267
+    },
+    {
+      "epoch": 2.180658714082077,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.127880704925441e-05,
+      "loss": 0.6619,
+      "step": 7268
+    },
+    {
+      "epoch": 2.1809588116137744,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.126374453984034e-05,
+      "loss": 0.4989,
+      "step": 7269
+    },
+    {
+      "epoch": 2.1812589091454724,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.124868203042627e-05,
+      "loss": 0.5578,
+      "step": 7270
+    },
+    {
+      "epoch": 2.18155900667717,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.123361952101221e-05,
+      "loss": 0.6773,
+      "step": 7271
+    },
+    {
+      "epoch": 2.181859104208868,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.121855701159813e-05,
+      "loss": 0.6061,
+      "step": 7272
+    },
+    {
+      "epoch": 2.1821592017405655,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.120349450218407e-05,
+      "loss": 0.529,
+      "step": 7273
+    },
+    {
+      "epoch": 2.1824592992722636,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.118843199277e-05,
+      "loss": 0.6002,
+      "step": 7274
+    },
+    {
+      "epoch": 2.182759396803961,
+      "grad_norm": 0.875,
+      "learning_rate": 9.117336948335593e-05,
+      "loss": 0.5248,
+      "step": 7275
+    },
+    {
+      "epoch": 2.183059494335659,
+      "grad_norm": 1.046875,
+      "learning_rate": 9.115830697394187e-05,
+      "loss": 0.4934,
+      "step": 7276
+    },
+    {
+      "epoch": 2.1833595918673567,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.11432444645278e-05,
+      "loss": 0.6192,
+      "step": 7277
+    },
+    {
+      "epoch": 2.1836596893990547,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.112818195511372e-05,
+      "loss": 0.6036,
+      "step": 7278
+    },
+    {
+      "epoch": 2.1839597869307523,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.111311944569966e-05,
+      "loss": 0.6542,
+      "step": 7279
+    },
+    {
+      "epoch": 2.1842598844624503,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.109805693628559e-05,
+      "loss": 0.6758,
+      "step": 7280
+    },
+    {
+      "epoch": 2.184559981994148,
+      "grad_norm": 1.125,
+      "learning_rate": 9.108299442687153e-05,
+      "loss": 0.6244,
+      "step": 7281
+    },
+    {
+      "epoch": 2.184860079525846,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.106793191745746e-05,
+      "loss": 0.6069,
+      "step": 7282
+    },
+    {
+      "epoch": 2.1851601770575435,
+      "grad_norm": 1.0234375,
+      "learning_rate": 9.105286940804338e-05,
+      "loss": 0.5855,
+      "step": 7283
+    },
+    {
+      "epoch": 2.1854602745892415,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.103780689862932e-05,
+      "loss": 0.6143,
+      "step": 7284
+    },
+    {
+      "epoch": 2.185760372120939,
+      "grad_norm": 1.203125,
+      "learning_rate": 9.102274438921525e-05,
+      "loss": 0.5701,
+      "step": 7285
+    },
+    {
+      "epoch": 2.186060469652637,
+      "grad_norm": 0.9296875,
+      "learning_rate": 9.100768187980118e-05,
+      "loss": 0.5842,
+      "step": 7286
+    },
+    {
+      "epoch": 2.1863605671843347,
+      "grad_norm": 1.125,
+      "learning_rate": 9.099261937038712e-05,
+      "loss": 0.6027,
+      "step": 7287
+    },
+    {
+      "epoch": 2.1866606647160327,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.097755686097304e-05,
+      "loss": 0.6489,
+      "step": 7288
+    },
+    {
+      "epoch": 2.1869607622477307,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.096249435155897e-05,
+      "loss": 0.5123,
+      "step": 7289
+    },
+    {
+      "epoch": 2.1872608597794283,
+      "grad_norm": 1.1796875,
+      "learning_rate": 9.09474318421449e-05,
+      "loss": 0.6059,
+      "step": 7290
+    },
+    {
+      "epoch": 2.1875609573111263,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.093236933273084e-05,
+      "loss": 0.5675,
+      "step": 7291
+    },
+    {
+      "epoch": 2.187861054842824,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.091730682331678e-05,
+      "loss": 0.5043,
+      "step": 7292
+    },
+    {
+      "epoch": 2.188161152374522,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.090224431390269e-05,
+      "loss": 0.6508,
+      "step": 7293
+    },
+    {
+      "epoch": 2.1884612499062195,
+      "grad_norm": 0.99609375,
+      "learning_rate": 9.088718180448863e-05,
+      "loss": 0.5711,
+      "step": 7294
+    },
+    {
+      "epoch": 2.1887613474379175,
+      "grad_norm": 0.98828125,
+      "learning_rate": 9.087211929507457e-05,
+      "loss": 0.5601,
+      "step": 7295
+    },
+    {
+      "epoch": 2.189061444969615,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.08570567856605e-05,
+      "loss": 0.4585,
+      "step": 7296
+    },
+    {
+      "epoch": 2.189361542501313,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.084199427624643e-05,
+      "loss": 0.5825,
+      "step": 7297
+    },
+    {
+      "epoch": 2.1896616400330107,
+      "grad_norm": 0.94140625,
+      "learning_rate": 9.082693176683235e-05,
+      "loss": 0.5202,
+      "step": 7298
+    },
+    {
+      "epoch": 2.1899617375647087,
+      "grad_norm": 1.0390625,
+      "learning_rate": 9.081186925741829e-05,
+      "loss": 0.5367,
+      "step": 7299
+    },
+    {
+      "epoch": 2.1902618350964063,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.079680674800423e-05,
+      "loss": 0.6231,
+      "step": 7300
+    },
+    {
+      "epoch": 2.1905619326281043,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.078174423859015e-05,
+      "loss": 0.5921,
+      "step": 7301
+    },
+    {
+      "epoch": 2.190862030159802,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.076668172917609e-05,
+      "loss": 0.5472,
+      "step": 7302
+    },
+    {
+      "epoch": 2.1911621276915,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.075161921976201e-05,
+      "loss": 0.6147,
+      "step": 7303
+    },
+    {
+      "epoch": 2.1914622252231974,
+      "grad_norm": 1.125,
+      "learning_rate": 9.073655671034795e-05,
+      "loss": 0.5654,
+      "step": 7304
+    },
+    {
+      "epoch": 2.1917623227548955,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.072149420093388e-05,
+      "loss": 0.5922,
+      "step": 7305
+    },
+    {
+      "epoch": 2.192062420286593,
+      "grad_norm": 0.890625,
+      "learning_rate": 9.07064316915198e-05,
+      "loss": 0.5745,
+      "step": 7306
+    },
+    {
+      "epoch": 2.192362517818291,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.069136918210575e-05,
+      "loss": 0.6663,
+      "step": 7307
+    },
+    {
+      "epoch": 2.1926626153499886,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.067630667269167e-05,
+      "loss": 0.5954,
+      "step": 7308
+    },
+    {
+      "epoch": 2.1929627128816866,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.06612441632776e-05,
+      "loss": 0.5624,
+      "step": 7309
+    },
+    {
+      "epoch": 2.1932628104133842,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.064618165386354e-05,
+      "loss": 0.4617,
+      "step": 7310
+    },
+    {
+      "epoch": 2.1935629079450822,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.063111914444947e-05,
+      "loss": 0.6391,
+      "step": 7311
+    },
+    {
+      "epoch": 2.19386300547678,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.061605663503541e-05,
+      "loss": 0.6867,
+      "step": 7312
+    },
+    {
+      "epoch": 2.194163103008478,
+      "grad_norm": 1.078125,
+      "learning_rate": 9.060099412562133e-05,
+      "loss": 0.5241,
+      "step": 7313
+    },
+    {
+      "epoch": 2.1944632005401754,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.058593161620726e-05,
+      "loss": 0.5096,
+      "step": 7314
+    },
+    {
+      "epoch": 2.1947632980718734,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.05708691067932e-05,
+      "loss": 0.5923,
+      "step": 7315
+    },
+    {
+      "epoch": 2.195063395603571,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.055580659737913e-05,
+      "loss": 0.5568,
+      "step": 7316
+    },
+    {
+      "epoch": 2.195363493135269,
+      "grad_norm": 0.9765625,
+      "learning_rate": 9.054074408796505e-05,
+      "loss": 0.6002,
+      "step": 7317
+    },
+    {
+      "epoch": 2.1956635906669666,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.0525681578551e-05,
+      "loss": 0.4751,
+      "step": 7318
+    },
+    {
+      "epoch": 2.1959636881986646,
+      "grad_norm": 1.09375,
+      "learning_rate": 9.051061906913692e-05,
+      "loss": 0.5767,
+      "step": 7319
+    },
+    {
+      "epoch": 2.196263785730362,
+      "grad_norm": 1.1015625,
+      "learning_rate": 9.049555655972285e-05,
+      "loss": 0.5952,
+      "step": 7320
+    },
+    {
+      "epoch": 2.19656388326206,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.048049405030879e-05,
+      "loss": 0.5774,
+      "step": 7321
+    },
+    {
+      "epoch": 2.1968639807937578,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.046543154089472e-05,
+      "loss": 0.5498,
+      "step": 7322
+    },
+    {
+      "epoch": 2.197164078325456,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.045036903148066e-05,
+      "loss": 0.5783,
+      "step": 7323
+    },
+    {
+      "epoch": 2.1974641758571534,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.043530652206658e-05,
+      "loss": 0.642,
+      "step": 7324
+    },
+    {
+      "epoch": 2.1977642733888514,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.042024401265251e-05,
+      "loss": 0.5121,
+      "step": 7325
+    },
+    {
+      "epoch": 2.198064370920549,
+      "grad_norm": 0.97265625,
+      "learning_rate": 9.040518150323845e-05,
+      "loss": 0.5121,
+      "step": 7326
+    },
+    {
+      "epoch": 2.198364468452247,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.039011899382438e-05,
+      "loss": 0.6423,
+      "step": 7327
+    },
+    {
+      "epoch": 2.198664565983945,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.03750564844103e-05,
+      "loss": 0.5713,
+      "step": 7328
+    },
+    {
+      "epoch": 2.1989646635156426,
+      "grad_norm": 0.96875,
+      "learning_rate": 9.035999397499624e-05,
+      "loss": 0.6167,
+      "step": 7329
+    },
+    {
+      "epoch": 2.1992647610473406,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.034493146558217e-05,
+      "loss": 0.4805,
+      "step": 7330
+    },
+    {
+      "epoch": 2.199564858579038,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.032986895616811e-05,
+      "loss": 0.476,
+      "step": 7331
+    },
+    {
+      "epoch": 2.199864956110736,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.031480644675402e-05,
+      "loss": 0.5269,
+      "step": 7332
+    },
+    {
+      "epoch": 2.2001650536424338,
+      "grad_norm": 0.93359375,
+      "learning_rate": 9.029974393733996e-05,
+      "loss": 0.583,
+      "step": 7333
+    },
+    {
+      "epoch": 2.2004651511741318,
+      "grad_norm": 1.1953125,
+      "learning_rate": 9.02846814279259e-05,
+      "loss": 0.5987,
+      "step": 7334
+    },
+    {
+      "epoch": 2.2007652487058293,
+      "grad_norm": 1.2109375,
+      "learning_rate": 9.026961891851183e-05,
+      "loss": 0.6884,
+      "step": 7335
+    },
+    {
+      "epoch": 2.2010653462375274,
+      "grad_norm": 0.9375,
+      "learning_rate": 9.025455640909776e-05,
+      "loss": 0.5234,
+      "step": 7336
+    },
+    {
+      "epoch": 2.201365443769225,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.023949389968368e-05,
+      "loss": 0.635,
+      "step": 7337
+    },
+    {
+      "epoch": 2.201665541300923,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.022443139026963e-05,
+      "loss": 0.5754,
+      "step": 7338
+    },
+    {
+      "epoch": 2.2019656388326205,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.020936888085557e-05,
+      "loss": 0.5191,
+      "step": 7339
+    },
+    {
+      "epoch": 2.2022657363643185,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.019430637144148e-05,
+      "loss": 0.5473,
+      "step": 7340
+    },
+    {
+      "epoch": 2.202565833896016,
+      "grad_norm": 1.2265625,
+      "learning_rate": 9.017924386202742e-05,
+      "loss": 0.5369,
+      "step": 7341
+    },
+    {
+      "epoch": 2.202865931427714,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.016418135261336e-05,
+      "loss": 0.5868,
+      "step": 7342
+    },
+    {
+      "epoch": 2.2031660289594117,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.014911884319927e-05,
+      "loss": 0.5947,
+      "step": 7343
+    },
+    {
+      "epoch": 2.2034661264911097,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.013405633378521e-05,
+      "loss": 0.6028,
+      "step": 7344
+    },
+    {
+      "epoch": 2.2037662240228073,
+      "grad_norm": 0.9921875,
+      "learning_rate": 9.011899382437114e-05,
+      "loss": 0.5814,
+      "step": 7345
+    },
+    {
+      "epoch": 2.2040663215545053,
+      "grad_norm": 1.1484375,
+      "learning_rate": 9.010393131495708e-05,
+      "loss": 0.6057,
+      "step": 7346
+    },
+    {
+      "epoch": 2.204366419086203,
+      "grad_norm": 1.140625,
+      "learning_rate": 9.0088868805543e-05,
+      "loss": 0.5632,
+      "step": 7347
+    },
+    {
+      "epoch": 2.204666516617901,
+      "grad_norm": 1.0859375,
+      "learning_rate": 9.007380629612893e-05,
+      "loss": 0.6148,
+      "step": 7348
+    },
+    {
+      "epoch": 2.2049666141495985,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.005874378671487e-05,
+      "loss": 0.6548,
+      "step": 7349
+    },
+    {
+      "epoch": 2.2052667116812965,
+      "grad_norm": 1.0078125,
+      "learning_rate": 9.00436812773008e-05,
+      "loss": 0.5472,
+      "step": 7350
+    },
+    {
+      "epoch": 2.205566809212994,
+      "grad_norm": 1.2890625,
+      "learning_rate": 9.002861876788673e-05,
+      "loss": 0.7015,
+      "step": 7351
+    },
+    {
+      "epoch": 2.205866906744692,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.001355625847267e-05,
+      "loss": 0.5506,
+      "step": 7352
+    },
+    {
+      "epoch": 2.2061670042763897,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.99984937490586e-05,
+      "loss": 0.5745,
+      "step": 7353
+    },
+    {
+      "epoch": 2.2064671018080877,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.998343123964453e-05,
+      "loss": 0.4445,
+      "step": 7354
+    },
+    {
+      "epoch": 2.2067671993397853,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.996836873023046e-05,
+      "loss": 0.6135,
+      "step": 7355
+    },
+    {
+      "epoch": 2.2070672968714833,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.995330622081639e-05,
+      "loss": 0.5681,
+      "step": 7356
+    },
+    {
+      "epoch": 2.207367394403181,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.993824371140233e-05,
+      "loss": 0.6254,
+      "step": 7357
+    },
+    {
+      "epoch": 2.207667491934879,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.992318120198826e-05,
+      "loss": 0.5356,
+      "step": 7358
+    },
+    {
+      "epoch": 2.2079675894665765,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.990811869257418e-05,
+      "loss": 0.745,
+      "step": 7359
+    },
+    {
+      "epoch": 2.2082676869982745,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.989305618316012e-05,
+      "loss": 0.5455,
+      "step": 7360
+    },
+    {
+      "epoch": 2.208567784529972,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.987799367374605e-05,
+      "loss": 0.5039,
+      "step": 7361
+    },
+    {
+      "epoch": 2.20886788206167,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.986293116433199e-05,
+      "loss": 0.4831,
+      "step": 7362
+    },
+    {
+      "epoch": 2.2091679795933676,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.984786865491792e-05,
+      "loss": 0.5222,
+      "step": 7363
+    },
+    {
+      "epoch": 2.2094680771250657,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.983280614550384e-05,
+      "loss": 0.5235,
+      "step": 7364
+    },
+    {
+      "epoch": 2.2097681746567632,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.981774363608978e-05,
+      "loss": 0.6157,
+      "step": 7365
+    },
+    {
+      "epoch": 2.2100682721884612,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.980268112667571e-05,
+      "loss": 0.7034,
+      "step": 7366
+    },
+    {
+      "epoch": 2.2103683697201593,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.978761861726164e-05,
+      "loss": 0.5453,
+      "step": 7367
+    },
+    {
+      "epoch": 2.210668467251857,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.977255610784758e-05,
+      "loss": 0.7258,
+      "step": 7368
+    },
+    {
+      "epoch": 2.210968564783555,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.97574935984335e-05,
+      "loss": 0.6714,
+      "step": 7369
+    },
+    {
+      "epoch": 2.2112686623152524,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.974243108901943e-05,
+      "loss": 0.6079,
+      "step": 7370
+    },
+    {
+      "epoch": 2.2115687598469504,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.972736857960537e-05,
+      "loss": 0.5769,
+      "step": 7371
+    },
+    {
+      "epoch": 2.211868857378648,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.97123060701913e-05,
+      "loss": 0.6067,
+      "step": 7372
+    },
+    {
+      "epoch": 2.212168954910346,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.969724356077724e-05,
+      "loss": 0.6317,
+      "step": 7373
+    },
+    {
+      "epoch": 2.2124690524420436,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.968218105136315e-05,
+      "loss": 0.4658,
+      "step": 7374
+    },
+    {
+      "epoch": 2.2127691499737416,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.966711854194909e-05,
+      "loss": 0.5666,
+      "step": 7375
+    },
+    {
+      "epoch": 2.213069247505439,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.965205603253503e-05,
+      "loss": 0.6645,
+      "step": 7376
+    },
+    {
+      "epoch": 2.2133693450371372,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.963699352312096e-05,
+      "loss": 0.6057,
+      "step": 7377
+    },
+    {
+      "epoch": 2.213669442568835,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.962193101370689e-05,
+      "loss": 0.5287,
+      "step": 7378
+    },
+    {
+      "epoch": 2.213969540100533,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.960686850429281e-05,
+      "loss": 0.489,
+      "step": 7379
+    },
+    {
+      "epoch": 2.2142696376322304,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.959180599487875e-05,
+      "loss": 0.6597,
+      "step": 7380
+    },
+    {
+      "epoch": 2.2145697351639284,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.957674348546469e-05,
+      "loss": 0.6035,
+      "step": 7381
+    },
+    {
+      "epoch": 2.214869832695626,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.95616809760506e-05,
+      "loss": 0.6191,
+      "step": 7382
+    },
+    {
+      "epoch": 2.215169930227324,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.954661846663655e-05,
+      "loss": 0.4766,
+      "step": 7383
+    },
+    {
+      "epoch": 2.2154700277590216,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.953155595722247e-05,
+      "loss": 0.8451,
+      "step": 7384
+    },
+    {
+      "epoch": 2.2157701252907196,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.951649344780841e-05,
+      "loss": 0.5472,
+      "step": 7385
+    },
+    {
+      "epoch": 2.216070222822417,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.950143093839434e-05,
+      "loss": 0.6402,
+      "step": 7386
+    },
+    {
+      "epoch": 2.216370320354115,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.948636842898027e-05,
+      "loss": 0.5537,
+      "step": 7387
+    },
+    {
+      "epoch": 2.2166704178858128,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.947130591956621e-05,
+      "loss": 0.6505,
+      "step": 7388
+    },
+    {
+      "epoch": 2.216970515417511,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.945624341015215e-05,
+      "loss": 0.4146,
+      "step": 7389
+    },
+    {
+      "epoch": 2.2172706129492084,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.944118090073806e-05,
+      "loss": 0.5357,
+      "step": 7390
+    },
+    {
+      "epoch": 2.2175707104809064,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.9426118391324e-05,
+      "loss": 0.5771,
+      "step": 7391
+    },
+    {
+      "epoch": 2.217870808012604,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.941105588190993e-05,
+      "loss": 0.5703,
+      "step": 7392
+    },
+    {
+      "epoch": 2.218170905544302,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.939599337249587e-05,
+      "loss": 0.5903,
+      "step": 7393
+    },
+    {
+      "epoch": 2.2184710030759995,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.93809308630818e-05,
+      "loss": 0.5259,
+      "step": 7394
+    },
+    {
+      "epoch": 2.2187711006076976,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.936586835366772e-05,
+      "loss": 0.6574,
+      "step": 7395
+    },
+    {
+      "epoch": 2.219071198139395,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.935080584425366e-05,
+      "loss": 0.6648,
+      "step": 7396
+    },
+    {
+      "epoch": 2.219371295671093,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.933574333483959e-05,
+      "loss": 0.5721,
+      "step": 7397
+    },
+    {
+      "epoch": 2.2196713932027907,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.932068082542552e-05,
+      "loss": 0.6007,
+      "step": 7398
+    },
+    {
+      "epoch": 2.2199714907344887,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.930561831601146e-05,
+      "loss": 0.5154,
+      "step": 7399
+    },
+    {
+      "epoch": 2.2202715882661863,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.929055580659738e-05,
+      "loss": 0.562,
+      "step": 7400
+    },
+    {
+      "epoch": 2.2205716857978843,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.927549329718331e-05,
+      "loss": 0.6659,
+      "step": 7401
+    },
+    {
+      "epoch": 2.2208717833295823,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.926043078776925e-05,
+      "loss": 0.5925,
+      "step": 7402
+    },
+    {
+      "epoch": 2.22117188086128,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.924536827835518e-05,
+      "loss": 0.5102,
+      "step": 7403
+    },
+    {
+      "epoch": 2.2214719783929775,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.923030576894112e-05,
+      "loss": 0.5744,
+      "step": 7404
+    },
+    {
+      "epoch": 2.2217720759246755,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.921524325952704e-05,
+      "loss": 0.5796,
+      "step": 7405
+    },
+    {
+      "epoch": 2.2220721734563735,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.920018075011297e-05,
+      "loss": 0.5061,
+      "step": 7406
+    },
+    {
+      "epoch": 2.222372270988071,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.918511824069891e-05,
+      "loss": 0.5806,
+      "step": 7407
+    },
+    {
+      "epoch": 2.222672368519769,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.917005573128484e-05,
+      "loss": 0.5454,
+      "step": 7408
+    },
+    {
+      "epoch": 2.2229724660514667,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.915499322187076e-05,
+      "loss": 0.6157,
+      "step": 7409
+    },
+    {
+      "epoch": 2.2232725635831647,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.91399307124567e-05,
+      "loss": 0.5007,
+      "step": 7410
+    },
+    {
+      "epoch": 2.2235726611148623,
+      "grad_norm": 1.0,
+      "learning_rate": 8.912486820304263e-05,
+      "loss": 0.5116,
+      "step": 7411
+    },
+    {
+      "epoch": 2.2238727586465603,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.910980569362857e-05,
+      "loss": 0.5498,
+      "step": 7412
+    },
+    {
+      "epoch": 2.224172856178258,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.90947431842145e-05,
+      "loss": 0.6781,
+      "step": 7413
+    },
+    {
+      "epoch": 2.224472953709956,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.907968067480042e-05,
+      "loss": 0.539,
+      "step": 7414
+    },
+    {
+      "epoch": 2.2247730512416535,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.906461816538636e-05,
+      "loss": 0.5933,
+      "step": 7415
+    },
+    {
+      "epoch": 2.2250731487733515,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.904955565597229e-05,
+      "loss": 0.599,
+      "step": 7416
+    },
+    {
+      "epoch": 2.225373246305049,
+      "grad_norm": 1.0,
+      "learning_rate": 8.903449314655822e-05,
+      "loss": 0.5988,
+      "step": 7417
+    },
+    {
+      "epoch": 2.225673343836747,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.901943063714416e-05,
+      "loss": 0.6903,
+      "step": 7418
+    },
+    {
+      "epoch": 2.2259734413684447,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.900436812773009e-05,
+      "loss": 0.5557,
+      "step": 7419
+    },
+    {
+      "epoch": 2.2262735389001427,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.898930561831603e-05,
+      "loss": 0.5461,
+      "step": 7420
+    },
+    {
+      "epoch": 2.2265736364318403,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.897424310890194e-05,
+      "loss": 0.5442,
+      "step": 7421
+    },
+    {
+      "epoch": 2.2268737339635383,
+      "grad_norm": 1.0,
+      "learning_rate": 8.895918059948788e-05,
+      "loss": 0.5822,
+      "step": 7422
+    },
+    {
+      "epoch": 2.227173831495236,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.894411809007382e-05,
+      "loss": 0.6066,
+      "step": 7423
+    },
+    {
+      "epoch": 2.227473929026934,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.892905558065973e-05,
+      "loss": 0.6065,
+      "step": 7424
+    },
+    {
+      "epoch": 2.2277740265586314,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.891399307124567e-05,
+      "loss": 0.5836,
+      "step": 7425
+    },
+    {
+      "epoch": 2.2280741240903295,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.88989305618316e-05,
+      "loss": 0.6169,
+      "step": 7426
+    },
+    {
+      "epoch": 2.228374221622027,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.888386805241754e-05,
+      "loss": 0.6031,
+      "step": 7427
+    },
+    {
+      "epoch": 2.228674319153725,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.886880554300347e-05,
+      "loss": 0.5609,
+      "step": 7428
+    },
+    {
+      "epoch": 2.2289744166854226,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.88537430335894e-05,
+      "loss": 0.5286,
+      "step": 7429
+    },
+    {
+      "epoch": 2.2292745142171206,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.883868052417533e-05,
+      "loss": 0.5731,
+      "step": 7430
+    },
+    {
+      "epoch": 2.229574611748818,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.882361801476127e-05,
+      "loss": 0.634,
+      "step": 7431
+    },
+    {
+      "epoch": 2.2298747092805162,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.880855550534719e-05,
+      "loss": 0.5976,
+      "step": 7432
+    },
+    {
+      "epoch": 2.230174806812214,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.879349299593313e-05,
+      "loss": 0.6267,
+      "step": 7433
+    },
+    {
+      "epoch": 2.230474904343912,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.877843048651905e-05,
+      "loss": 0.5972,
+      "step": 7434
+    },
+    {
+      "epoch": 2.2307750018756094,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.8763367977105e-05,
+      "loss": 0.4843,
+      "step": 7435
+    },
+    {
+      "epoch": 2.2310750994073074,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.874830546769092e-05,
+      "loss": 0.5551,
+      "step": 7436
+    },
+    {
+      "epoch": 2.231375196939005,
+      "grad_norm": 1.125,
+      "learning_rate": 8.873324295827685e-05,
+      "loss": 0.5129,
+      "step": 7437
+    },
+    {
+      "epoch": 2.231675294470703,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.871818044886279e-05,
+      "loss": 0.471,
+      "step": 7438
+    },
+    {
+      "epoch": 2.2319753920024006,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.870311793944872e-05,
+      "loss": 0.5391,
+      "step": 7439
+    },
+    {
+      "epoch": 2.2322754895340986,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.868805543003464e-05,
+      "loss": 0.6278,
+      "step": 7440
+    },
+    {
+      "epoch": 2.2325755870657966,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.867299292062058e-05,
+      "loss": 0.5549,
+      "step": 7441
+    },
+    {
+      "epoch": 2.232875684597494,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.865793041120651e-05,
+      "loss": 0.5555,
+      "step": 7442
+    },
+    {
+      "epoch": 2.2331757821291918,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.864286790179245e-05,
+      "loss": 0.5891,
+      "step": 7443
+    },
+    {
+      "epoch": 2.23347587966089,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.862780539237838e-05,
+      "loss": 0.5528,
+      "step": 7444
+    },
+    {
+      "epoch": 2.233775977192588,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.86127428829643e-05,
+      "loss": 0.5108,
+      "step": 7445
+    },
+    {
+      "epoch": 2.2340760747242854,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.859768037355024e-05,
+      "loss": 0.6954,
+      "step": 7446
+    },
+    {
+      "epoch": 2.2343761722559834,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.858261786413617e-05,
+      "loss": 0.6005,
+      "step": 7447
+    },
+    {
+      "epoch": 2.234676269787681,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.85675553547221e-05,
+      "loss": 0.5003,
+      "step": 7448
+    },
+    {
+      "epoch": 2.234976367319379,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.855249284530804e-05,
+      "loss": 0.5857,
+      "step": 7449
+    },
+    {
+      "epoch": 2.2352764648510766,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.853743033589396e-05,
+      "loss": 0.598,
+      "step": 7450
+    },
+    {
+      "epoch": 2.2355765623827746,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.852236782647989e-05,
+      "loss": 0.581,
+      "step": 7451
+    },
+    {
+      "epoch": 2.235876659914472,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.850730531706583e-05,
+      "loss": 0.5834,
+      "step": 7452
+    },
+    {
+      "epoch": 2.23617675744617,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.849224280765176e-05,
+      "loss": 0.6195,
+      "step": 7453
+    },
+    {
+      "epoch": 2.2364768549778677,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.84771802982377e-05,
+      "loss": 0.5806,
+      "step": 7454
+    },
+    {
+      "epoch": 2.2367769525095658,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.846211778882361e-05,
+      "loss": 0.6075,
+      "step": 7455
+    },
+    {
+      "epoch": 2.2370770500412633,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.844705527940955e-05,
+      "loss": 0.5023,
+      "step": 7456
+    },
+    {
+      "epoch": 2.2373771475729614,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.843199276999549e-05,
+      "loss": 0.6808,
+      "step": 7457
+    },
+    {
+      "epoch": 2.237677245104659,
+      "grad_norm": 1.125,
+      "learning_rate": 8.841693026058142e-05,
+      "loss": 0.7177,
+      "step": 7458
+    },
+    {
+      "epoch": 2.237977342636357,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.840186775116735e-05,
+      "loss": 0.4611,
+      "step": 7459
+    },
+    {
+      "epoch": 2.2382774401680545,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.838680524175329e-05,
+      "loss": 0.5548,
+      "step": 7460
+    },
+    {
+      "epoch": 2.2385775376997525,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.837174273233921e-05,
+      "loss": 0.6345,
+      "step": 7461
+    },
+    {
+      "epoch": 2.23887763523145,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.835668022292515e-05,
+      "loss": 0.6176,
+      "step": 7462
+    },
+    {
+      "epoch": 2.239177732763148,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.834161771351107e-05,
+      "loss": 0.5294,
+      "step": 7463
+    },
+    {
+      "epoch": 2.2394778302948457,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.8326555204097e-05,
+      "loss": 0.5465,
+      "step": 7464
+    },
+    {
+      "epoch": 2.2397779278265437,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.831149269468295e-05,
+      "loss": 0.5799,
+      "step": 7465
+    },
+    {
+      "epoch": 2.2400780253582413,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.829643018526887e-05,
+      "loss": 0.5583,
+      "step": 7466
+    },
+    {
+      "epoch": 2.2403781228899393,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.82813676758548e-05,
+      "loss": 0.5947,
+      "step": 7467
+    },
+    {
+      "epoch": 2.240678220421637,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.826630516644073e-05,
+      "loss": 0.6835,
+      "step": 7468
+    },
+    {
+      "epoch": 2.240978317953335,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.825124265702667e-05,
+      "loss": 0.5823,
+      "step": 7469
+    },
+    {
+      "epoch": 2.2412784154850325,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.823618014761261e-05,
+      "loss": 0.5859,
+      "step": 7470
+    },
+    {
+      "epoch": 2.2415785130167305,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.822111763819852e-05,
+      "loss": 0.6196,
+      "step": 7471
+    },
+    {
+      "epoch": 2.241878610548428,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.820605512878446e-05,
+      "loss": 0.5793,
+      "step": 7472
+    },
+    {
+      "epoch": 2.242178708080126,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.819099261937039e-05,
+      "loss": 0.5429,
+      "step": 7473
+    },
+    {
+      "epoch": 2.2424788056118237,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.817593010995633e-05,
+      "loss": 0.6461,
+      "step": 7474
+    },
+    {
+      "epoch": 2.2427789031435217,
+      "grad_norm": 1.421875,
+      "learning_rate": 8.816086760054225e-05,
+      "loss": 0.7839,
+      "step": 7475
+    },
+    {
+      "epoch": 2.2430790006752193,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.814580509112818e-05,
+      "loss": 0.5496,
+      "step": 7476
+    },
+    {
+      "epoch": 2.2433790982069173,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.813074258171412e-05,
+      "loss": 0.5121,
+      "step": 7477
+    },
+    {
+      "epoch": 2.243679195738615,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.811568007230005e-05,
+      "loss": 0.586,
+      "step": 7478
+    },
+    {
+      "epoch": 2.243979293270313,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.810061756288598e-05,
+      "loss": 0.7382,
+      "step": 7479
+    },
+    {
+      "epoch": 2.244279390802011,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.808555505347192e-05,
+      "loss": 0.5743,
+      "step": 7480
+    },
+    {
+      "epoch": 2.2445794883337085,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.807049254405784e-05,
+      "loss": 0.5978,
+      "step": 7481
+    },
+    {
+      "epoch": 2.244879585865406,
+      "grad_norm": 1.125,
+      "learning_rate": 8.805543003464377e-05,
+      "loss": 0.5535,
+      "step": 7482
+    },
+    {
+      "epoch": 2.245179683397104,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.804036752522971e-05,
+      "loss": 0.5354,
+      "step": 7483
+    },
+    {
+      "epoch": 2.245479780928802,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.802530501581564e-05,
+      "loss": 0.5068,
+      "step": 7484
+    },
+    {
+      "epoch": 2.2457798784604996,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.801024250640158e-05,
+      "loss": 0.5573,
+      "step": 7485
+    },
+    {
+      "epoch": 2.2460799759921977,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.79951799969875e-05,
+      "loss": 0.6286,
+      "step": 7486
+    },
+    {
+      "epoch": 2.2463800735238952,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.798011748757343e-05,
+      "loss": 0.5455,
+      "step": 7487
+    },
+    {
+      "epoch": 2.2466801710555933,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.796505497815937e-05,
+      "loss": 0.5382,
+      "step": 7488
+    },
+    {
+      "epoch": 2.246980268587291,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.79499924687453e-05,
+      "loss": 0.5962,
+      "step": 7489
+    },
+    {
+      "epoch": 2.247280366118989,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.793492995933122e-05,
+      "loss": 0.7215,
+      "step": 7490
+    },
+    {
+      "epoch": 2.2475804636506864,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.791986744991716e-05,
+      "loss": 0.6629,
+      "step": 7491
+    },
+    {
+      "epoch": 2.2478805611823844,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.790480494050309e-05,
+      "loss": 0.6912,
+      "step": 7492
+    },
+    {
+      "epoch": 2.248180658714082,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.788974243108903e-05,
+      "loss": 0.5209,
+      "step": 7493
+    },
+    {
+      "epoch": 2.24848075624578,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.787467992167496e-05,
+      "loss": 0.607,
+      "step": 7494
+    },
+    {
+      "epoch": 2.2487808537774776,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.785961741226088e-05,
+      "loss": 0.4229,
+      "step": 7495
+    },
+    {
+      "epoch": 2.2490809513091756,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.784455490284682e-05,
+      "loss": 0.6055,
+      "step": 7496
+    },
+    {
+      "epoch": 2.249381048840873,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.782949239343275e-05,
+      "loss": 0.5834,
+      "step": 7497
+    },
+    {
+      "epoch": 2.249681146372571,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.781442988401868e-05,
+      "loss": 0.6449,
+      "step": 7498
+    },
+    {
+      "epoch": 2.249981243904269,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.779936737460462e-05,
+      "loss": 0.5848,
+      "step": 7499
+    },
+    {
+      "epoch": 2.250281341435967,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.778430486519055e-05,
+      "loss": 0.5075,
+      "step": 7500
+    },
+    {
+      "epoch": 2.2505814389676644,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.776924235577649e-05,
+      "loss": 0.7063,
+      "step": 7501
+    },
+    {
+      "epoch": 2.2508815364993624,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.77541798463624e-05,
+      "loss": 0.5606,
+      "step": 7502
+    },
+    {
+      "epoch": 2.25118163403106,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.773911733694834e-05,
+      "loss": 0.5724,
+      "step": 7503
+    },
+    {
+      "epoch": 2.251481731562758,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.772405482753428e-05,
+      "loss": 0.6321,
+      "step": 7504
+    },
+    {
+      "epoch": 2.2517818290944556,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.770899231812019e-05,
+      "loss": 0.5317,
+      "step": 7505
+    },
+    {
+      "epoch": 2.2520819266261536,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.769392980870613e-05,
+      "loss": 0.6045,
+      "step": 7506
+    },
+    {
+      "epoch": 2.252382024157851,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.767886729929207e-05,
+      "loss": 0.5186,
+      "step": 7507
+    },
+    {
+      "epoch": 2.252682121689549,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.7663804789878e-05,
+      "loss": 0.6414,
+      "step": 7508
+    },
+    {
+      "epoch": 2.2529822192212468,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.764874228046393e-05,
+      "loss": 0.6132,
+      "step": 7509
+    },
+    {
+      "epoch": 2.2532823167529448,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.763367977104985e-05,
+      "loss": 0.6508,
+      "step": 7510
+    },
+    {
+      "epoch": 2.2535824142846423,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.76186172616358e-05,
+      "loss": 0.6677,
+      "step": 7511
+    },
+    {
+      "epoch": 2.2538825118163404,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.760355475222173e-05,
+      "loss": 0.5441,
+      "step": 7512
+    },
+    {
+      "epoch": 2.254182609348038,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.758849224280765e-05,
+      "loss": 0.673,
+      "step": 7513
+    },
+    {
+      "epoch": 2.254482706879736,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.757342973339359e-05,
+      "loss": 0.5032,
+      "step": 7514
+    },
+    {
+      "epoch": 2.2547828044114335,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.755836722397951e-05,
+      "loss": 0.6277,
+      "step": 7515
+    },
+    {
+      "epoch": 2.2550829019431315,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.754330471456545e-05,
+      "loss": 0.7414,
+      "step": 7516
+    },
+    {
+      "epoch": 2.255382999474829,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.752824220515138e-05,
+      "loss": 0.5728,
+      "step": 7517
+    },
+    {
+      "epoch": 2.255683097006527,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.751317969573731e-05,
+      "loss": 0.5299,
+      "step": 7518
+    },
+    {
+      "epoch": 2.255983194538225,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.749811718632325e-05,
+      "loss": 0.6567,
+      "step": 7519
+    },
+    {
+      "epoch": 2.2562832920699227,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.748305467690918e-05,
+      "loss": 0.5573,
+      "step": 7520
+    },
+    {
+      "epoch": 2.2565833896016203,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.74679921674951e-05,
+      "loss": 0.587,
+      "step": 7521
+    },
+    {
+      "epoch": 2.2568834871333183,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.745292965808104e-05,
+      "loss": 0.6235,
+      "step": 7522
+    },
+    {
+      "epoch": 2.2571835846650163,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.743786714866697e-05,
+      "loss": 0.6094,
+      "step": 7523
+    },
+    {
+      "epoch": 2.257483682196714,
+      "grad_norm": 0.8828125,
+      "learning_rate": 8.742280463925291e-05,
+      "loss": 0.4933,
+      "step": 7524
+    },
+    {
+      "epoch": 2.2577837797284115,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.740774212983884e-05,
+      "loss": 0.5671,
+      "step": 7525
+    },
+    {
+      "epoch": 2.2580838772601095,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.739267962042476e-05,
+      "loss": 0.6515,
+      "step": 7526
+    },
+    {
+      "epoch": 2.2583839747918075,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.73776171110107e-05,
+      "loss": 0.5957,
+      "step": 7527
+    },
+    {
+      "epoch": 2.258684072323505,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.736255460159663e-05,
+      "loss": 0.6586,
+      "step": 7528
+    },
+    {
+      "epoch": 2.258984169855203,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.734749209218256e-05,
+      "loss": 0.6142,
+      "step": 7529
+    },
+    {
+      "epoch": 2.2592842673869007,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.73324295827685e-05,
+      "loss": 0.6116,
+      "step": 7530
+    },
+    {
+      "epoch": 2.2595843649185987,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.731736707335442e-05,
+      "loss": 0.5714,
+      "step": 7531
+    },
+    {
+      "epoch": 2.2598844624502963,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.730230456394035e-05,
+      "loss": 0.552,
+      "step": 7532
+    },
+    {
+      "epoch": 2.2601845599819943,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.728724205452629e-05,
+      "loss": 0.4905,
+      "step": 7533
+    },
+    {
+      "epoch": 2.260484657513692,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.727217954511222e-05,
+      "loss": 0.6491,
+      "step": 7534
+    },
+    {
+      "epoch": 2.26078475504539,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.725711703569816e-05,
+      "loss": 0.5791,
+      "step": 7535
+    },
+    {
+      "epoch": 2.2610848525770875,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.724205452628408e-05,
+      "loss": 0.4599,
+      "step": 7536
+    },
+    {
+      "epoch": 2.2613849501087855,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.722699201687001e-05,
+      "loss": 0.585,
+      "step": 7537
+    },
+    {
+      "epoch": 2.261685047640483,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.721192950745595e-05,
+      "loss": 0.5259,
+      "step": 7538
+    },
+    {
+      "epoch": 2.261985145172181,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.719686699804188e-05,
+      "loss": 0.6228,
+      "step": 7539
+    },
+    {
+      "epoch": 2.2622852427038787,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.71818044886278e-05,
+      "loss": 0.5648,
+      "step": 7540
+    },
+    {
+      "epoch": 2.2625853402355767,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.716674197921375e-05,
+      "loss": 0.652,
+      "step": 7541
+    },
+    {
+      "epoch": 2.2628854377672742,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.715167946979967e-05,
+      "loss": 0.5402,
+      "step": 7542
+    },
+    {
+      "epoch": 2.2631855352989723,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.713661696038561e-05,
+      "loss": 0.5496,
+      "step": 7543
+    },
+    {
+      "epoch": 2.26348563283067,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.712155445097153e-05,
+      "loss": 0.5037,
+      "step": 7544
+    },
+    {
+      "epoch": 2.263785730362368,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.710649194155747e-05,
+      "loss": 0.621,
+      "step": 7545
+    },
+    {
+      "epoch": 2.2640858278940654,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.70914294321434e-05,
+      "loss": 0.604,
+      "step": 7546
+    },
+    {
+      "epoch": 2.2643859254257634,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.707636692272933e-05,
+      "loss": 0.6706,
+      "step": 7547
+    },
+    {
+      "epoch": 2.264686022957461,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.706130441331526e-05,
+      "loss": 0.5481,
+      "step": 7548
+    },
+    {
+      "epoch": 2.264986120489159,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.70462419039012e-05,
+      "loss": 0.6784,
+      "step": 7549
+    },
+    {
+      "epoch": 2.2652862180208566,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.703117939448713e-05,
+      "loss": 0.5841,
+      "step": 7550
+    },
+    {
+      "epoch": 2.2655863155525546,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.701611688507307e-05,
+      "loss": 0.5578,
+      "step": 7551
+    },
+    {
+      "epoch": 2.265886413084252,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.700105437565898e-05,
+      "loss": 0.6243,
+      "step": 7552
+    },
+    {
+      "epoch": 2.2661865106159502,
+      "grad_norm": 1.125,
+      "learning_rate": 8.698599186624492e-05,
+      "loss": 0.5999,
+      "step": 7553
+    },
+    {
+      "epoch": 2.266486608147648,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.697092935683086e-05,
+      "loss": 0.5458,
+      "step": 7554
+    },
+    {
+      "epoch": 2.266786705679346,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.695586684741679e-05,
+      "loss": 0.5787,
+      "step": 7555
+    },
+    {
+      "epoch": 2.2670868032110434,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.694080433800271e-05,
+      "loss": 0.5578,
+      "step": 7556
+    },
+    {
+      "epoch": 2.2673869007427414,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.692574182858864e-05,
+      "loss": 0.6064,
+      "step": 7557
+    },
+    {
+      "epoch": 2.2676869982744394,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.691067931917458e-05,
+      "loss": 0.5646,
+      "step": 7558
+    },
+    {
+      "epoch": 2.267987095806137,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.689561680976051e-05,
+      "loss": 0.5316,
+      "step": 7559
+    },
+    {
+      "epoch": 2.2682871933378346,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.688055430034644e-05,
+      "loss": 0.5074,
+      "step": 7560
+    },
+    {
+      "epoch": 2.2685872908695326,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.686549179093238e-05,
+      "loss": 0.5861,
+      "step": 7561
+    },
+    {
+      "epoch": 2.2688873884012306,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.68504292815183e-05,
+      "loss": 0.6041,
+      "step": 7562
+    },
+    {
+      "epoch": 2.269187485932928,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.683536677210423e-05,
+      "loss": 0.5821,
+      "step": 7563
+    },
+    {
+      "epoch": 2.2694875834646258,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.682030426269017e-05,
+      "loss": 0.6758,
+      "step": 7564
+    },
+    {
+      "epoch": 2.2697876809963238,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.68052417532761e-05,
+      "loss": 0.6664,
+      "step": 7565
+    },
+    {
+      "epoch": 2.270087778528022,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.679017924386204e-05,
+      "loss": 0.6546,
+      "step": 7566
+    },
+    {
+      "epoch": 2.2703878760597194,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.677511673444796e-05,
+      "loss": 0.5251,
+      "step": 7567
+    },
+    {
+      "epoch": 2.2706879735914174,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.676005422503389e-05,
+      "loss": 0.69,
+      "step": 7568
+    },
+    {
+      "epoch": 2.270988071123115,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.674499171561983e-05,
+      "loss": 0.5391,
+      "step": 7569
+    },
+    {
+      "epoch": 2.271288168654813,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.672992920620576e-05,
+      "loss": 0.5009,
+      "step": 7570
+    },
+    {
+      "epoch": 2.2715882661865106,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.671486669679168e-05,
+      "loss": 0.5344,
+      "step": 7571
+    },
+    {
+      "epoch": 2.2718883637182086,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.669980418737762e-05,
+      "loss": 0.4989,
+      "step": 7572
+    },
+    {
+      "epoch": 2.272188461249906,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.668474167796355e-05,
+      "loss": 0.6677,
+      "step": 7573
+    },
+    {
+      "epoch": 2.272488558781604,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.666967916854949e-05,
+      "loss": 0.545,
+      "step": 7574
+    },
+    {
+      "epoch": 2.2727886563133017,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.665461665913542e-05,
+      "loss": 0.6539,
+      "step": 7575
+    },
+    {
+      "epoch": 2.2730887538449998,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.663955414972134e-05,
+      "loss": 0.5907,
+      "step": 7576
+    },
+    {
+      "epoch": 2.2733888513766973,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.662449164030728e-05,
+      "loss": 0.5262,
+      "step": 7577
+    },
+    {
+      "epoch": 2.2736889489083953,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.660942913089321e-05,
+      "loss": 0.5569,
+      "step": 7578
+    },
+    {
+      "epoch": 2.273989046440093,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.659436662147914e-05,
+      "loss": 0.6257,
+      "step": 7579
+    },
+    {
+      "epoch": 2.274289143971791,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.657930411206508e-05,
+      "loss": 0.5657,
+      "step": 7580
+    },
+    {
+      "epoch": 2.2745892415034885,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.6564241602651e-05,
+      "loss": 0.5924,
+      "step": 7581
+    },
+    {
+      "epoch": 2.2748893390351865,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.654917909323695e-05,
+      "loss": 0.6698,
+      "step": 7582
+    },
+    {
+      "epoch": 2.275189436566884,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.653411658382287e-05,
+      "loss": 0.558,
+      "step": 7583
+    },
+    {
+      "epoch": 2.275489534098582,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.65190540744088e-05,
+      "loss": 0.5466,
+      "step": 7584
+    },
+    {
+      "epoch": 2.2757896316302797,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.650399156499474e-05,
+      "loss": 0.5835,
+      "step": 7585
+    },
+    {
+      "epoch": 2.2760897291619777,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.648892905558065e-05,
+      "loss": 0.5259,
+      "step": 7586
+    },
+    {
+      "epoch": 2.2763898266936753,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.647386654616659e-05,
+      "loss": 0.6108,
+      "step": 7587
+    },
+    {
+      "epoch": 2.2766899242253733,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.645880403675253e-05,
+      "loss": 0.6463,
+      "step": 7588
+    },
+    {
+      "epoch": 2.276990021757071,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.644374152733846e-05,
+      "loss": 0.5751,
+      "step": 7589
+    },
+    {
+      "epoch": 2.277290119288769,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.642867901792439e-05,
+      "loss": 0.5849,
+      "step": 7590
+    },
+    {
+      "epoch": 2.2775902168204665,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.641361650851031e-05,
+      "loss": 0.5373,
+      "step": 7591
+    },
+    {
+      "epoch": 2.2778903143521645,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.639855399909625e-05,
+      "loss": 0.5407,
+      "step": 7592
+    },
+    {
+      "epoch": 2.278190411883862,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.63834914896822e-05,
+      "loss": 0.519,
+      "step": 7593
+    },
+    {
+      "epoch": 2.27849050941556,
+      "grad_norm": 0.875,
+      "learning_rate": 8.636842898026811e-05,
+      "loss": 0.5062,
+      "step": 7594
+    },
+    {
+      "epoch": 2.2787906069472577,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.635336647085405e-05,
+      "loss": 0.5906,
+      "step": 7595
+    },
+    {
+      "epoch": 2.2790907044789557,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.633830396143999e-05,
+      "loss": 0.5555,
+      "step": 7596
+    },
+    {
+      "epoch": 2.2793908020106537,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.632324145202591e-05,
+      "loss": 0.6143,
+      "step": 7597
+    },
+    {
+      "epoch": 2.2796908995423513,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.630817894261184e-05,
+      "loss": 0.5258,
+      "step": 7598
+    },
+    {
+      "epoch": 2.279990997074049,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.629311643319777e-05,
+      "loss": 0.5412,
+      "step": 7599
+    },
+    {
+      "epoch": 2.280291094605747,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.627805392378371e-05,
+      "loss": 0.5646,
+      "step": 7600
+    },
+    {
+      "epoch": 2.280591192137445,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.626299141436965e-05,
+      "loss": 0.6087,
+      "step": 7601
+    },
+    {
+      "epoch": 2.2808912896691425,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.624792890495556e-05,
+      "loss": 0.5638,
+      "step": 7602
+    },
+    {
+      "epoch": 2.28119138720084,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.62328663955415e-05,
+      "loss": 0.5868,
+      "step": 7603
+    },
+    {
+      "epoch": 2.281491484732538,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.621780388612743e-05,
+      "loss": 0.566,
+      "step": 7604
+    },
+    {
+      "epoch": 2.281791582264236,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.620274137671337e-05,
+      "loss": 0.4795,
+      "step": 7605
+    },
+    {
+      "epoch": 2.2820916797959336,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.61876788672993e-05,
+      "loss": 0.522,
+      "step": 7606
+    },
+    {
+      "epoch": 2.2823917773276317,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.617261635788522e-05,
+      "loss": 0.5996,
+      "step": 7607
+    },
+    {
+      "epoch": 2.2826918748593292,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.615755384847116e-05,
+      "loss": 0.6532,
+      "step": 7608
+    },
+    {
+      "epoch": 2.2829919723910272,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.614249133905709e-05,
+      "loss": 0.4921,
+      "step": 7609
+    },
+    {
+      "epoch": 2.283292069922725,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.612742882964302e-05,
+      "loss": 0.6046,
+      "step": 7610
+    },
+    {
+      "epoch": 2.283592167454423,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.611236632022896e-05,
+      "loss": 0.6289,
+      "step": 7611
+    },
+    {
+      "epoch": 2.2838922649861204,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.609730381081488e-05,
+      "loss": 0.6033,
+      "step": 7612
+    },
+    {
+      "epoch": 2.2841923625178184,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.608224130140081e-05,
+      "loss": 0.5316,
+      "step": 7613
+    },
+    {
+      "epoch": 2.284492460049516,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.606717879198675e-05,
+      "loss": 0.6537,
+      "step": 7614
+    },
+    {
+      "epoch": 2.284792557581214,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.605211628257268e-05,
+      "loss": 0.6152,
+      "step": 7615
+    },
+    {
+      "epoch": 2.2850926551129116,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.603705377315862e-05,
+      "loss": 0.525,
+      "step": 7616
+    },
+    {
+      "epoch": 2.2853927526446096,
+      "grad_norm": 1.125,
+      "learning_rate": 8.602199126374454e-05,
+      "loss": 0.6308,
+      "step": 7617
+    },
+    {
+      "epoch": 2.285692850176307,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.600692875433047e-05,
+      "loss": 0.5935,
+      "step": 7618
+    },
+    {
+      "epoch": 2.285992947708005,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.599186624491641e-05,
+      "loss": 0.5662,
+      "step": 7619
+    },
+    {
+      "epoch": 2.286293045239703,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.597680373550234e-05,
+      "loss": 0.539,
+      "step": 7620
+    },
+    {
+      "epoch": 2.286593142771401,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.596174122608827e-05,
+      "loss": 0.7731,
+      "step": 7621
+    },
+    {
+      "epoch": 2.2868932403030984,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.59466787166742e-05,
+      "loss": 0.5983,
+      "step": 7622
+    },
+    {
+      "epoch": 2.2871933378347964,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.593161620726013e-05,
+      "loss": 0.7112,
+      "step": 7623
+    },
+    {
+      "epoch": 2.287493435366494,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.591655369784607e-05,
+      "loss": 0.5841,
+      "step": 7624
+    },
+    {
+      "epoch": 2.287793532898192,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.5901491188432e-05,
+      "loss": 0.5816,
+      "step": 7625
+    },
+    {
+      "epoch": 2.2880936304298896,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.588642867901793e-05,
+      "loss": 0.6434,
+      "step": 7626
+    },
+    {
+      "epoch": 2.2883937279615876,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.587136616960387e-05,
+      "loss": 0.6221,
+      "step": 7627
+    },
+    {
+      "epoch": 2.288693825493285,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.585630366018979e-05,
+      "loss": 0.5641,
+      "step": 7628
+    },
+    {
+      "epoch": 2.288993923024983,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.584124115077572e-05,
+      "loss": 0.5056,
+      "step": 7629
+    },
+    {
+      "epoch": 2.2892940205566807,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.582617864136166e-05,
+      "loss": 0.6418,
+      "step": 7630
+    },
+    {
+      "epoch": 2.2895941180883788,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.581111613194759e-05,
+      "loss": 0.6287,
+      "step": 7631
+    },
+    {
+      "epoch": 2.2898942156200763,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.579605362253353e-05,
+      "loss": 0.6159,
+      "step": 7632
+    },
+    {
+      "epoch": 2.2901943131517744,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.578099111311944e-05,
+      "loss": 0.5228,
+      "step": 7633
+    },
+    {
+      "epoch": 2.290494410683472,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.576592860370538e-05,
+      "loss": 0.4528,
+      "step": 7634
+    },
+    {
+      "epoch": 2.29079450821517,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.575086609429132e-05,
+      "loss": 0.5735,
+      "step": 7635
+    },
+    {
+      "epoch": 2.291094605746868,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.573580358487725e-05,
+      "loss": 0.5673,
+      "step": 7636
+    },
+    {
+      "epoch": 2.2913947032785655,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.572074107546317e-05,
+      "loss": 0.5442,
+      "step": 7637
+    },
+    {
+      "epoch": 2.291694800810263,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.57056785660491e-05,
+      "loss": 0.5664,
+      "step": 7638
+    },
+    {
+      "epoch": 2.291994898341961,
+      "grad_norm": 0.875,
+      "learning_rate": 8.569061605663504e-05,
+      "loss": 0.5063,
+      "step": 7639
+    },
+    {
+      "epoch": 2.292294995873659,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.567555354722097e-05,
+      "loss": 0.5117,
+      "step": 7640
+    },
+    {
+      "epoch": 2.2925950934053567,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.56604910378069e-05,
+      "loss": 0.6208,
+      "step": 7641
+    },
+    {
+      "epoch": 2.2928951909370543,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.564542852839284e-05,
+      "loss": 0.5605,
+      "step": 7642
+    },
+    {
+      "epoch": 2.2931952884687523,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.563036601897878e-05,
+      "loss": 0.6142,
+      "step": 7643
+    },
+    {
+      "epoch": 2.2934953860004503,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.561530350956469e-05,
+      "loss": 0.559,
+      "step": 7644
+    },
+    {
+      "epoch": 2.293795483532148,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.560024100015063e-05,
+      "loss": 0.5329,
+      "step": 7645
+    },
+    {
+      "epoch": 2.294095581063846,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.558517849073656e-05,
+      "loss": 0.5662,
+      "step": 7646
+    },
+    {
+      "epoch": 2.2943956785955435,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.55701159813225e-05,
+      "loss": 0.5497,
+      "step": 7647
+    },
+    {
+      "epoch": 2.2946957761272415,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.555505347190842e-05,
+      "loss": 0.5958,
+      "step": 7648
+    },
+    {
+      "epoch": 2.294995873658939,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.553999096249435e-05,
+      "loss": 0.5522,
+      "step": 7649
+    },
+    {
+      "epoch": 2.295295971190637,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.552492845308029e-05,
+      "loss": 0.4633,
+      "step": 7650
+    },
+    {
+      "epoch": 2.2955960687223347,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.550986594366622e-05,
+      "loss": 0.5196,
+      "step": 7651
+    },
+    {
+      "epoch": 2.2958961662540327,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.549480343425214e-05,
+      "loss": 0.5733,
+      "step": 7652
+    },
+    {
+      "epoch": 2.2961962637857303,
+      "grad_norm": 1.125,
+      "learning_rate": 8.547974092483808e-05,
+      "loss": 0.539,
+      "step": 7653
+    },
+    {
+      "epoch": 2.2964963613174283,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.546467841542401e-05,
+      "loss": 0.5647,
+      "step": 7654
+    },
+    {
+      "epoch": 2.296796458849126,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.544961590600995e-05,
+      "loss": 0.6444,
+      "step": 7655
+    },
+    {
+      "epoch": 2.297096556380824,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.543455339659588e-05,
+      "loss": 0.4996,
+      "step": 7656
+    },
+    {
+      "epoch": 2.2973966539125215,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.54194908871818e-05,
+      "loss": 0.613,
+      "step": 7657
+    },
+    {
+      "epoch": 2.2976967514442195,
+      "grad_norm": 1.2734375,
+      "learning_rate": 8.540442837776774e-05,
+      "loss": 0.6413,
+      "step": 7658
+    },
+    {
+      "epoch": 2.297996848975917,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.538936586835367e-05,
+      "loss": 0.5702,
+      "step": 7659
+    },
+    {
+      "epoch": 2.298296946507615,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.53743033589396e-05,
+      "loss": 0.5561,
+      "step": 7660
+    },
+    {
+      "epoch": 2.2985970440393126,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.535924084952554e-05,
+      "loss": 0.5315,
+      "step": 7661
+    },
+    {
+      "epoch": 2.2988971415710107,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.534417834011147e-05,
+      "loss": 0.6621,
+      "step": 7662
+    },
+    {
+      "epoch": 2.2991972391027082,
+      "grad_norm": 1.125,
+      "learning_rate": 8.53291158306974e-05,
+      "loss": 0.7064,
+      "step": 7663
+    },
+    {
+      "epoch": 2.2994973366344063,
+      "grad_norm": 1.0,
+      "learning_rate": 8.531405332128333e-05,
+      "loss": 0.5641,
+      "step": 7664
+    },
+    {
+      "epoch": 2.299797434166104,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.529899081186926e-05,
+      "loss": 0.569,
+      "step": 7665
+    },
+    {
+      "epoch": 2.300097531697802,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.52839283024552e-05,
+      "loss": 0.5988,
+      "step": 7666
+    },
+    {
+      "epoch": 2.3003976292294994,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.526886579304113e-05,
+      "loss": 0.6354,
+      "step": 7667
+    },
+    {
+      "epoch": 2.3006977267611974,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.525380328362705e-05,
+      "loss": 0.5539,
+      "step": 7668
+    },
+    {
+      "epoch": 2.300997824292895,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.523874077421299e-05,
+      "loss": 0.5477,
+      "step": 7669
+    },
+    {
+      "epoch": 2.301297921824593,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.522367826479892e-05,
+      "loss": 0.6653,
+      "step": 7670
+    },
+    {
+      "epoch": 2.3015980193562906,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.520861575538485e-05,
+      "loss": 0.5969,
+      "step": 7671
+    },
+    {
+      "epoch": 2.3018981168879886,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.519355324597079e-05,
+      "loss": 0.5269,
+      "step": 7672
+    },
+    {
+      "epoch": 2.302198214419686,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.517849073655671e-05,
+      "loss": 0.6335,
+      "step": 7673
+    },
+    {
+      "epoch": 2.302498311951384,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.516342822714265e-05,
+      "loss": 0.5086,
+      "step": 7674
+    },
+    {
+      "epoch": 2.3027984094830822,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.514836571772857e-05,
+      "loss": 0.6755,
+      "step": 7675
+    },
+    {
+      "epoch": 2.30309850701478,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.513330320831451e-05,
+      "loss": 0.5536,
+      "step": 7676
+    },
+    {
+      "epoch": 2.3033986045464774,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.511824069890045e-05,
+      "loss": 0.6201,
+      "step": 7677
+    },
+    {
+      "epoch": 2.3036987020781754,
+      "grad_norm": 0.92578125,
+      "learning_rate": 8.510317818948637e-05,
+      "loss": 0.4876,
+      "step": 7678
+    },
+    {
+      "epoch": 2.3039987996098734,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.50881156800723e-05,
+      "loss": 0.5825,
+      "step": 7679
+    },
+    {
+      "epoch": 2.304298897141571,
+      "grad_norm": 1.0,
+      "learning_rate": 8.507305317065823e-05,
+      "loss": 0.5875,
+      "step": 7680
+    },
+    {
+      "epoch": 2.3045989946732686,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.505799066124417e-05,
+      "loss": 0.5945,
+      "step": 7681
+    },
+    {
+      "epoch": 2.3048990922049666,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.504292815183011e-05,
+      "loss": 0.5235,
+      "step": 7682
+    },
+    {
+      "epoch": 2.3051991897366646,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.502786564241602e-05,
+      "loss": 0.573,
+      "step": 7683
+    },
+    {
+      "epoch": 2.305499287268362,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.501280313300196e-05,
+      "loss": 0.5743,
+      "step": 7684
+    },
+    {
+      "epoch": 2.30579938480006,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.49977406235879e-05,
+      "loss": 0.6054,
+      "step": 7685
+    },
+    {
+      "epoch": 2.3060994823317578,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.498267811417383e-05,
+      "loss": 0.5376,
+      "step": 7686
+    },
+    {
+      "epoch": 2.306399579863456,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.496761560475976e-05,
+      "loss": 0.6194,
+      "step": 7687
+    },
+    {
+      "epoch": 2.3066996773951534,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.495255309534568e-05,
+      "loss": 0.5316,
+      "step": 7688
+    },
+    {
+      "epoch": 2.3069997749268514,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.493749058593162e-05,
+      "loss": 0.5909,
+      "step": 7689
+    },
+    {
+      "epoch": 2.307299872458549,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.492242807651756e-05,
+      "loss": 0.5304,
+      "step": 7690
+    },
+    {
+      "epoch": 2.307599969990247,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.490736556710348e-05,
+      "loss": 0.5183,
+      "step": 7691
+    },
+    {
+      "epoch": 2.3079000675219445,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.489230305768942e-05,
+      "loss": 0.5533,
+      "step": 7692
+    },
+    {
+      "epoch": 2.3082001650536426,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.487724054827534e-05,
+      "loss": 0.5457,
+      "step": 7693
+    },
+    {
+      "epoch": 2.30850026258534,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.486217803886127e-05,
+      "loss": 0.6351,
+      "step": 7694
+    },
+    {
+      "epoch": 2.308800360117038,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.484711552944721e-05,
+      "loss": 0.5887,
+      "step": 7695
+    },
+    {
+      "epoch": 2.3091004576487357,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.483205302003314e-05,
+      "loss": 0.6089,
+      "step": 7696
+    },
+    {
+      "epoch": 2.3094005551804337,
+      "grad_norm": 1.2578125,
+      "learning_rate": 8.481699051061908e-05,
+      "loss": 0.5987,
+      "step": 7697
+    },
+    {
+      "epoch": 2.3097006527121313,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.4801928001205e-05,
+      "loss": 0.5387,
+      "step": 7698
+    },
+    {
+      "epoch": 2.3100007502438293,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.478686549179093e-05,
+      "loss": 0.4946,
+      "step": 7699
+    },
+    {
+      "epoch": 2.310300847775527,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.477180298237687e-05,
+      "loss": 0.6088,
+      "step": 7700
+    },
+    {
+      "epoch": 2.310600945307225,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.47567404729628e-05,
+      "loss": 0.6098,
+      "step": 7701
+    },
+    {
+      "epoch": 2.3109010428389225,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.474167796354873e-05,
+      "loss": 0.5262,
+      "step": 7702
+    },
+    {
+      "epoch": 2.3112011403706205,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.472661545413467e-05,
+      "loss": 0.5336,
+      "step": 7703
+    },
+    {
+      "epoch": 2.311501237902318,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.471155294472059e-05,
+      "loss": 0.573,
+      "step": 7704
+    },
+    {
+      "epoch": 2.311801335434016,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.469649043530653e-05,
+      "loss": 0.5695,
+      "step": 7705
+    },
+    {
+      "epoch": 2.3121014329657137,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.468142792589246e-05,
+      "loss": 0.5508,
+      "step": 7706
+    },
+    {
+      "epoch": 2.3124015304974117,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.466636541647839e-05,
+      "loss": 0.5555,
+      "step": 7707
+    },
+    {
+      "epoch": 2.3127016280291093,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.465130290706433e-05,
+      "loss": 0.5642,
+      "step": 7708
+    },
+    {
+      "epoch": 2.3130017255608073,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.463624039765025e-05,
+      "loss": 0.5386,
+      "step": 7709
+    },
+    {
+      "epoch": 2.313301823092505,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.462117788823618e-05,
+      "loss": 0.6036,
+      "step": 7710
+    },
+    {
+      "epoch": 2.313601920624203,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.460611537882212e-05,
+      "loss": 0.5083,
+      "step": 7711
+    },
+    {
+      "epoch": 2.3139020181559005,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.459105286940805e-05,
+      "loss": 0.5873,
+      "step": 7712
+    },
+    {
+      "epoch": 2.3142021156875985,
+      "grad_norm": 1.0,
+      "learning_rate": 8.457599035999399e-05,
+      "loss": 0.4915,
+      "step": 7713
+    },
+    {
+      "epoch": 2.3145022132192965,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.456092785057991e-05,
+      "loss": 0.4667,
+      "step": 7714
+    },
+    {
+      "epoch": 2.314802310750994,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.454586534116584e-05,
+      "loss": 0.652,
+      "step": 7715
+    },
+    {
+      "epoch": 2.3151024082826916,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.453080283175178e-05,
+      "loss": 0.6089,
+      "step": 7716
+    },
+    {
+      "epoch": 2.3154025058143897,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.451574032233771e-05,
+      "loss": 0.6214,
+      "step": 7717
+    },
+    {
+      "epoch": 2.3157026033460877,
+      "grad_norm": 1.0,
+      "learning_rate": 8.450067781292363e-05,
+      "loss": 0.5766,
+      "step": 7718
+    },
+    {
+      "epoch": 2.3160027008777853,
+      "grad_norm": 1.21875,
+      "learning_rate": 8.448561530350957e-05,
+      "loss": 0.6604,
+      "step": 7719
+    },
+    {
+      "epoch": 2.316302798409483,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.44705527940955e-05,
+      "loss": 0.586,
+      "step": 7720
+    },
+    {
+      "epoch": 2.316602895941181,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.445549028468143e-05,
+      "loss": 0.6384,
+      "step": 7721
+    },
+    {
+      "epoch": 2.316902993472879,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.444042777526736e-05,
+      "loss": 0.6896,
+      "step": 7722
+    },
+    {
+      "epoch": 2.3172030910045764,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.44253652658533e-05,
+      "loss": 0.6339,
+      "step": 7723
+    },
+    {
+      "epoch": 2.3175031885362745,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.441030275643924e-05,
+      "loss": 0.6097,
+      "step": 7724
+    },
+    {
+      "epoch": 2.317803286067972,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.439524024702515e-05,
+      "loss": 0.5756,
+      "step": 7725
+    },
+    {
+      "epoch": 2.31810338359967,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.438017773761109e-05,
+      "loss": 0.6595,
+      "step": 7726
+    },
+    {
+      "epoch": 2.3184034811313676,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.436511522819702e-05,
+      "loss": 0.6282,
+      "step": 7727
+    },
+    {
+      "epoch": 2.3187035786630656,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.435005271878296e-05,
+      "loss": 0.5418,
+      "step": 7728
+    },
+    {
+      "epoch": 2.319003676194763,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.433499020936888e-05,
+      "loss": 0.4325,
+      "step": 7729
+    },
+    {
+      "epoch": 2.3193037737264612,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.431992769995481e-05,
+      "loss": 0.6363,
+      "step": 7730
+    },
+    {
+      "epoch": 2.319603871258159,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.430486519054075e-05,
+      "loss": 0.6492,
+      "step": 7731
+    },
+    {
+      "epoch": 2.319903968789857,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.428980268112669e-05,
+      "loss": 0.5141,
+      "step": 7732
+    },
+    {
+      "epoch": 2.3202040663215544,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.42747401717126e-05,
+      "loss": 0.587,
+      "step": 7733
+    },
+    {
+      "epoch": 2.3205041638532524,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.425967766229854e-05,
+      "loss": 0.6826,
+      "step": 7734
+    },
+    {
+      "epoch": 2.32080426138495,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.424461515288447e-05,
+      "loss": 0.5896,
+      "step": 7735
+    },
+    {
+      "epoch": 2.321104358916648,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.422955264347041e-05,
+      "loss": 0.5619,
+      "step": 7736
+    },
+    {
+      "epoch": 2.3214044564483456,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.421449013405634e-05,
+      "loss": 0.5802,
+      "step": 7737
+    },
+    {
+      "epoch": 2.3217045539800436,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.419942762464226e-05,
+      "loss": 0.561,
+      "step": 7738
+    },
+    {
+      "epoch": 2.322004651511741,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.41843651152282e-05,
+      "loss": 0.4946,
+      "step": 7739
+    },
+    {
+      "epoch": 2.322304749043439,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.416930260581413e-05,
+      "loss": 0.5698,
+      "step": 7740
+    },
+    {
+      "epoch": 2.3226048465751368,
+      "grad_norm": 1.28125,
+      "learning_rate": 8.415424009640006e-05,
+      "loss": 0.6111,
+      "step": 7741
+    },
+    {
+      "epoch": 2.322904944106835,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.4139177586986e-05,
+      "loss": 0.5956,
+      "step": 7742
+    },
+    {
+      "epoch": 2.3232050416385324,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.412411507757193e-05,
+      "loss": 0.5469,
+      "step": 7743
+    },
+    {
+      "epoch": 2.3235051391702304,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.410905256815787e-05,
+      "loss": 0.4533,
+      "step": 7744
+    },
+    {
+      "epoch": 2.323805236701928,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.409399005874379e-05,
+      "loss": 0.5676,
+      "step": 7745
+    },
+    {
+      "epoch": 2.324105334233626,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.407892754932972e-05,
+      "loss": 0.6066,
+      "step": 7746
+    },
+    {
+      "epoch": 2.3244054317653235,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.406386503991566e-05,
+      "loss": 0.5857,
+      "step": 7747
+    },
+    {
+      "epoch": 2.3247055292970216,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.404880253050159e-05,
+      "loss": 0.5675,
+      "step": 7748
+    },
+    {
+      "epoch": 2.325005626828719,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.403374002108751e-05,
+      "loss": 0.5958,
+      "step": 7749
+    },
+    {
+      "epoch": 2.325305724360417,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.401867751167345e-05,
+      "loss": 0.6059,
+      "step": 7750
+    },
+    {
+      "epoch": 2.3256058218921147,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.400361500225938e-05,
+      "loss": 0.6954,
+      "step": 7751
+    },
+    {
+      "epoch": 2.3259059194238128,
+      "grad_norm": 1.125,
+      "learning_rate": 8.39885524928453e-05,
+      "loss": 0.5943,
+      "step": 7752
+    },
+    {
+      "epoch": 2.3262060169555108,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.397348998343125e-05,
+      "loss": 0.5283,
+      "step": 7753
+    },
+    {
+      "epoch": 2.3265061144872083,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.395842747401717e-05,
+      "loss": 0.6109,
+      "step": 7754
+    },
+    {
+      "epoch": 2.326806212018906,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.394336496460311e-05,
+      "loss": 0.5839,
+      "step": 7755
+    },
+    {
+      "epoch": 2.327106309550604,
+      "grad_norm": 0.875,
+      "learning_rate": 8.392830245518903e-05,
+      "loss": 0.574,
+      "step": 7756
+    },
+    {
+      "epoch": 2.327406407082302,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.391323994577497e-05,
+      "loss": 0.6609,
+      "step": 7757
+    },
+    {
+      "epoch": 2.3277065046139995,
+      "grad_norm": 1.25,
+      "learning_rate": 8.389817743636091e-05,
+      "loss": 0.5788,
+      "step": 7758
+    },
+    {
+      "epoch": 2.3280066021456975,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.388311492694683e-05,
+      "loss": 0.6324,
+      "step": 7759
+    },
+    {
+      "epoch": 2.328306699677395,
+      "grad_norm": 1.125,
+      "learning_rate": 8.386805241753276e-05,
+      "loss": 0.5921,
+      "step": 7760
+    },
+    {
+      "epoch": 2.328606797209093,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.38529899081187e-05,
+      "loss": 0.555,
+      "step": 7761
+    },
+    {
+      "epoch": 2.3289068947407907,
+      "grad_norm": 1.296875,
+      "learning_rate": 8.383792739870463e-05,
+      "loss": 0.5733,
+      "step": 7762
+    },
+    {
+      "epoch": 2.3292069922724887,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.382286488929057e-05,
+      "loss": 0.5836,
+      "step": 7763
+    },
+    {
+      "epoch": 2.3295070898041863,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.380780237987648e-05,
+      "loss": 0.6428,
+      "step": 7764
+    },
+    {
+      "epoch": 2.3298071873358843,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.379273987046242e-05,
+      "loss": 0.5124,
+      "step": 7765
+    },
+    {
+      "epoch": 2.330107284867582,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.377767736104836e-05,
+      "loss": 0.5957,
+      "step": 7766
+    },
+    {
+      "epoch": 2.33040738239928,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.376261485163429e-05,
+      "loss": 0.6035,
+      "step": 7767
+    },
+    {
+      "epoch": 2.3307074799309775,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.374755234222022e-05,
+      "loss": 0.7172,
+      "step": 7768
+    },
+    {
+      "epoch": 2.3310075774626755,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.373248983280614e-05,
+      "loss": 0.4969,
+      "step": 7769
+    },
+    {
+      "epoch": 2.331307674994373,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.371742732339208e-05,
+      "loss": 0.5423,
+      "step": 7770
+    },
+    {
+      "epoch": 2.331607772526071,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.370236481397802e-05,
+      "loss": 0.5133,
+      "step": 7771
+    },
+    {
+      "epoch": 2.3319078700577687,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.368730230456394e-05,
+      "loss": 0.5825,
+      "step": 7772
+    },
+    {
+      "epoch": 2.3322079675894667,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.367223979514988e-05,
+      "loss": 0.663,
+      "step": 7773
+    },
+    {
+      "epoch": 2.3325080651211643,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.36571772857358e-05,
+      "loss": 0.5567,
+      "step": 7774
+    },
+    {
+      "epoch": 2.3328081626528623,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.364211477632173e-05,
+      "loss": 0.6029,
+      "step": 7775
+    },
+    {
+      "epoch": 2.33310826018456,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.362705226690767e-05,
+      "loss": 0.6084,
+      "step": 7776
+    },
+    {
+      "epoch": 2.333408357716258,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.36119897574936e-05,
+      "loss": 0.6263,
+      "step": 7777
+    },
+    {
+      "epoch": 2.3337084552479554,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.359692724807954e-05,
+      "loss": 0.625,
+      "step": 7778
+    },
+    {
+      "epoch": 2.3340085527796535,
+      "grad_norm": 1.40625,
+      "learning_rate": 8.358186473866546e-05,
+      "loss": 0.5791,
+      "step": 7779
+    },
+    {
+      "epoch": 2.334308650311351,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.356680222925139e-05,
+      "loss": 0.5199,
+      "step": 7780
+    },
+    {
+      "epoch": 2.334608747843049,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.355173971983733e-05,
+      "loss": 0.5254,
+      "step": 7781
+    },
+    {
+      "epoch": 2.3349088453747466,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.353667721042326e-05,
+      "loss": 0.6169,
+      "step": 7782
+    },
+    {
+      "epoch": 2.3352089429064447,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.352161470100919e-05,
+      "loss": 0.5114,
+      "step": 7783
+    },
+    {
+      "epoch": 2.3355090404381422,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.350655219159513e-05,
+      "loss": 0.5248,
+      "step": 7784
+    },
+    {
+      "epoch": 2.3358091379698402,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.349148968218105e-05,
+      "loss": 0.5941,
+      "step": 7785
+    },
+    {
+      "epoch": 2.336109235501538,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.347642717276699e-05,
+      "loss": 0.6284,
+      "step": 7786
+    },
+    {
+      "epoch": 2.336409333033236,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.346136466335292e-05,
+      "loss": 0.5893,
+      "step": 7787
+    },
+    {
+      "epoch": 2.3367094305649334,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.344630215393885e-05,
+      "loss": 0.5422,
+      "step": 7788
+    },
+    {
+      "epoch": 2.3370095280966314,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.343123964452479e-05,
+      "loss": 0.6254,
+      "step": 7789
+    },
+    {
+      "epoch": 2.337309625628329,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.341617713511071e-05,
+      "loss": 0.5665,
+      "step": 7790
+    },
+    {
+      "epoch": 2.337609723160027,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.340111462569664e-05,
+      "loss": 0.5888,
+      "step": 7791
+    },
+    {
+      "epoch": 2.337909820691725,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.338605211628258e-05,
+      "loss": 0.5584,
+      "step": 7792
+    },
+    {
+      "epoch": 2.3382099182234226,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.337098960686851e-05,
+      "loss": 0.56,
+      "step": 7793
+    },
+    {
+      "epoch": 2.33851001575512,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.335592709745445e-05,
+      "loss": 0.586,
+      "step": 7794
+    },
+    {
+      "epoch": 2.338810113286818,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.334086458804037e-05,
+      "loss": 0.6314,
+      "step": 7795
+    },
+    {
+      "epoch": 2.3391102108185162,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.33258020786263e-05,
+      "loss": 0.5499,
+      "step": 7796
+    },
+    {
+      "epoch": 2.339410308350214,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.331073956921224e-05,
+      "loss": 0.5523,
+      "step": 7797
+    },
+    {
+      "epoch": 2.339710405881912,
+      "grad_norm": 1.234375,
+      "learning_rate": 8.329567705979817e-05,
+      "loss": 0.6414,
+      "step": 7798
+    },
+    {
+      "epoch": 2.3400105034136094,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.32806145503841e-05,
+      "loss": 0.5831,
+      "step": 7799
+    },
+    {
+      "epoch": 2.3403106009453074,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.326555204097003e-05,
+      "loss": 0.5487,
+      "step": 7800
+    },
+    {
+      "epoch": 2.340610698477005,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.325048953155596e-05,
+      "loss": 0.479,
+      "step": 7801
+    },
+    {
+      "epoch": 2.340910796008703,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.323542702214189e-05,
+      "loss": 0.5739,
+      "step": 7802
+    },
+    {
+      "epoch": 2.3412108935404006,
+      "grad_norm": 1.265625,
+      "learning_rate": 8.322036451272783e-05,
+      "loss": 0.6743,
+      "step": 7803
+    },
+    {
+      "epoch": 2.3415109910720986,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.320530200331376e-05,
+      "loss": 0.6847,
+      "step": 7804
+    },
+    {
+      "epoch": 2.341811088603796,
+      "grad_norm": 1.3203125,
+      "learning_rate": 8.31902394938997e-05,
+      "loss": 0.5912,
+      "step": 7805
+    },
+    {
+      "epoch": 2.342111186135494,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.317517698448561e-05,
+      "loss": 0.5652,
+      "step": 7806
+    },
+    {
+      "epoch": 2.3424112836671918,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.316011447507155e-05,
+      "loss": 0.5224,
+      "step": 7807
+    },
+    {
+      "epoch": 2.3427113811988898,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.314505196565749e-05,
+      "loss": 0.5355,
+      "step": 7808
+    },
+    {
+      "epoch": 2.3430114787305873,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.312998945624342e-05,
+      "loss": 0.5455,
+      "step": 7809
+    },
+    {
+      "epoch": 2.3433115762622854,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.311492694682934e-05,
+      "loss": 0.6367,
+      "step": 7810
+    },
+    {
+      "epoch": 2.343611673793983,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.309986443741527e-05,
+      "loss": 0.6504,
+      "step": 7811
+    },
+    {
+      "epoch": 2.343911771325681,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.308480192800121e-05,
+      "loss": 0.6317,
+      "step": 7812
+    },
+    {
+      "epoch": 2.3442118688573785,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.306973941858715e-05,
+      "loss": 0.5117,
+      "step": 7813
+    },
+    {
+      "epoch": 2.3445119663890766,
+      "grad_norm": 1.125,
+      "learning_rate": 8.305467690917306e-05,
+      "loss": 0.5404,
+      "step": 7814
+    },
+    {
+      "epoch": 2.344812063920774,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.3039614399759e-05,
+      "loss": 0.5369,
+      "step": 7815
+    },
+    {
+      "epoch": 2.345112161452472,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.302455189034493e-05,
+      "loss": 0.6153,
+      "step": 7816
+    },
+    {
+      "epoch": 2.3454122589841697,
+      "grad_norm": 1.0,
+      "learning_rate": 8.300948938093087e-05,
+      "loss": 0.5755,
+      "step": 7817
+    },
+    {
+      "epoch": 2.3457123565158677,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.29944268715168e-05,
+      "loss": 0.5564,
+      "step": 7818
+    },
+    {
+      "epoch": 2.3460124540475653,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.297936436210272e-05,
+      "loss": 0.5095,
+      "step": 7819
+    },
+    {
+      "epoch": 2.3463125515792633,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.296430185268866e-05,
+      "loss": 0.6102,
+      "step": 7820
+    },
+    {
+      "epoch": 2.346612649110961,
+      "grad_norm": 1.125,
+      "learning_rate": 8.29492393432746e-05,
+      "loss": 0.6045,
+      "step": 7821
+    },
+    {
+      "epoch": 2.346912746642659,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.293417683386052e-05,
+      "loss": 0.4143,
+      "step": 7822
+    },
+    {
+      "epoch": 2.3472128441743565,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.291911432444646e-05,
+      "loss": 0.5539,
+      "step": 7823
+    },
+    {
+      "epoch": 2.3475129417060545,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.290405181503239e-05,
+      "loss": 0.5956,
+      "step": 7824
+    },
+    {
+      "epoch": 2.347813039237752,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.288898930561833e-05,
+      "loss": 0.5147,
+      "step": 7825
+    },
+    {
+      "epoch": 2.34811313676945,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.287392679620425e-05,
+      "loss": 0.6455,
+      "step": 7826
+    },
+    {
+      "epoch": 2.348413234301148,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.285886428679018e-05,
+      "loss": 0.5599,
+      "step": 7827
+    },
+    {
+      "epoch": 2.3487133318328457,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.284380177737612e-05,
+      "loss": 0.6412,
+      "step": 7828
+    },
+    {
+      "epoch": 2.3490134293645433,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.282873926796205e-05,
+      "loss": 0.6887,
+      "step": 7829
+    },
+    {
+      "epoch": 2.3493135268962413,
+      "grad_norm": 1.25,
+      "learning_rate": 8.281367675854797e-05,
+      "loss": 0.5401,
+      "step": 7830
+    },
+    {
+      "epoch": 2.3496136244279393,
+      "grad_norm": 0.91796875,
+      "learning_rate": 8.279861424913391e-05,
+      "loss": 0.5172,
+      "step": 7831
+    },
+    {
+      "epoch": 2.349913721959637,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.278355173971984e-05,
+      "loss": 0.6708,
+      "step": 7832
+    },
+    {
+      "epoch": 2.3502138194913345,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.276848923030577e-05,
+      "loss": 0.524,
+      "step": 7833
+    },
+    {
+      "epoch": 2.3505139170230325,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.275342672089171e-05,
+      "loss": 0.5886,
+      "step": 7834
+    },
+    {
+      "epoch": 2.3508140145547305,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.273836421147763e-05,
+      "loss": 0.5556,
+      "step": 7835
+    },
+    {
+      "epoch": 2.351114112086428,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.272330170206357e-05,
+      "loss": 0.5457,
+      "step": 7836
+    },
+    {
+      "epoch": 2.351414209618126,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.27082391926495e-05,
+      "loss": 0.5788,
+      "step": 7837
+    },
+    {
+      "epoch": 2.3517143071498237,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.269317668323543e-05,
+      "loss": 0.5595,
+      "step": 7838
+    },
+    {
+      "epoch": 2.3520144046815217,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.267811417382137e-05,
+      "loss": 0.5333,
+      "step": 7839
+    },
+    {
+      "epoch": 2.3523145022132192,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.26630516644073e-05,
+      "loss": 0.5333,
+      "step": 7840
+    },
+    {
+      "epoch": 2.3526145997449173,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.264798915499322e-05,
+      "loss": 0.5696,
+      "step": 7841
+    },
+    {
+      "epoch": 2.352914697276615,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.263292664557916e-05,
+      "loss": 0.5717,
+      "step": 7842
+    },
+    {
+      "epoch": 2.353214794808313,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.261786413616509e-05,
+      "loss": 0.5013,
+      "step": 7843
+    },
+    {
+      "epoch": 2.3535148923400104,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.260280162675103e-05,
+      "loss": 0.5049,
+      "step": 7844
+    },
+    {
+      "epoch": 2.3538149898717085,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.258773911733694e-05,
+      "loss": 0.6387,
+      "step": 7845
+    },
+    {
+      "epoch": 2.354115087403406,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.257267660792288e-05,
+      "loss": 0.6605,
+      "step": 7846
+    },
+    {
+      "epoch": 2.354415184935104,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.255761409850882e-05,
+      "loss": 0.5309,
+      "step": 7847
+    },
+    {
+      "epoch": 2.3547152824668016,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.254255158909475e-05,
+      "loss": 0.4512,
+      "step": 7848
+    },
+    {
+      "epoch": 2.3550153799984996,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.252748907968068e-05,
+      "loss": 0.5272,
+      "step": 7849
+    },
+    {
+      "epoch": 2.355315477530197,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.251242657026662e-05,
+      "loss": 0.5874,
+      "step": 7850
+    },
+    {
+      "epoch": 2.3556155750618952,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.249736406085254e-05,
+      "loss": 0.5696,
+      "step": 7851
+    },
+    {
+      "epoch": 2.355915672593593,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.248230155143848e-05,
+      "loss": 0.598,
+      "step": 7852
+    },
+    {
+      "epoch": 2.356215770125291,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.24672390420244e-05,
+      "loss": 0.5436,
+      "step": 7853
+    },
+    {
+      "epoch": 2.3565158676569884,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.245217653261034e-05,
+      "loss": 0.5825,
+      "step": 7854
+    },
+    {
+      "epoch": 2.3568159651886864,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.243711402319628e-05,
+      "loss": 0.5174,
+      "step": 7855
+    },
+    {
+      "epoch": 2.357116062720384,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.242205151378219e-05,
+      "loss": 0.6044,
+      "step": 7856
+    },
+    {
+      "epoch": 2.357416160252082,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.240698900436813e-05,
+      "loss": 0.5529,
+      "step": 7857
+    },
+    {
+      "epoch": 2.3577162577837796,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.239192649495406e-05,
+      "loss": 0.5299,
+      "step": 7858
+    },
+    {
+      "epoch": 2.3580163553154776,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.237686398554e-05,
+      "loss": 0.6632,
+      "step": 7859
+    },
+    {
+      "epoch": 2.358316452847175,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.236180147612592e-05,
+      "loss": 0.6127,
+      "step": 7860
+    },
+    {
+      "epoch": 2.358616550378873,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.234673896671185e-05,
+      "loss": 0.5621,
+      "step": 7861
+    },
+    {
+      "epoch": 2.3589166479105708,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.233167645729779e-05,
+      "loss": 0.5651,
+      "step": 7862
+    },
+    {
+      "epoch": 2.359216745442269,
+      "grad_norm": 1.125,
+      "learning_rate": 8.231661394788372e-05,
+      "loss": 0.5114,
+      "step": 7863
+    },
+    {
+      "epoch": 2.3595168429739664,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.230155143846965e-05,
+      "loss": 0.5676,
+      "step": 7864
+    },
+    {
+      "epoch": 2.3598169405056644,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.228648892905559e-05,
+      "loss": 0.5449,
+      "step": 7865
+    },
+    {
+      "epoch": 2.3601170380373624,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.227142641964151e-05,
+      "loss": 0.5656,
+      "step": 7866
+    },
+    {
+      "epoch": 2.36041713556906,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.225636391022745e-05,
+      "loss": 0.6438,
+      "step": 7867
+    },
+    {
+      "epoch": 2.3607172331007575,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.224130140081338e-05,
+      "loss": 0.5645,
+      "step": 7868
+    },
+    {
+      "epoch": 2.3610173306324556,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.22262388913993e-05,
+      "loss": 0.6305,
+      "step": 7869
+    },
+    {
+      "epoch": 2.3613174281641536,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.221117638198525e-05,
+      "loss": 0.6059,
+      "step": 7870
+    },
+    {
+      "epoch": 2.361617525695851,
+      "grad_norm": 1.1640625,
+      "learning_rate": 8.219611387257117e-05,
+      "loss": 0.5295,
+      "step": 7871
+    },
+    {
+      "epoch": 2.3619176232275487,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.21810513631571e-05,
+      "loss": 0.5141,
+      "step": 7872
+    },
+    {
+      "epoch": 2.3622177207592467,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.216598885374304e-05,
+      "loss": 0.6101,
+      "step": 7873
+    },
+    {
+      "epoch": 2.3625178182909448,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.215092634432897e-05,
+      "loss": 0.6567,
+      "step": 7874
+    },
+    {
+      "epoch": 2.3628179158226423,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.213586383491491e-05,
+      "loss": 0.5546,
+      "step": 7875
+    },
+    {
+      "epoch": 2.3631180133543404,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.212080132550083e-05,
+      "loss": 0.5828,
+      "step": 7876
+    },
+    {
+      "epoch": 2.363418110886038,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.210573881608676e-05,
+      "loss": 0.5858,
+      "step": 7877
+    },
+    {
+      "epoch": 2.363718208417736,
+      "grad_norm": 1.125,
+      "learning_rate": 8.20906763066727e-05,
+      "loss": 0.6747,
+      "step": 7878
+    },
+    {
+      "epoch": 2.3640183059494335,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.207561379725863e-05,
+      "loss": 0.6834,
+      "step": 7879
+    },
+    {
+      "epoch": 2.3643184034811315,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.206055128784455e-05,
+      "loss": 0.5871,
+      "step": 7880
+    },
+    {
+      "epoch": 2.364618501012829,
+      "grad_norm": 0.95703125,
+      "learning_rate": 8.20454887784305e-05,
+      "loss": 0.5274,
+      "step": 7881
+    },
+    {
+      "epoch": 2.364918598544527,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.203042626901642e-05,
+      "loss": 0.5976,
+      "step": 7882
+    },
+    {
+      "epoch": 2.3652186960762247,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.201536375960235e-05,
+      "loss": 0.5515,
+      "step": 7883
+    },
+    {
+      "epoch": 2.3655187936079227,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.200030125018829e-05,
+      "loss": 0.6578,
+      "step": 7884
+    },
+    {
+      "epoch": 2.3658188911396203,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.198523874077422e-05,
+      "loss": 0.556,
+      "step": 7885
+    },
+    {
+      "epoch": 2.3661189886713183,
+      "grad_norm": 1.125,
+      "learning_rate": 8.197017623136016e-05,
+      "loss": 0.66,
+      "step": 7886
+    },
+    {
+      "epoch": 2.366419086203016,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.195511372194607e-05,
+      "loss": 0.52,
+      "step": 7887
+    },
+    {
+      "epoch": 2.366719183734714,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.194005121253201e-05,
+      "loss": 0.5342,
+      "step": 7888
+    },
+    {
+      "epoch": 2.3670192812664115,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.192498870311795e-05,
+      "loss": 0.5224,
+      "step": 7889
+    },
+    {
+      "epoch": 2.3673193787981095,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.190992619370388e-05,
+      "loss": 0.5373,
+      "step": 7890
+    },
+    {
+      "epoch": 2.367619476329807,
+      "grad_norm": 1.125,
+      "learning_rate": 8.18948636842898e-05,
+      "loss": 0.6543,
+      "step": 7891
+    },
+    {
+      "epoch": 2.367919573861505,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.187980117487573e-05,
+      "loss": 0.5381,
+      "step": 7892
+    },
+    {
+      "epoch": 2.3682196713932027,
+      "grad_norm": 0.9921875,
+      "learning_rate": 8.186473866546167e-05,
+      "loss": 0.578,
+      "step": 7893
+    },
+    {
+      "epoch": 2.3685197689249007,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.184967615604761e-05,
+      "loss": 0.5728,
+      "step": 7894
+    },
+    {
+      "epoch": 2.3688198664565983,
+      "grad_norm": 1.15625,
+      "learning_rate": 8.183461364663352e-05,
+      "loss": 0.579,
+      "step": 7895
+    },
+    {
+      "epoch": 2.3691199639882963,
+      "grad_norm": 1.1796875,
+      "learning_rate": 8.181955113721946e-05,
+      "loss": 0.643,
+      "step": 7896
+    },
+    {
+      "epoch": 2.369420061519994,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.18044886278054e-05,
+      "loss": 0.631,
+      "step": 7897
+    },
+    {
+      "epoch": 2.369720159051692,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.178942611839133e-05,
+      "loss": 0.58,
+      "step": 7898
+    },
+    {
+      "epoch": 2.3700202565833894,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.177436360897726e-05,
+      "loss": 0.4601,
+      "step": 7899
+    },
+    {
+      "epoch": 2.3703203541150875,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.175930109956318e-05,
+      "loss": 0.4807,
+      "step": 7900
+    },
+    {
+      "epoch": 2.370620451646785,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.174423859014912e-05,
+      "loss": 0.4873,
+      "step": 7901
+    },
+    {
+      "epoch": 2.370920549178483,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.172917608073507e-05,
+      "loss": 0.5208,
+      "step": 7902
+    },
+    {
+      "epoch": 2.3712206467101806,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.171411357132098e-05,
+      "loss": 0.531,
+      "step": 7903
+    },
+    {
+      "epoch": 2.3715207442418786,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.169905106190692e-05,
+      "loss": 0.5206,
+      "step": 7904
+    },
+    {
+      "epoch": 2.3718208417735767,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.168398855249285e-05,
+      "loss": 0.5201,
+      "step": 7905
+    },
+    {
+      "epoch": 2.3721209393052742,
+      "grad_norm": 1.1484375,
+      "learning_rate": 8.166892604307879e-05,
+      "loss": 0.6386,
+      "step": 7906
+    },
+    {
+      "epoch": 2.372421036836972,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.165386353366471e-05,
+      "loss": 0.5569,
+      "step": 7907
+    },
+    {
+      "epoch": 2.37272113436867,
+      "grad_norm": 1.2421875,
+      "learning_rate": 8.163880102425064e-05,
+      "loss": 0.5881,
+      "step": 7908
+    },
+    {
+      "epoch": 2.373021231900368,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.162373851483658e-05,
+      "loss": 0.4832,
+      "step": 7909
+    },
+    {
+      "epoch": 2.3733213294320654,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.16086760054225e-05,
+      "loss": 0.4804,
+      "step": 7910
+    },
+    {
+      "epoch": 2.373621426963763,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.159361349600843e-05,
+      "loss": 0.607,
+      "step": 7911
+    },
+    {
+      "epoch": 2.373921524495461,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.157855098659437e-05,
+      "loss": 0.5803,
+      "step": 7912
+    },
+    {
+      "epoch": 2.374221622027159,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.15634884771803e-05,
+      "loss": 0.5018,
+      "step": 7913
+    },
+    {
+      "epoch": 2.3745217195588566,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.154842596776623e-05,
+      "loss": 0.5452,
+      "step": 7914
+    },
+    {
+      "epoch": 2.3748218170905546,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.153336345835217e-05,
+      "loss": 0.5614,
+      "step": 7915
+    },
+    {
+      "epoch": 2.375121914622252,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.15183009489381e-05,
+      "loss": 0.6293,
+      "step": 7916
+    },
+    {
+      "epoch": 2.37542201215395,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.150323843952403e-05,
+      "loss": 0.6003,
+      "step": 7917
+    },
+    {
+      "epoch": 2.375722109685648,
+      "grad_norm": 1.2109375,
+      "learning_rate": 8.148817593010996e-05,
+      "loss": 0.6605,
+      "step": 7918
+    },
+    {
+      "epoch": 2.376022207217346,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.147311342069589e-05,
+      "loss": 0.6136,
+      "step": 7919
+    },
+    {
+      "epoch": 2.3763223047490434,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.145805091128183e-05,
+      "loss": 0.5237,
+      "step": 7920
+    },
+    {
+      "epoch": 2.3766224022807414,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.144298840186775e-05,
+      "loss": 0.5248,
+      "step": 7921
+    },
+    {
+      "epoch": 2.376922499812439,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.142792589245368e-05,
+      "loss": 0.508,
+      "step": 7922
+    },
+    {
+      "epoch": 2.377222597344137,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.141286338303962e-05,
+      "loss": 0.5535,
+      "step": 7923
+    },
+    {
+      "epoch": 2.3775226948758346,
+      "grad_norm": 0.875,
+      "learning_rate": 8.139780087362555e-05,
+      "loss": 0.5785,
+      "step": 7924
+    },
+    {
+      "epoch": 2.3778227924075326,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.138273836421149e-05,
+      "loss": 0.6022,
+      "step": 7925
+    },
+    {
+      "epoch": 2.37812288993923,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.136767585479742e-05,
+      "loss": 0.4888,
+      "step": 7926
+    },
+    {
+      "epoch": 2.378422987470928,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.135261334538334e-05,
+      "loss": 0.5438,
+      "step": 7927
+    },
+    {
+      "epoch": 2.3787230850026257,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.133755083596928e-05,
+      "loss": 0.4961,
+      "step": 7928
+    },
+    {
+      "epoch": 2.3790231825343238,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.132248832655521e-05,
+      "loss": 0.6301,
+      "step": 7929
+    },
+    {
+      "epoch": 2.3793232800660213,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.130742581714114e-05,
+      "loss": 0.5335,
+      "step": 7930
+    },
+    {
+      "epoch": 2.3796233775977194,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.129236330772708e-05,
+      "loss": 0.5414,
+      "step": 7931
+    },
+    {
+      "epoch": 2.379923475129417,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.1277300798313e-05,
+      "loss": 0.6003,
+      "step": 7932
+    },
+    {
+      "epoch": 2.380223572661115,
+      "grad_norm": 1.1875,
+      "learning_rate": 8.126223828889894e-05,
+      "loss": 0.501,
+      "step": 7933
+    },
+    {
+      "epoch": 2.3805236701928125,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.124717577948486e-05,
+      "loss": 0.6294,
+      "step": 7934
+    },
+    {
+      "epoch": 2.3808237677245105,
+      "grad_norm": 1.125,
+      "learning_rate": 8.12321132700708e-05,
+      "loss": 0.5745,
+      "step": 7935
+    },
+    {
+      "epoch": 2.381123865256208,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.121705076065674e-05,
+      "loss": 0.5222,
+      "step": 7936
+    },
+    {
+      "epoch": 2.381423962787906,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.120198825124265e-05,
+      "loss": 0.5327,
+      "step": 7937
+    },
+    {
+      "epoch": 2.3817240603196037,
+      "grad_norm": 1.359375,
+      "learning_rate": 8.118692574182859e-05,
+      "loss": 0.5253,
+      "step": 7938
+    },
+    {
+      "epoch": 2.3820241578513017,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.117186323241453e-05,
+      "loss": 0.6121,
+      "step": 7939
+    },
+    {
+      "epoch": 2.3823242553829993,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.115680072300046e-05,
+      "loss": 0.604,
+      "step": 7940
+    },
+    {
+      "epoch": 2.3826243529146973,
+      "grad_norm": 0.9765625,
+      "learning_rate": 8.114173821358638e-05,
+      "loss": 0.6423,
+      "step": 7941
+    },
+    {
+      "epoch": 2.382924450446395,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.112667570417231e-05,
+      "loss": 0.528,
+      "step": 7942
+    },
+    {
+      "epoch": 2.383224547978093,
+      "grad_norm": 1.2265625,
+      "learning_rate": 8.111161319475825e-05,
+      "loss": 0.6144,
+      "step": 7943
+    },
+    {
+      "epoch": 2.383524645509791,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.109655068534419e-05,
+      "loss": 0.4993,
+      "step": 7944
+    },
+    {
+      "epoch": 2.3838247430414885,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.10814881759301e-05,
+      "loss": 0.6245,
+      "step": 7945
+    },
+    {
+      "epoch": 2.384124840573186,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.106642566651605e-05,
+      "loss": 0.548,
+      "step": 7946
+    },
+    {
+      "epoch": 2.384424938104884,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.105136315710197e-05,
+      "loss": 0.578,
+      "step": 7947
+    },
+    {
+      "epoch": 2.384725035636582,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.103630064768791e-05,
+      "loss": 0.5932,
+      "step": 7948
+    },
+    {
+      "epoch": 2.3850251331682797,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.102123813827384e-05,
+      "loss": 0.535,
+      "step": 7949
+    },
+    {
+      "epoch": 2.3853252306999773,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.100617562885977e-05,
+      "loss": 0.6475,
+      "step": 7950
+    },
+    {
+      "epoch": 2.3856253282316753,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.09911131194457e-05,
+      "loss": 0.5352,
+      "step": 7951
+    },
+    {
+      "epoch": 2.3859254257633733,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.097605061003163e-05,
+      "loss": 0.4933,
+      "step": 7952
+    },
+    {
+      "epoch": 2.386225523295071,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.096098810061756e-05,
+      "loss": 0.6316,
+      "step": 7953
+    },
+    {
+      "epoch": 2.386525620826769,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.09459255912035e-05,
+      "loss": 0.5721,
+      "step": 7954
+    },
+    {
+      "epoch": 2.3868257183584665,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.093086308178943e-05,
+      "loss": 0.5295,
+      "step": 7955
+    },
+    {
+      "epoch": 2.3871258158901645,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.091580057237537e-05,
+      "loss": 0.5462,
+      "step": 7956
+    },
+    {
+      "epoch": 2.387425913421862,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.09007380629613e-05,
+      "loss": 0.5253,
+      "step": 7957
+    },
+    {
+      "epoch": 2.38772601095356,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.088567555354722e-05,
+      "loss": 0.5452,
+      "step": 7958
+    },
+    {
+      "epoch": 2.3880261084852576,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.087061304413316e-05,
+      "loss": 0.5538,
+      "step": 7959
+    },
+    {
+      "epoch": 2.3883262060169557,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.085555053471909e-05,
+      "loss": 0.5315,
+      "step": 7960
+    },
+    {
+      "epoch": 2.3886263035486532,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.084048802530501e-05,
+      "loss": 0.481,
+      "step": 7961
+    },
+    {
+      "epoch": 2.3889264010803513,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.082542551589095e-05,
+      "loss": 0.6675,
+      "step": 7962
+    },
+    {
+      "epoch": 2.389226498612049,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.081036300647688e-05,
+      "loss": 0.4172,
+      "step": 7963
+    },
+    {
+      "epoch": 2.389526596143747,
+      "grad_norm": 1.109375,
+      "learning_rate": 8.079530049706281e-05,
+      "loss": 0.6675,
+      "step": 7964
+    },
+    {
+      "epoch": 2.3898266936754444,
+      "grad_norm": 1.3125,
+      "learning_rate": 8.078023798764875e-05,
+      "loss": 0.5301,
+      "step": 7965
+    },
+    {
+      "epoch": 2.3901267912071424,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.076517547823468e-05,
+      "loss": 0.5374,
+      "step": 7966
+    },
+    {
+      "epoch": 2.39042688873884,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.075011296882062e-05,
+      "loss": 0.5408,
+      "step": 7967
+    },
+    {
+      "epoch": 2.390726986270538,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.073505045940654e-05,
+      "loss": 0.6073,
+      "step": 7968
+    },
+    {
+      "epoch": 2.3910270838022356,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.071998794999247e-05,
+      "loss": 0.5325,
+      "step": 7969
+    },
+    {
+      "epoch": 2.3913271813339336,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.070492544057841e-05,
+      "loss": 0.4953,
+      "step": 7970
+    },
+    {
+      "epoch": 2.391627278865631,
+      "grad_norm": 0.94921875,
+      "learning_rate": 8.068986293116434e-05,
+      "loss": 0.5233,
+      "step": 7971
+    },
+    {
+      "epoch": 2.391927376397329,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.067480042175026e-05,
+      "loss": 0.5249,
+      "step": 7972
+    },
+    {
+      "epoch": 2.392227473929027,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.06597379123362e-05,
+      "loss": 0.5946,
+      "step": 7973
+    },
+    {
+      "epoch": 2.392527571460725,
+      "grad_norm": 0.96484375,
+      "learning_rate": 8.064467540292213e-05,
+      "loss": 0.4571,
+      "step": 7974
+    },
+    {
+      "epoch": 2.3928276689924224,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.062961289350807e-05,
+      "loss": 0.6497,
+      "step": 7975
+    },
+    {
+      "epoch": 2.3931277665241204,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.061455038409398e-05,
+      "loss": 0.6891,
+      "step": 7976
+    },
+    {
+      "epoch": 2.393427864055818,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.059948787467992e-05,
+      "loss": 0.6269,
+      "step": 7977
+    },
+    {
+      "epoch": 2.393727961587516,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.058442536526586e-05,
+      "loss": 0.5434,
+      "step": 7978
+    },
+    {
+      "epoch": 2.3940280591192136,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.056936285585179e-05,
+      "loss": 0.6291,
+      "step": 7979
+    },
+    {
+      "epoch": 2.3943281566509116,
+      "grad_norm": 1.09375,
+      "learning_rate": 8.055430034643772e-05,
+      "loss": 0.5754,
+      "step": 7980
+    },
+    {
+      "epoch": 2.394628254182609,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.053923783702364e-05,
+      "loss": 0.5313,
+      "step": 7981
+    },
+    {
+      "epoch": 2.394928351714307,
+      "grad_norm": 0.9453125,
+      "learning_rate": 8.052417532760958e-05,
+      "loss": 0.5715,
+      "step": 7982
+    },
+    {
+      "epoch": 2.395228449246005,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.050911281819553e-05,
+      "loss": 0.5371,
+      "step": 7983
+    },
+    {
+      "epoch": 2.3955285467777028,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.049405030878144e-05,
+      "loss": 0.5539,
+      "step": 7984
+    },
+    {
+      "epoch": 2.3958286443094003,
+      "grad_norm": 1.0390625,
+      "learning_rate": 8.047898779936738e-05,
+      "loss": 0.5331,
+      "step": 7985
+    },
+    {
+      "epoch": 2.3961287418410984,
+      "grad_norm": 1.203125,
+      "learning_rate": 8.046392528995332e-05,
+      "loss": 0.574,
+      "step": 7986
+    },
+    {
+      "epoch": 2.3964288393727964,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.044886278053925e-05,
+      "loss": 0.6401,
+      "step": 7987
+    },
+    {
+      "epoch": 2.396728936904494,
+      "grad_norm": 0.953125,
+      "learning_rate": 8.043380027112517e-05,
+      "loss": 0.533,
+      "step": 7988
+    },
+    {
+      "epoch": 2.3970290344361915,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.04187377617111e-05,
+      "loss": 0.6057,
+      "step": 7989
+    },
+    {
+      "epoch": 2.3973291319678895,
+      "grad_norm": 1.0859375,
+      "learning_rate": 8.040367525229704e-05,
+      "loss": 0.5474,
+      "step": 7990
+    },
+    {
+      "epoch": 2.3976292294995876,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.038861274288297e-05,
+      "loss": 0.5519,
+      "step": 7991
+    },
+    {
+      "epoch": 2.397929327031285,
+      "grad_norm": 0.94140625,
+      "learning_rate": 8.037355023346889e-05,
+      "loss": 0.5416,
+      "step": 7992
+    },
+    {
+      "epoch": 2.398229424562983,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.035848772405483e-05,
+      "loss": 0.5996,
+      "step": 7993
+    },
+    {
+      "epoch": 2.3985295220946807,
+      "grad_norm": 0.984375,
+      "learning_rate": 8.034342521464076e-05,
+      "loss": 0.5909,
+      "step": 7994
+    },
+    {
+      "epoch": 2.3988296196263788,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.032836270522669e-05,
+      "loss": 0.5832,
+      "step": 7995
+    },
+    {
+      "epoch": 2.3991297171580763,
+      "grad_norm": 1.1015625,
+      "learning_rate": 8.031330019581263e-05,
+      "loss": 0.6031,
+      "step": 7996
+    },
+    {
+      "epoch": 2.3994298146897743,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.029823768639855e-05,
+      "loss": 0.5523,
+      "step": 7997
+    },
+    {
+      "epoch": 2.399729912221472,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.02831751769845e-05,
+      "loss": 0.6315,
+      "step": 7998
+    },
+    {
+      "epoch": 2.40003000975317,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.026811266757042e-05,
+      "loss": 0.5729,
+      "step": 7999
+    },
+    {
+      "epoch": 2.4003301072848675,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.025305015815635e-05,
+      "loss": 0.5276,
+      "step": 8000
+    },
+    {
+      "epoch": 2.4006302048165655,
+      "grad_norm": 1.1171875,
+      "learning_rate": 8.023798764874229e-05,
+      "loss": 0.5786,
+      "step": 8001
+    },
+    {
+      "epoch": 2.400930302348263,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.022292513932821e-05,
+      "loss": 0.5676,
+      "step": 8002
+    },
+    {
+      "epoch": 2.401230399879961,
+      "grad_norm": 1.0703125,
+      "learning_rate": 8.020786262991414e-05,
+      "loss": 0.5739,
+      "step": 8003
+    },
+    {
+      "epoch": 2.4015304974116587,
+      "grad_norm": 0.91015625,
+      "learning_rate": 8.019280012050008e-05,
+      "loss": 0.5184,
+      "step": 8004
+    },
+    {
+      "epoch": 2.4018305949433567,
+      "grad_norm": 1.046875,
+      "learning_rate": 8.017773761108601e-05,
+      "loss": 0.5754,
+      "step": 8005
+    },
+    {
+      "epoch": 2.4021306924750543,
+      "grad_norm": 0.97265625,
+      "learning_rate": 8.016267510167195e-05,
+      "loss": 0.5314,
+      "step": 8006
+    },
+    {
+      "epoch": 2.4024307900067523,
+      "grad_norm": 1.171875,
+      "learning_rate": 8.014761259225788e-05,
+      "loss": 0.6531,
+      "step": 8007
+    },
+    {
+      "epoch": 2.40273088753845,
+      "grad_norm": 1.0625,
+      "learning_rate": 8.01325500828438e-05,
+      "loss": 0.5142,
+      "step": 8008
+    },
+    {
+      "epoch": 2.403030985070148,
+      "grad_norm": 1.0234375,
+      "learning_rate": 8.011748757342974e-05,
+      "loss": 0.6462,
+      "step": 8009
+    },
+    {
+      "epoch": 2.4033310826018455,
+      "grad_norm": 1.34375,
+      "learning_rate": 8.010242506401567e-05,
+      "loss": 0.5129,
+      "step": 8010
+    },
+    {
+      "epoch": 2.4036311801335435,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.00873625546016e-05,
+      "loss": 0.5484,
+      "step": 8011
+    },
+    {
+      "epoch": 2.403931277665241,
+      "grad_norm": 1.0546875,
+      "learning_rate": 8.007230004518754e-05,
+      "loss": 0.5814,
+      "step": 8012
+    },
+    {
+      "epoch": 2.404231375196939,
+      "grad_norm": 1.25,
+      "learning_rate": 8.005723753577346e-05,
+      "loss": 0.6284,
+      "step": 8013
+    },
+    {
+      "epoch": 2.4045314727286367,
+      "grad_norm": 1.0078125,
+      "learning_rate": 8.00421750263594e-05,
+      "loss": 0.462,
+      "step": 8014
+    },
+    {
+      "epoch": 2.4048315702603347,
+      "grad_norm": 1.1328125,
+      "learning_rate": 8.002711251694533e-05,
+      "loss": 0.6153,
+      "step": 8015
+    },
+    {
+      "epoch": 2.4051316677920322,
+      "grad_norm": 1.1953125,
+      "learning_rate": 8.001205000753126e-05,
+      "loss": 0.6402,
+      "step": 8016
+    },
+    {
+      "epoch": 2.4054317653237303,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.99969874981172e-05,
+      "loss": 0.5926,
+      "step": 8017
+    },
+    {
+      "epoch": 2.405731862855428,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.998192498870311e-05,
+      "loss": 0.5405,
+      "step": 8018
+    },
+    {
+      "epoch": 2.406031960387126,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.996686247928905e-05,
+      "loss": 0.5536,
+      "step": 8019
+    },
+    {
+      "epoch": 2.4063320579188234,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.995179996987499e-05,
+      "loss": 0.5763,
+      "step": 8020
+    },
+    {
+      "epoch": 2.4066321554505214,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.993673746046092e-05,
+      "loss": 0.5441,
+      "step": 8021
+    },
+    {
+      "epoch": 2.4069322529822195,
+      "grad_norm": 1.125,
+      "learning_rate": 7.992167495104684e-05,
+      "loss": 0.574,
+      "step": 8022
+    },
+    {
+      "epoch": 2.407232350513917,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.990661244163277e-05,
+      "loss": 0.6589,
+      "step": 8023
+    },
+    {
+      "epoch": 2.4075324480456146,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.989154993221871e-05,
+      "loss": 0.6306,
+      "step": 8024
+    },
+    {
+      "epoch": 2.4078325455773126,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.987648742280465e-05,
+      "loss": 0.5266,
+      "step": 8025
+    },
+    {
+      "epoch": 2.4081326431090107,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.986142491339057e-05,
+      "loss": 0.5637,
+      "step": 8026
+    },
+    {
+      "epoch": 2.4084327406407082,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.98463624039765e-05,
+      "loss": 0.6447,
+      "step": 8027
+    },
+    {
+      "epoch": 2.408732838172406,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.983129989456245e-05,
+      "loss": 0.5499,
+      "step": 8028
+    },
+    {
+      "epoch": 2.409032935704104,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.981623738514837e-05,
+      "loss": 0.5156,
+      "step": 8029
+    },
+    {
+      "epoch": 2.409333033235802,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.98011748757343e-05,
+      "loss": 0.5537,
+      "step": 8030
+    },
+    {
+      "epoch": 2.4096331307674994,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.978611236632023e-05,
+      "loss": 0.677,
+      "step": 8031
+    },
+    {
+      "epoch": 2.4099332282991974,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.977104985690617e-05,
+      "loss": 0.6444,
+      "step": 8032
+    },
+    {
+      "epoch": 2.410233325830895,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.97559873474921e-05,
+      "loss": 0.5381,
+      "step": 8033
+    },
+    {
+      "epoch": 2.410533423362593,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.974092483807802e-05,
+      "loss": 0.6364,
+      "step": 8034
+    },
+    {
+      "epoch": 2.4108335208942906,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.972586232866396e-05,
+      "loss": 0.6262,
+      "step": 8035
+    },
+    {
+      "epoch": 2.4111336184259886,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.971079981924989e-05,
+      "loss": 0.6603,
+      "step": 8036
+    },
+    {
+      "epoch": 2.411433715957686,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.969573730983583e-05,
+      "loss": 0.525,
+      "step": 8037
+    },
+    {
+      "epoch": 2.411733813489384,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.968067480042175e-05,
+      "loss": 0.5772,
+      "step": 8038
+    },
+    {
+      "epoch": 2.4120339110210818,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.966561229100768e-05,
+      "loss": 0.4807,
+      "step": 8039
+    },
+    {
+      "epoch": 2.41233400855278,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.965054978159362e-05,
+      "loss": 0.5795,
+      "step": 8040
+    },
+    {
+      "epoch": 2.4126341060844774,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.963548727217955e-05,
+      "loss": 0.5437,
+      "step": 8041
+    },
+    {
+      "epoch": 2.4129342036161754,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.962042476276547e-05,
+      "loss": 0.6042,
+      "step": 8042
+    },
+    {
+      "epoch": 2.413234301147873,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.960536225335141e-05,
+      "loss": 0.5805,
+      "step": 8043
+    },
+    {
+      "epoch": 2.413534398679571,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.959029974393734e-05,
+      "loss": 0.5198,
+      "step": 8044
+    },
+    {
+      "epoch": 2.4138344962112686,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.957523723452327e-05,
+      "loss": 0.5092,
+      "step": 8045
+    },
+    {
+      "epoch": 2.4141345937429666,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.956017472510921e-05,
+      "loss": 0.6138,
+      "step": 8046
+    },
+    {
+      "epoch": 2.414434691274664,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.954511221569514e-05,
+      "loss": 0.4701,
+      "step": 8047
+    },
+    {
+      "epoch": 2.414734788806362,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.953004970628108e-05,
+      "loss": 0.5409,
+      "step": 8048
+    },
+    {
+      "epoch": 2.4150348863380597,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.9514987196867e-05,
+      "loss": 0.6083,
+      "step": 8049
+    },
+    {
+      "epoch": 2.4153349838697578,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.949992468745293e-05,
+      "loss": 0.4874,
+      "step": 8050
+    },
+    {
+      "epoch": 2.4156350814014553,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.948486217803887e-05,
+      "loss": 0.5641,
+      "step": 8051
+    },
+    {
+      "epoch": 2.4159351789331533,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.94697996686248e-05,
+      "loss": 0.5606,
+      "step": 8052
+    },
+    {
+      "epoch": 2.416235276464851,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.945473715921072e-05,
+      "loss": 0.6084,
+      "step": 8053
+    },
+    {
+      "epoch": 2.416535373996549,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.943967464979666e-05,
+      "loss": 0.5618,
+      "step": 8054
+    },
+    {
+      "epoch": 2.4168354715282465,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.942461214038259e-05,
+      "loss": 0.6037,
+      "step": 8055
+    },
+    {
+      "epoch": 2.4171355690599445,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.940954963096853e-05,
+      "loss": 0.5844,
+      "step": 8056
+    },
+    {
+      "epoch": 2.417435666591642,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.939448712155446e-05,
+      "loss": 0.5966,
+      "step": 8057
+    },
+    {
+      "epoch": 2.41773576412334,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.937942461214038e-05,
+      "loss": 0.4847,
+      "step": 8058
+    },
+    {
+      "epoch": 2.4180358616550377,
+      "grad_norm": 1.0,
+      "learning_rate": 7.936436210272632e-05,
+      "loss": 0.5956,
+      "step": 8059
+    },
+    {
+      "epoch": 2.4183359591867357,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.934929959331225e-05,
+      "loss": 0.6073,
+      "step": 8060
+    },
+    {
+      "epoch": 2.4186360567184337,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.933423708389818e-05,
+      "loss": 0.6669,
+      "step": 8061
+    },
+    {
+      "epoch": 2.4189361542501313,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.931917457448412e-05,
+      "loss": 0.5809,
+      "step": 8062
+    },
+    {
+      "epoch": 2.419236251781829,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.930411206507004e-05,
+      "loss": 0.488,
+      "step": 8063
+    },
+    {
+      "epoch": 2.419536349313527,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.928904955565599e-05,
+      "loss": 0.5247,
+      "step": 8064
+    },
+    {
+      "epoch": 2.419836446845225,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.92739870462419e-05,
+      "loss": 0.5419,
+      "step": 8065
+    },
+    {
+      "epoch": 2.4201365443769225,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.925892453682784e-05,
+      "loss": 0.5477,
+      "step": 8066
+    },
+    {
+      "epoch": 2.42043664190862,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.924386202741378e-05,
+      "loss": 0.5457,
+      "step": 8067
+    },
+    {
+      "epoch": 2.420736739440318,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.92287995179997e-05,
+      "loss": 0.5744,
+      "step": 8068
+    },
+    {
+      "epoch": 2.421036836972016,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.921373700858563e-05,
+      "loss": 0.5937,
+      "step": 8069
+    },
+    {
+      "epoch": 2.4213369345037137,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.919867449917156e-05,
+      "loss": 0.5734,
+      "step": 8070
+    },
+    {
+      "epoch": 2.4216370320354117,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.91836119897575e-05,
+      "loss": 0.4997,
+      "step": 8071
+    },
+    {
+      "epoch": 2.4219371295671093,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.916854948034343e-05,
+      "loss": 0.5145,
+      "step": 8072
+    },
+    {
+      "epoch": 2.4222372270988073,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.915348697092935e-05,
+      "loss": 0.6672,
+      "step": 8073
+    },
+    {
+      "epoch": 2.422537324630505,
+      "grad_norm": 1.125,
+      "learning_rate": 7.91384244615153e-05,
+      "loss": 0.6225,
+      "step": 8074
+    },
+    {
+      "epoch": 2.422837422162203,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.912336195210123e-05,
+      "loss": 0.6415,
+      "step": 8075
+    },
+    {
+      "epoch": 2.4231375196939005,
+      "grad_norm": 1.125,
+      "learning_rate": 7.910829944268715e-05,
+      "loss": 0.5807,
+      "step": 8076
+    },
+    {
+      "epoch": 2.4234376172255985,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.909323693327309e-05,
+      "loss": 0.5386,
+      "step": 8077
+    },
+    {
+      "epoch": 2.423737714757296,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.907817442385901e-05,
+      "loss": 0.55,
+      "step": 8078
+    },
+    {
+      "epoch": 2.424037812288994,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.906311191444495e-05,
+      "loss": 0.5759,
+      "step": 8079
+    },
+    {
+      "epoch": 2.4243379098206916,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.904804940503088e-05,
+      "loss": 0.5984,
+      "step": 8080
+    },
+    {
+      "epoch": 2.4246380073523897,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.903298689561681e-05,
+      "loss": 0.5283,
+      "step": 8081
+    },
+    {
+      "epoch": 2.4249381048840872,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.901792438620275e-05,
+      "loss": 0.5679,
+      "step": 8082
+    },
+    {
+      "epoch": 2.4252382024157852,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.900286187678867e-05,
+      "loss": 0.6704,
+      "step": 8083
+    },
+    {
+      "epoch": 2.425538299947483,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.89877993673746e-05,
+      "loss": 0.5621,
+      "step": 8084
+    },
+    {
+      "epoch": 2.425838397479181,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.897273685796054e-05,
+      "loss": 0.5282,
+      "step": 8085
+    },
+    {
+      "epoch": 2.4261384950108784,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.895767434854647e-05,
+      "loss": 0.6798,
+      "step": 8086
+    },
+    {
+      "epoch": 2.4264385925425764,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.894261183913241e-05,
+      "loss": 0.4552,
+      "step": 8087
+    },
+    {
+      "epoch": 2.426738690074274,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.892754932971834e-05,
+      "loss": 0.5276,
+      "step": 8088
+    },
+    {
+      "epoch": 2.427038787605972,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.891248682030426e-05,
+      "loss": 0.5413,
+      "step": 8089
+    },
+    {
+      "epoch": 2.4273388851376696,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.88974243108902e-05,
+      "loss": 0.6034,
+      "step": 8090
+    },
+    {
+      "epoch": 2.4276389826693676,
+      "grad_norm": 1.125,
+      "learning_rate": 7.888236180147613e-05,
+      "loss": 0.595,
+      "step": 8091
+    },
+    {
+      "epoch": 2.427939080201065,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.886729929206206e-05,
+      "loss": 0.4738,
+      "step": 8092
+    },
+    {
+      "epoch": 2.428239177732763,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.8852236782648e-05,
+      "loss": 0.5457,
+      "step": 8093
+    },
+    {
+      "epoch": 2.428539275264461,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.883717427323392e-05,
+      "loss": 0.564,
+      "step": 8094
+    },
+    {
+      "epoch": 2.428839372796159,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.882211176381986e-05,
+      "loss": 0.6599,
+      "step": 8095
+    },
+    {
+      "epoch": 2.4291394703278564,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.880704925440579e-05,
+      "loss": 0.5277,
+      "step": 8096
+    },
+    {
+      "epoch": 2.4294395678595544,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.879198674499172e-05,
+      "loss": 0.5334,
+      "step": 8097
+    },
+    {
+      "epoch": 2.429739665391252,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.877692423557766e-05,
+      "loss": 0.6059,
+      "step": 8098
+    },
+    {
+      "epoch": 2.43003976292295,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.876186172616357e-05,
+      "loss": 0.5412,
+      "step": 8099
+    },
+    {
+      "epoch": 2.430339860454648,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.874679921674951e-05,
+      "loss": 0.5705,
+      "step": 8100
+    },
+    {
+      "epoch": 2.4306399579863456,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.873173670733545e-05,
+      "loss": 0.5497,
+      "step": 8101
+    },
+    {
+      "epoch": 2.430940055518043,
+      "grad_norm": 1.3203125,
+      "learning_rate": 7.871667419792138e-05,
+      "loss": 0.8076,
+      "step": 8102
+    },
+    {
+      "epoch": 2.431240153049741,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.87016116885073e-05,
+      "loss": 0.6678,
+      "step": 8103
+    },
+    {
+      "epoch": 2.431540250581439,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.868654917909324e-05,
+      "loss": 0.5322,
+      "step": 8104
+    },
+    {
+      "epoch": 2.4318403481131368,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.867148666967917e-05,
+      "loss": 0.521,
+      "step": 8105
+    },
+    {
+      "epoch": 2.4321404456448343,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.865642416026511e-05,
+      "loss": 0.5579,
+      "step": 8106
+    },
+    {
+      "epoch": 2.4324405431765324,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.864136165085103e-05,
+      "loss": 0.5258,
+      "step": 8107
+    },
+    {
+      "epoch": 2.4327406407082304,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.862629914143697e-05,
+      "loss": 0.4753,
+      "step": 8108
+    },
+    {
+      "epoch": 2.433040738239928,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.86112366320229e-05,
+      "loss": 0.669,
+      "step": 8109
+    },
+    {
+      "epoch": 2.433340835771626,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.859617412260883e-05,
+      "loss": 0.6077,
+      "step": 8110
+    },
+    {
+      "epoch": 2.4336409333033235,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.858111161319476e-05,
+      "loss": 0.5956,
+      "step": 8111
+    },
+    {
+      "epoch": 2.4339410308350216,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.856604910378069e-05,
+      "loss": 0.5037,
+      "step": 8112
+    },
+    {
+      "epoch": 2.434241128366719,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.855098659436663e-05,
+      "loss": 0.6168,
+      "step": 8113
+    },
+    {
+      "epoch": 2.434541225898417,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.853592408495257e-05,
+      "loss": 0.6392,
+      "step": 8114
+    },
+    {
+      "epoch": 2.4348413234301147,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.852086157553848e-05,
+      "loss": 0.5256,
+      "step": 8115
+    },
+    {
+      "epoch": 2.4351414209618127,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.850579906612442e-05,
+      "loss": 0.5836,
+      "step": 8116
+    },
+    {
+      "epoch": 2.4354415184935103,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.849073655671035e-05,
+      "loss": 0.446,
+      "step": 8117
+    },
+    {
+      "epoch": 2.4357416160252083,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.847567404729629e-05,
+      "loss": 0.5627,
+      "step": 8118
+    },
+    {
+      "epoch": 2.436041713556906,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.846061153788221e-05,
+      "loss": 0.5912,
+      "step": 8119
+    },
+    {
+      "epoch": 2.436341811088604,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.844554902846814e-05,
+      "loss": 0.5542,
+      "step": 8120
+    },
+    {
+      "epoch": 2.4366419086203015,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.843048651905408e-05,
+      "loss": 0.6496,
+      "step": 8121
+    },
+    {
+      "epoch": 2.4369420061519995,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.841542400964002e-05,
+      "loss": 0.6122,
+      "step": 8122
+    },
+    {
+      "epoch": 2.437242103683697,
+      "grad_norm": 1.0,
+      "learning_rate": 7.840036150022593e-05,
+      "loss": 0.5584,
+      "step": 8123
+    },
+    {
+      "epoch": 2.437542201215395,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.838529899081187e-05,
+      "loss": 0.6185,
+      "step": 8124
+    },
+    {
+      "epoch": 2.4378422987470927,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.83702364813978e-05,
+      "loss": 0.5868,
+      "step": 8125
+    },
+    {
+      "epoch": 2.4381423962787907,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.835517397198373e-05,
+      "loss": 0.6229,
+      "step": 8126
+    },
+    {
+      "epoch": 2.4384424938104883,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.834011146256967e-05,
+      "loss": 0.6379,
+      "step": 8127
+    },
+    {
+      "epoch": 2.4387425913421863,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.83250489531556e-05,
+      "loss": 0.4598,
+      "step": 8128
+    },
+    {
+      "epoch": 2.439042688873884,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.830998644374154e-05,
+      "loss": 0.6193,
+      "step": 8129
+    },
+    {
+      "epoch": 2.439342786405582,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.829492393432746e-05,
+      "loss": 0.5775,
+      "step": 8130
+    },
+    {
+      "epoch": 2.4396428839372795,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.827986142491339e-05,
+      "loss": 0.5839,
+      "step": 8131
+    },
+    {
+      "epoch": 2.4399429814689775,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.826479891549933e-05,
+      "loss": 0.6188,
+      "step": 8132
+    },
+    {
+      "epoch": 2.440243079000675,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.824973640608526e-05,
+      "loss": 0.4742,
+      "step": 8133
+    },
+    {
+      "epoch": 2.440543176532373,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.823467389667118e-05,
+      "loss": 0.5344,
+      "step": 8134
+    },
+    {
+      "epoch": 2.4408432740640706,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.821961138725712e-05,
+      "loss": 0.5752,
+      "step": 8135
+    },
+    {
+      "epoch": 2.4411433715957687,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.820454887784305e-05,
+      "loss": 0.5706,
+      "step": 8136
+    },
+    {
+      "epoch": 2.4414434691274662,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.818948636842899e-05,
+      "loss": 0.5855,
+      "step": 8137
+    },
+    {
+      "epoch": 2.4417435666591643,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.817442385901492e-05,
+      "loss": 0.5441,
+      "step": 8138
+    },
+    {
+      "epoch": 2.4420436641908623,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.815936134960084e-05,
+      "loss": 0.5552,
+      "step": 8139
+    },
+    {
+      "epoch": 2.44234376172256,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.814429884018678e-05,
+      "loss": 0.5261,
+      "step": 8140
+    },
+    {
+      "epoch": 2.4426438592542574,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.812923633077271e-05,
+      "loss": 0.5968,
+      "step": 8141
+    },
+    {
+      "epoch": 2.4429439567859554,
+      "grad_norm": 1.2578125,
+      "learning_rate": 7.811417382135864e-05,
+      "loss": 0.6602,
+      "step": 8142
+    },
+    {
+      "epoch": 2.4432440543176535,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.809911131194458e-05,
+      "loss": 0.5651,
+      "step": 8143
+    },
+    {
+      "epoch": 2.443544151849351,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.80840488025305e-05,
+      "loss": 0.5946,
+      "step": 8144
+    },
+    {
+      "epoch": 2.4438442493810486,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.806898629311645e-05,
+      "loss": 0.6383,
+      "step": 8145
+    },
+    {
+      "epoch": 2.4441443469127466,
+      "grad_norm": 1.3125,
+      "learning_rate": 7.805392378370237e-05,
+      "loss": 0.5756,
+      "step": 8146
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.80388612742883e-05,
+      "loss": 0.5279,
+      "step": 8147
+    },
+    {
+      "epoch": 2.444744541976142,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.802379876487424e-05,
+      "loss": 0.5932,
+      "step": 8148
+    },
+    {
+      "epoch": 2.4450446395078402,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.800873625546017e-05,
+      "loss": 0.5384,
+      "step": 8149
+    },
+    {
+      "epoch": 2.445344737039538,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.799367374604609e-05,
+      "loss": 0.5488,
+      "step": 8150
+    },
+    {
+      "epoch": 2.445644834571236,
+      "grad_norm": 1.0,
+      "learning_rate": 7.797861123663203e-05,
+      "loss": 0.5705,
+      "step": 8151
+    },
+    {
+      "epoch": 2.4459449321029334,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.796354872721796e-05,
+      "loss": 0.6074,
+      "step": 8152
+    },
+    {
+      "epoch": 2.4462450296346314,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.794848621780389e-05,
+      "loss": 0.5124,
+      "step": 8153
+    },
+    {
+      "epoch": 2.446545127166329,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.793342370838981e-05,
+      "loss": 0.5881,
+      "step": 8154
+    },
+    {
+      "epoch": 2.446845224698027,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.791836119897575e-05,
+      "loss": 0.4855,
+      "step": 8155
+    },
+    {
+      "epoch": 2.4471453222297246,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.79032986895617e-05,
+      "loss": 0.4864,
+      "step": 8156
+    },
+    {
+      "epoch": 2.4474454197614226,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.788823618014761e-05,
+      "loss": 0.5735,
+      "step": 8157
+    },
+    {
+      "epoch": 2.44774551729312,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.787317367073355e-05,
+      "loss": 0.5552,
+      "step": 8158
+    },
+    {
+      "epoch": 2.448045614824818,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.785811116131947e-05,
+      "loss": 0.5296,
+      "step": 8159
+    },
+    {
+      "epoch": 2.4483457123565158,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.784304865190541e-05,
+      "loss": 0.5123,
+      "step": 8160
+    },
+    {
+      "epoch": 2.448645809888214,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.782798614249134e-05,
+      "loss": 0.5891,
+      "step": 8161
+    },
+    {
+      "epoch": 2.4489459074199114,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.781292363307727e-05,
+      "loss": 0.5254,
+      "step": 8162
+    },
+    {
+      "epoch": 2.4492460049516094,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.779786112366321e-05,
+      "loss": 0.5813,
+      "step": 8163
+    },
+    {
+      "epoch": 2.449546102483307,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.778279861424915e-05,
+      "loss": 0.4828,
+      "step": 8164
+    },
+    {
+      "epoch": 2.449846200015005,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.776773610483506e-05,
+      "loss": 0.5926,
+      "step": 8165
+    },
+    {
+      "epoch": 2.4501462975467025,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.7752673595421e-05,
+      "loss": 0.5364,
+      "step": 8166
+    },
+    {
+      "epoch": 2.4504463950784006,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.773761108600693e-05,
+      "loss": 0.5667,
+      "step": 8167
+    },
+    {
+      "epoch": 2.450746492610098,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.772254857659287e-05,
+      "loss": 0.6618,
+      "step": 8168
+    },
+    {
+      "epoch": 2.451046590141796,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.77074860671788e-05,
+      "loss": 0.4577,
+      "step": 8169
+    },
+    {
+      "epoch": 2.4513466876734937,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.769242355776472e-05,
+      "loss": 0.5774,
+      "step": 8170
+    },
+    {
+      "epoch": 2.4516467852051917,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.767736104835066e-05,
+      "loss": 0.606,
+      "step": 8171
+    },
+    {
+      "epoch": 2.4519468827368893,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.766229853893659e-05,
+      "loss": 0.5197,
+      "step": 8172
+    },
+    {
+      "epoch": 2.4522469802685873,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.764723602952252e-05,
+      "loss": 0.5796,
+      "step": 8173
+    },
+    {
+      "epoch": 2.452547077800285,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.763217352010846e-05,
+      "loss": 0.5097,
+      "step": 8174
+    },
+    {
+      "epoch": 2.452847175331983,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.761711101069438e-05,
+      "loss": 0.5214,
+      "step": 8175
+    },
+    {
+      "epoch": 2.4531472728636805,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.760204850128032e-05,
+      "loss": 0.6094,
+      "step": 8176
+    },
+    {
+      "epoch": 2.4534473703953785,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.758698599186625e-05,
+      "loss": 0.5993,
+      "step": 8177
+    },
+    {
+      "epoch": 2.4537474679270765,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.757192348245218e-05,
+      "loss": 0.5807,
+      "step": 8178
+    },
+    {
+      "epoch": 2.454047565458774,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.755686097303812e-05,
+      "loss": 0.5732,
+      "step": 8179
+    },
+    {
+      "epoch": 2.4543476629904717,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.754179846362404e-05,
+      "loss": 0.5464,
+      "step": 8180
+    },
+    {
+      "epoch": 2.4546477605221697,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.752673595420997e-05,
+      "loss": 0.5052,
+      "step": 8181
+    },
+    {
+      "epoch": 2.4549478580538677,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.751167344479591e-05,
+      "loss": 0.5681,
+      "step": 8182
+    },
+    {
+      "epoch": 2.4552479555855653,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.749661093538184e-05,
+      "loss": 0.5695,
+      "step": 8183
+    },
+    {
+      "epoch": 2.455548053117263,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.748154842596776e-05,
+      "loss": 0.5299,
+      "step": 8184
+    },
+    {
+      "epoch": 2.455848150648961,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.74664859165537e-05,
+      "loss": 0.5566,
+      "step": 8185
+    },
+    {
+      "epoch": 2.456148248180659,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.745142340713963e-05,
+      "loss": 0.5822,
+      "step": 8186
+    },
+    {
+      "epoch": 2.4564483457123565,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.743636089772557e-05,
+      "loss": 0.5333,
+      "step": 8187
+    },
+    {
+      "epoch": 2.4567484432440545,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.742129838831149e-05,
+      "loss": 0.5508,
+      "step": 8188
+    },
+    {
+      "epoch": 2.457048540775752,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.740623587889743e-05,
+      "loss": 0.5783,
+      "step": 8189
+    },
+    {
+      "epoch": 2.45734863830745,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.739117336948337e-05,
+      "loss": 0.675,
+      "step": 8190
+    },
+    {
+      "epoch": 2.4576487358391477,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.737611086006929e-05,
+      "loss": 0.5574,
+      "step": 8191
+    },
+    {
+      "epoch": 2.4579488333708457,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.736104835065522e-05,
+      "loss": 0.6041,
+      "step": 8192
+    },
+    {
+      "epoch": 2.4582489309025433,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.734598584124116e-05,
+      "loss": 0.5614,
+      "step": 8193
+    },
+    {
+      "epoch": 2.4585490284342413,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.733092333182709e-05,
+      "loss": 0.476,
+      "step": 8194
+    },
+    {
+      "epoch": 2.458849125965939,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.731586082241303e-05,
+      "loss": 0.5639,
+      "step": 8195
+    },
+    {
+      "epoch": 2.459149223497637,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.730079831299894e-05,
+      "loss": 0.5944,
+      "step": 8196
+    },
+    {
+      "epoch": 2.4594493210293344,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.728573580358488e-05,
+      "loss": 0.4974,
+      "step": 8197
+    },
+    {
+      "epoch": 2.4597494185610325,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.727067329417082e-05,
+      "loss": 0.6825,
+      "step": 8198
+    },
+    {
+      "epoch": 2.46004951609273,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.725561078475675e-05,
+      "loss": 0.5454,
+      "step": 8199
+    },
+    {
+      "epoch": 2.460349613624428,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.724054827534267e-05,
+      "loss": 0.5667,
+      "step": 8200
+    },
+    {
+      "epoch": 2.4606497111561256,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.72254857659286e-05,
+      "loss": 0.4711,
+      "step": 8201
+    },
+    {
+      "epoch": 2.4609498086878236,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.721042325651454e-05,
+      "loss": 0.6042,
+      "step": 8202
+    },
+    {
+      "epoch": 2.461249906219521,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.719536074710048e-05,
+      "loss": 0.645,
+      "step": 8203
+    },
+    {
+      "epoch": 2.4615500037512192,
+      "grad_norm": 1.125,
+      "learning_rate": 7.71802982376864e-05,
+      "loss": 0.5893,
+      "step": 8204
+    },
+    {
+      "epoch": 2.461850101282917,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.716523572827233e-05,
+      "loss": 0.6508,
+      "step": 8205
+    },
+    {
+      "epoch": 2.462150198814615,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.715017321885826e-05,
+      "loss": 0.5885,
+      "step": 8206
+    },
+    {
+      "epoch": 2.4624502963463124,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.713511070944419e-05,
+      "loss": 0.6263,
+      "step": 8207
+    },
+    {
+      "epoch": 2.4627503938780104,
+      "grad_norm": 1.0,
+      "learning_rate": 7.712004820003013e-05,
+      "loss": 0.6015,
+      "step": 8208
+    },
+    {
+      "epoch": 2.463050491409708,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.710498569061606e-05,
+      "loss": 0.5829,
+      "step": 8209
+    },
+    {
+      "epoch": 2.463350588941406,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.7089923181202e-05,
+      "loss": 0.6356,
+      "step": 8210
+    },
+    {
+      "epoch": 2.4636506864731036,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.707486067178792e-05,
+      "loss": 0.5954,
+      "step": 8211
+    },
+    {
+      "epoch": 2.4639507840048016,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.705979816237385e-05,
+      "loss": 0.4996,
+      "step": 8212
+    },
+    {
+      "epoch": 2.464250881536499,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.704473565295979e-05,
+      "loss": 0.5379,
+      "step": 8213
+    },
+    {
+      "epoch": 2.464550979068197,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.702967314354572e-05,
+      "loss": 0.5749,
+      "step": 8214
+    },
+    {
+      "epoch": 2.4648510765998948,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.701461063413164e-05,
+      "loss": 0.5859,
+      "step": 8215
+    },
+    {
+      "epoch": 2.465151174131593,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.699954812471758e-05,
+      "loss": 0.5693,
+      "step": 8216
+    },
+    {
+      "epoch": 2.465451271663291,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.698448561530351e-05,
+      "loss": 0.5269,
+      "step": 8217
+    },
+    {
+      "epoch": 2.4657513691949884,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.696942310588945e-05,
+      "loss": 0.6627,
+      "step": 8218
+    },
+    {
+      "epoch": 2.466051466726686,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.695436059647538e-05,
+      "loss": 0.518,
+      "step": 8219
+    },
+    {
+      "epoch": 2.466351564258384,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.69392980870613e-05,
+      "loss": 0.5327,
+      "step": 8220
+    },
+    {
+      "epoch": 2.466651661790082,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.692423557764724e-05,
+      "loss": 0.5826,
+      "step": 8221
+    },
+    {
+      "epoch": 2.4669517593217796,
+      "grad_norm": 1.0,
+      "learning_rate": 7.690917306823317e-05,
+      "loss": 0.4929,
+      "step": 8222
+    },
+    {
+      "epoch": 2.467251856853477,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.68941105588191e-05,
+      "loss": 0.5667,
+      "step": 8223
+    },
+    {
+      "epoch": 2.467551954385175,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.687904804940504e-05,
+      "loss": 0.5487,
+      "step": 8224
+    },
+    {
+      "epoch": 2.467852051916873,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.686398553999096e-05,
+      "loss": 0.524,
+      "step": 8225
+    },
+    {
+      "epoch": 2.4681521494485708,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.68489230305769e-05,
+      "loss": 0.5703,
+      "step": 8226
+    },
+    {
+      "epoch": 2.4684522469802688,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.683386052116283e-05,
+      "loss": 0.5399,
+      "step": 8227
+    },
+    {
+      "epoch": 2.4687523445119663,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.681879801174876e-05,
+      "loss": 0.5513,
+      "step": 8228
+    },
+    {
+      "epoch": 2.4690524420436644,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.68037355023347e-05,
+      "loss": 0.5109,
+      "step": 8229
+    },
+    {
+      "epoch": 2.469352539575362,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.678867299292063e-05,
+      "loss": 0.6312,
+      "step": 8230
+    },
+    {
+      "epoch": 2.46965263710706,
+      "grad_norm": 1.125,
+      "learning_rate": 7.677361048350655e-05,
+      "loss": 0.5382,
+      "step": 8231
+    },
+    {
+      "epoch": 2.4699527346387575,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.675854797409249e-05,
+      "loss": 0.5423,
+      "step": 8232
+    },
+    {
+      "epoch": 2.4702528321704555,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.674348546467842e-05,
+      "loss": 0.4728,
+      "step": 8233
+    },
+    {
+      "epoch": 2.470552929702153,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.672842295526435e-05,
+      "loss": 0.4979,
+      "step": 8234
+    },
+    {
+      "epoch": 2.470853027233851,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.671336044585027e-05,
+      "loss": 0.5184,
+      "step": 8235
+    },
+    {
+      "epoch": 2.4711531247655487,
+      "grad_norm": 1.125,
+      "learning_rate": 7.669829793643621e-05,
+      "loss": 0.5316,
+      "step": 8236
+    },
+    {
+      "epoch": 2.4714532222972467,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.668323542702215e-05,
+      "loss": 0.5523,
+      "step": 8237
+    },
+    {
+      "epoch": 2.4717533198289443,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.666817291760807e-05,
+      "loss": 0.6254,
+      "step": 8238
+    },
+    {
+      "epoch": 2.4720534173606423,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.665311040819401e-05,
+      "loss": 0.5573,
+      "step": 8239
+    },
+    {
+      "epoch": 2.47235351489234,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.663804789877995e-05,
+      "loss": 0.5945,
+      "step": 8240
+    },
+    {
+      "epoch": 2.472653612424038,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.662298538936587e-05,
+      "loss": 0.6311,
+      "step": 8241
+    },
+    {
+      "epoch": 2.4729537099557355,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.66079228799518e-05,
+      "loss": 0.6353,
+      "step": 8242
+    },
+    {
+      "epoch": 2.4732538074874335,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.659286037053773e-05,
+      "loss": 0.5246,
+      "step": 8243
+    },
+    {
+      "epoch": 2.473553905019131,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.657779786112367e-05,
+      "loss": 0.5994,
+      "step": 8244
+    },
+    {
+      "epoch": 2.473854002550829,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.656273535170961e-05,
+      "loss": 0.5992,
+      "step": 8245
+    },
+    {
+      "epoch": 2.4741541000825267,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.654767284229552e-05,
+      "loss": 0.5054,
+      "step": 8246
+    },
+    {
+      "epoch": 2.4744541976142247,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.653261033288146e-05,
+      "loss": 0.6235,
+      "step": 8247
+    },
+    {
+      "epoch": 2.4747542951459223,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.651754782346739e-05,
+      "loss": 0.6329,
+      "step": 8248
+    },
+    {
+      "epoch": 2.4750543926776203,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.650248531405333e-05,
+      "loss": 0.6859,
+      "step": 8249
+    },
+    {
+      "epoch": 2.475354490209318,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.648742280463926e-05,
+      "loss": 0.5193,
+      "step": 8250
+    },
+    {
+      "epoch": 2.475654587741016,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.647236029522518e-05,
+      "loss": 0.5245,
+      "step": 8251
+    },
+    {
+      "epoch": 2.4759546852727135,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.645729778581112e-05,
+      "loss": 0.6226,
+      "step": 8252
+    },
+    {
+      "epoch": 2.4762547828044115,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.644223527639705e-05,
+      "loss": 0.6241,
+      "step": 8253
+    },
+    {
+      "epoch": 2.476554880336109,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.642717276698298e-05,
+      "loss": 0.5065,
+      "step": 8254
+    },
+    {
+      "epoch": 2.476854977867807,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.641211025756892e-05,
+      "loss": 0.5008,
+      "step": 8255
+    },
+    {
+      "epoch": 2.477155075399505,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.639704774815484e-05,
+      "loss": 0.551,
+      "step": 8256
+    },
+    {
+      "epoch": 2.4774551729312027,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.638198523874078e-05,
+      "loss": 0.5236,
+      "step": 8257
+    },
+    {
+      "epoch": 2.4777552704629002,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.636692272932671e-05,
+      "loss": 0.579,
+      "step": 8258
+    },
+    {
+      "epoch": 2.4780553679945982,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.635186021991264e-05,
+      "loss": 0.5604,
+      "step": 8259
+    },
+    {
+      "epoch": 2.4783554655262963,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.633679771049858e-05,
+      "loss": 0.5176,
+      "step": 8260
+    },
+    {
+      "epoch": 2.478655563057994,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.63217352010845e-05,
+      "loss": 0.4982,
+      "step": 8261
+    },
+    {
+      "epoch": 2.4789556605896914,
+      "grad_norm": 1.0,
+      "learning_rate": 7.630667269167043e-05,
+      "loss": 0.5433,
+      "step": 8262
+    },
+    {
+      "epoch": 2.4792557581213894,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.629161018225637e-05,
+      "loss": 0.5625,
+      "step": 8263
+    },
+    {
+      "epoch": 2.4795558556530874,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.62765476728423e-05,
+      "loss": 0.5945,
+      "step": 8264
+    },
+    {
+      "epoch": 2.479855953184785,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.626148516342822e-05,
+      "loss": 0.5871,
+      "step": 8265
+    },
+    {
+      "epoch": 2.480156050716483,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.624642265401416e-05,
+      "loss": 0.6681,
+      "step": 8266
+    },
+    {
+      "epoch": 2.4804561482481806,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.623136014460009e-05,
+      "loss": 0.5161,
+      "step": 8267
+    },
+    {
+      "epoch": 2.4807562457798786,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.621629763518603e-05,
+      "loss": 0.5831,
+      "step": 8268
+    },
+    {
+      "epoch": 2.481056343311576,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.620123512577196e-05,
+      "loss": 0.5435,
+      "step": 8269
+    },
+    {
+      "epoch": 2.4813564408432742,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.618617261635789e-05,
+      "loss": 0.5073,
+      "step": 8270
+    },
+    {
+      "epoch": 2.481656538374972,
+      "grad_norm": 1.3359375,
+      "learning_rate": 7.617111010694383e-05,
+      "loss": 0.6116,
+      "step": 8271
+    },
+    {
+      "epoch": 2.48195663590667,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.615604759752975e-05,
+      "loss": 0.5401,
+      "step": 8272
+    },
+    {
+      "epoch": 2.4822567334383674,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.614098508811568e-05,
+      "loss": 0.5987,
+      "step": 8273
+    },
+    {
+      "epoch": 2.4825568309700654,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.612592257870162e-05,
+      "loss": 0.5511,
+      "step": 8274
+    },
+    {
+      "epoch": 2.482856928501763,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.611086006928755e-05,
+      "loss": 0.5672,
+      "step": 8275
+    },
+    {
+      "epoch": 2.483157026033461,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.609579755987349e-05,
+      "loss": 0.5255,
+      "step": 8276
+    },
+    {
+      "epoch": 2.4834571235651586,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.60807350504594e-05,
+      "loss": 0.5565,
+      "step": 8277
+    },
+    {
+      "epoch": 2.4837572210968566,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.606567254104534e-05,
+      "loss": 0.6331,
+      "step": 8278
+    },
+    {
+      "epoch": 2.484057318628554,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.605061003163128e-05,
+      "loss": 0.5014,
+      "step": 8279
+    },
+    {
+      "epoch": 2.484357416160252,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.603554752221721e-05,
+      "loss": 0.5538,
+      "step": 8280
+    },
+    {
+      "epoch": 2.4846575136919498,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.602048501280313e-05,
+      "loss": 0.5768,
+      "step": 8281
+    },
+    {
+      "epoch": 2.4849576112236478,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.600542250338907e-05,
+      "loss": 0.5878,
+      "step": 8282
+    },
+    {
+      "epoch": 2.4852577087553454,
+      "grad_norm": 1.296875,
+      "learning_rate": 7.5990359993975e-05,
+      "loss": 0.513,
+      "step": 8283
+    },
+    {
+      "epoch": 2.4855578062870434,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.597529748456094e-05,
+      "loss": 0.5478,
+      "step": 8284
+    },
+    {
+      "epoch": 2.485857903818741,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.596023497514685e-05,
+      "loss": 0.567,
+      "step": 8285
+    },
+    {
+      "epoch": 2.486158001350439,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.59451724657328e-05,
+      "loss": 0.4916,
+      "step": 8286
+    },
+    {
+      "epoch": 2.4864580988821365,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.593010995631874e-05,
+      "loss": 0.5482,
+      "step": 8287
+    },
+    {
+      "epoch": 2.4867581964138346,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.591504744690465e-05,
+      "loss": 0.6203,
+      "step": 8288
+    },
+    {
+      "epoch": 2.487058293945532,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.589998493749059e-05,
+      "loss": 0.549,
+      "step": 8289
+    },
+    {
+      "epoch": 2.48735839147723,
+      "grad_norm": 1.125,
+      "learning_rate": 7.588492242807652e-05,
+      "loss": 0.6093,
+      "step": 8290
+    },
+    {
+      "epoch": 2.4876584890089277,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.586985991866246e-05,
+      "loss": 0.6031,
+      "step": 8291
+    },
+    {
+      "epoch": 2.4879585865406257,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.585479740924838e-05,
+      "loss": 0.5428,
+      "step": 8292
+    },
+    {
+      "epoch": 2.4882586840723233,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.583973489983431e-05,
+      "loss": 0.5654,
+      "step": 8293
+    },
+    {
+      "epoch": 2.4885587816040213,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.582467239042025e-05,
+      "loss": 0.5804,
+      "step": 8294
+    },
+    {
+      "epoch": 2.4888588791357193,
+      "grad_norm": 1.0,
+      "learning_rate": 7.580960988100618e-05,
+      "loss": 0.6275,
+      "step": 8295
+    },
+    {
+      "epoch": 2.489158976667417,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.57945473715921e-05,
+      "loss": 0.6627,
+      "step": 8296
+    },
+    {
+      "epoch": 2.4894590741991145,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.577948486217804e-05,
+      "loss": 0.4927,
+      "step": 8297
+    },
+    {
+      "epoch": 2.4897591717308125,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.576442235276397e-05,
+      "loss": 0.573,
+      "step": 8298
+    },
+    {
+      "epoch": 2.4900592692625105,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.574935984334991e-05,
+      "loss": 0.5624,
+      "step": 8299
+    },
+    {
+      "epoch": 2.490359366794208,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.573429733393584e-05,
+      "loss": 0.5223,
+      "step": 8300
+    },
+    {
+      "epoch": 2.4906594643259057,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.571923482452176e-05,
+      "loss": 0.4902,
+      "step": 8301
+    },
+    {
+      "epoch": 2.4909595618576037,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.57041723151077e-05,
+      "loss": 0.634,
+      "step": 8302
+    },
+    {
+      "epoch": 2.4912596593893017,
+      "grad_norm": 1.0,
+      "learning_rate": 7.568910980569363e-05,
+      "loss": 0.6179,
+      "step": 8303
+    },
+    {
+      "epoch": 2.4915597569209993,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.567404729627956e-05,
+      "loss": 0.583,
+      "step": 8304
+    },
+    {
+      "epoch": 2.4918598544526973,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.56589847868655e-05,
+      "loss": 0.5331,
+      "step": 8305
+    },
+    {
+      "epoch": 2.492159951984395,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.564392227745142e-05,
+      "loss": 0.5238,
+      "step": 8306
+    },
+    {
+      "epoch": 2.492460049516093,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.562885976803737e-05,
+      "loss": 0.6129,
+      "step": 8307
+    },
+    {
+      "epoch": 2.4927601470477905,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.561379725862329e-05,
+      "loss": 0.5622,
+      "step": 8308
+    },
+    {
+      "epoch": 2.4930602445794885,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.559873474920922e-05,
+      "loss": 0.5576,
+      "step": 8309
+    },
+    {
+      "epoch": 2.493360342111186,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.558367223979516e-05,
+      "loss": 0.4467,
+      "step": 8310
+    },
+    {
+      "epoch": 2.493660439642884,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.556860973038109e-05,
+      "loss": 0.55,
+      "step": 8311
+    },
+    {
+      "epoch": 2.4939605371745817,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.555354722096701e-05,
+      "loss": 0.6822,
+      "step": 8312
+    },
+    {
+      "epoch": 2.4942606347062797,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.553848471155295e-05,
+      "loss": 0.5119,
+      "step": 8313
+    },
+    {
+      "epoch": 2.4945607322379773,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.552342220213888e-05,
+      "loss": 0.6142,
+      "step": 8314
+    },
+    {
+      "epoch": 2.4948608297696753,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.55083596927248e-05,
+      "loss": 0.5909,
+      "step": 8315
+    },
+    {
+      "epoch": 2.495160927301373,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.549329718331075e-05,
+      "loss": 0.5836,
+      "step": 8316
+    },
+    {
+      "epoch": 2.495461024833071,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.547823467389667e-05,
+      "loss": 0.5175,
+      "step": 8317
+    },
+    {
+      "epoch": 2.4957611223647684,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.546317216448261e-05,
+      "loss": 0.6029,
+      "step": 8318
+    },
+    {
+      "epoch": 2.4960612198964665,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.544810965506853e-05,
+      "loss": 0.5422,
+      "step": 8319
+    },
+    {
+      "epoch": 2.496361317428164,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.543304714565447e-05,
+      "loss": 0.5188,
+      "step": 8320
+    },
+    {
+      "epoch": 2.496661414959862,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.541798463624041e-05,
+      "loss": 0.5709,
+      "step": 8321
+    },
+    {
+      "epoch": 2.4969615124915596,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.540292212682633e-05,
+      "loss": 0.5933,
+      "step": 8322
+    },
+    {
+      "epoch": 2.4972616100232576,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.538785961741226e-05,
+      "loss": 0.5245,
+      "step": 8323
+    },
+    {
+      "epoch": 2.497561707554955,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.537279710799819e-05,
+      "loss": 0.5792,
+      "step": 8324
+    },
+    {
+      "epoch": 2.4978618050866532,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.535773459858413e-05,
+      "loss": 0.556,
+      "step": 8325
+    },
+    {
+      "epoch": 2.498161902618351,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.534267208917007e-05,
+      "loss": 0.6065,
+      "step": 8326
+    },
+    {
+      "epoch": 2.498462000150049,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.532760957975598e-05,
+      "loss": 0.6215,
+      "step": 8327
+    },
+    {
+      "epoch": 2.4987620976817464,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.531254707034192e-05,
+      "loss": 0.6667,
+      "step": 8328
+    },
+    {
+      "epoch": 2.4990621952134444,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.529748456092786e-05,
+      "loss": 0.5067,
+      "step": 8329
+    },
+    {
+      "epoch": 2.499362292745142,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.528242205151379e-05,
+      "loss": 0.4826,
+      "step": 8330
+    },
+    {
+      "epoch": 2.49966239027684,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.526735954209972e-05,
+      "loss": 0.5868,
+      "step": 8331
+    },
+    {
+      "epoch": 2.4999624878085376,
+      "grad_norm": 1.125,
+      "learning_rate": 7.525229703268564e-05,
+      "loss": 0.5325,
+      "step": 8332
+    },
+    {
+      "epoch": 2.5002625853402356,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.523723452327158e-05,
+      "loss": 0.5209,
+      "step": 8333
+    },
+    {
+      "epoch": 2.5005626828719336,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.522217201385752e-05,
+      "loss": 0.5801,
+      "step": 8334
+    },
+    {
+      "epoch": 2.500862780403631,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.520710950444344e-05,
+      "loss": 0.5675,
+      "step": 8335
+    },
+    {
+      "epoch": 2.5011628779353288,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.519204699502938e-05,
+      "loss": 0.5783,
+      "step": 8336
+    },
+    {
+      "epoch": 2.501462975467027,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.51769844856153e-05,
+      "loss": 0.6045,
+      "step": 8337
+    },
+    {
+      "epoch": 2.501763072998725,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.516192197620124e-05,
+      "loss": 0.6579,
+      "step": 8338
+    },
+    {
+      "epoch": 2.5020631705304224,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.514685946678717e-05,
+      "loss": 0.6573,
+      "step": 8339
+    },
+    {
+      "epoch": 2.50236326806212,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.51317969573731e-05,
+      "loss": 0.6043,
+      "step": 8340
+    },
+    {
+      "epoch": 2.502663365593818,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.511673444795904e-05,
+      "loss": 0.5461,
+      "step": 8341
+    },
+    {
+      "epoch": 2.502963463125516,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.510167193854496e-05,
+      "loss": 0.4775,
+      "step": 8342
+    },
+    {
+      "epoch": 2.5032635606572136,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.508660942913089e-05,
+      "loss": 0.5428,
+      "step": 8343
+    },
+    {
+      "epoch": 2.503563658188911,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.507154691971683e-05,
+      "loss": 0.6606,
+      "step": 8344
+    },
+    {
+      "epoch": 2.503863755720609,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.505648441030276e-05,
+      "loss": 0.5279,
+      "step": 8345
+    },
+    {
+      "epoch": 2.504163853252307,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.504142190088868e-05,
+      "loss": 0.4953,
+      "step": 8346
+    },
+    {
+      "epoch": 2.5044639507840047,
+      "grad_norm": 1.125,
+      "learning_rate": 7.502635939147463e-05,
+      "loss": 0.6093,
+      "step": 8347
+    },
+    {
+      "epoch": 2.5047640483157028,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.501129688206055e-05,
+      "loss": 0.6247,
+      "step": 8348
+    },
+    {
+      "epoch": 2.5050641458474003,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.499623437264649e-05,
+      "loss": 0.515,
+      "step": 8349
+    },
+    {
+      "epoch": 2.5053642433790984,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.498117186323242e-05,
+      "loss": 0.5916,
+      "step": 8350
+    },
+    {
+      "epoch": 2.505664340910796,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.496610935381835e-05,
+      "loss": 0.5686,
+      "step": 8351
+    },
+    {
+      "epoch": 2.505964438442494,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.495104684440429e-05,
+      "loss": 0.6039,
+      "step": 8352
+    },
+    {
+      "epoch": 2.5062645359741915,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.493598433499021e-05,
+      "loss": 0.4895,
+      "step": 8353
+    },
+    {
+      "epoch": 2.5065646335058895,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.492092182557614e-05,
+      "loss": 0.5569,
+      "step": 8354
+    },
+    {
+      "epoch": 2.506864731037587,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.490585931616208e-05,
+      "loss": 0.5775,
+      "step": 8355
+    },
+    {
+      "epoch": 2.507164828569285,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.4890796806748e-05,
+      "loss": 0.4477,
+      "step": 8356
+    },
+    {
+      "epoch": 2.5074649261009827,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.487573429733395e-05,
+      "loss": 0.5185,
+      "step": 8357
+    },
+    {
+      "epoch": 2.5077650236326807,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.486067178791987e-05,
+      "loss": 0.5229,
+      "step": 8358
+    },
+    {
+      "epoch": 2.5080651211643783,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.48456092785058e-05,
+      "loss": 0.5468,
+      "step": 8359
+    },
+    {
+      "epoch": 2.5083652186960763,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.483054676909174e-05,
+      "loss": 0.6101,
+      "step": 8360
+    },
+    {
+      "epoch": 2.508665316227774,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.481548425967767e-05,
+      "loss": 0.4711,
+      "step": 8361
+    },
+    {
+      "epoch": 2.508965413759472,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.48004217502636e-05,
+      "loss": 0.695,
+      "step": 8362
+    },
+    {
+      "epoch": 2.5092655112911695,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.478535924084953e-05,
+      "loss": 0.4416,
+      "step": 8363
+    },
+    {
+      "epoch": 2.5095656088228675,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.477029673143546e-05,
+      "loss": 0.5433,
+      "step": 8364
+    },
+    {
+      "epoch": 2.509865706354565,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.47552342220214e-05,
+      "loss": 0.5227,
+      "step": 8365
+    },
+    {
+      "epoch": 2.510165803886263,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.474017171260731e-05,
+      "loss": 0.5167,
+      "step": 8366
+    },
+    {
+      "epoch": 2.5104659014179607,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.472510920319325e-05,
+      "loss": 0.6346,
+      "step": 8367
+    },
+    {
+      "epoch": 2.5107659989496587,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.47100466937792e-05,
+      "loss": 0.6242,
+      "step": 8368
+    },
+    {
+      "epoch": 2.5110660964813567,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.469498418436512e-05,
+      "loss": 0.5031,
+      "step": 8369
+    },
+    {
+      "epoch": 2.5113661940130543,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.467992167495105e-05,
+      "loss": 0.5531,
+      "step": 8370
+    },
+    {
+      "epoch": 2.511666291544752,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.466485916553698e-05,
+      "loss": 0.4801,
+      "step": 8371
+    },
+    {
+      "epoch": 2.51196638907645,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.464979665612292e-05,
+      "loss": 0.6452,
+      "step": 8372
+    },
+    {
+      "epoch": 2.512266486608148,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.463473414670884e-05,
+      "loss": 0.6519,
+      "step": 8373
+    },
+    {
+      "epoch": 2.5125665841398455,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.461967163729477e-05,
+      "loss": 0.6282,
+      "step": 8374
+    },
+    {
+      "epoch": 2.512866681671543,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.460460912788071e-05,
+      "loss": 0.4729,
+      "step": 8375
+    },
+    {
+      "epoch": 2.513166779203241,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.458954661846665e-05,
+      "loss": 0.5791,
+      "step": 8376
+    },
+    {
+      "epoch": 2.513466876734939,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.457448410905256e-05,
+      "loss": 0.5212,
+      "step": 8377
+    },
+    {
+      "epoch": 2.5137669742666366,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.45594215996385e-05,
+      "loss": 0.5066,
+      "step": 8378
+    },
+    {
+      "epoch": 2.514067071798334,
+      "grad_norm": 1.0,
+      "learning_rate": 7.454435909022443e-05,
+      "loss": 0.5783,
+      "step": 8379
+    },
+    {
+      "epoch": 2.5143671693300322,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.452929658081037e-05,
+      "loss": 0.5771,
+      "step": 8380
+    },
+    {
+      "epoch": 2.5146672668617303,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.45142340713963e-05,
+      "loss": 0.5716,
+      "step": 8381
+    },
+    {
+      "epoch": 2.514967364393428,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.449917156198222e-05,
+      "loss": 0.5801,
+      "step": 8382
+    },
+    {
+      "epoch": 2.5152674619251254,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.448410905256816e-05,
+      "loss": 0.5372,
+      "step": 8383
+    },
+    {
+      "epoch": 2.5155675594568234,
+      "grad_norm": 1.3515625,
+      "learning_rate": 7.446904654315409e-05,
+      "loss": 0.726,
+      "step": 8384
+    },
+    {
+      "epoch": 2.5158676569885214,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.445398403374002e-05,
+      "loss": 0.5574,
+      "step": 8385
+    },
+    {
+      "epoch": 2.516167754520219,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.443892152432596e-05,
+      "loss": 0.4899,
+      "step": 8386
+    },
+    {
+      "epoch": 2.516467852051917,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.442385901491188e-05,
+      "loss": 0.4592,
+      "step": 8387
+    },
+    {
+      "epoch": 2.5167679495836146,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.440879650549783e-05,
+      "loss": 0.5818,
+      "step": 8388
+    },
+    {
+      "epoch": 2.5170680471153126,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.439373399608375e-05,
+      "loss": 0.6003,
+      "step": 8389
+    },
+    {
+      "epoch": 2.51736814464701,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.437867148666968e-05,
+      "loss": 0.5851,
+      "step": 8390
+    },
+    {
+      "epoch": 2.517668242178708,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.436360897725562e-05,
+      "loss": 0.5674,
+      "step": 8391
+    },
+    {
+      "epoch": 2.517968339710406,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.434854646784155e-05,
+      "loss": 0.6663,
+      "step": 8392
+    },
+    {
+      "epoch": 2.518268437242104,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.433348395842747e-05,
+      "loss": 0.5712,
+      "step": 8393
+    },
+    {
+      "epoch": 2.5185685347738014,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.431842144901341e-05,
+      "loss": 0.6585,
+      "step": 8394
+    },
+    {
+      "epoch": 2.5188686323054994,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.430335893959934e-05,
+      "loss": 0.5853,
+      "step": 8395
+    },
+    {
+      "epoch": 2.519168729837197,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.428829643018528e-05,
+      "loss": 0.4757,
+      "step": 8396
+    },
+    {
+      "epoch": 2.519468827368895,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.42732339207712e-05,
+      "loss": 0.5801,
+      "step": 8397
+    },
+    {
+      "epoch": 2.5197689249005926,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.425817141135713e-05,
+      "loss": 0.6421,
+      "step": 8398
+    },
+    {
+      "epoch": 2.5200690224322906,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.424310890194307e-05,
+      "loss": 0.5759,
+      "step": 8399
+    },
+    {
+      "epoch": 2.520369119963988,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.4228046392529e-05,
+      "loss": 0.4764,
+      "step": 8400
+    },
+    {
+      "epoch": 2.520669217495686,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.421298388311493e-05,
+      "loss": 0.4652,
+      "step": 8401
+    },
+    {
+      "epoch": 2.5209693150273838,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.419792137370087e-05,
+      "loss": 0.5097,
+      "step": 8402
+    },
+    {
+      "epoch": 2.5212694125590818,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.41828588642868e-05,
+      "loss": 0.5905,
+      "step": 8403
+    },
+    {
+      "epoch": 2.5215695100907793,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.416779635487272e-05,
+      "loss": 0.5017,
+      "step": 8404
+    },
+    {
+      "epoch": 2.5218696076224774,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.415273384545866e-05,
+      "loss": 0.6341,
+      "step": 8405
+    },
+    {
+      "epoch": 2.522169705154175,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.413767133604459e-05,
+      "loss": 0.4531,
+      "step": 8406
+    },
+    {
+      "epoch": 2.522469802685873,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.412260882663053e-05,
+      "loss": 0.5268,
+      "step": 8407
+    },
+    {
+      "epoch": 2.522769900217571,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.410754631721644e-05,
+      "loss": 0.6068,
+      "step": 8408
+    },
+    {
+      "epoch": 2.5230699977492685,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.409248380780238e-05,
+      "loss": 0.5193,
+      "step": 8409
+    },
+    {
+      "epoch": 2.523370095280966,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.407742129838832e-05,
+      "loss": 0.4851,
+      "step": 8410
+    },
+    {
+      "epoch": 2.523670192812664,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.406235878897425e-05,
+      "loss": 0.5883,
+      "step": 8411
+    },
+    {
+      "epoch": 2.523970290344362,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.404729627956018e-05,
+      "loss": 0.6073,
+      "step": 8412
+    },
+    {
+      "epoch": 2.5242703878760597,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.40322337701461e-05,
+      "loss": 0.5144,
+      "step": 8413
+    },
+    {
+      "epoch": 2.5245704854077573,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.401717126073204e-05,
+      "loss": 0.5422,
+      "step": 8414
+    },
+    {
+      "epoch": 2.5248705829394553,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.400210875131798e-05,
+      "loss": 0.5488,
+      "step": 8415
+    },
+    {
+      "epoch": 2.5251706804711533,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.39870462419039e-05,
+      "loss": 0.566,
+      "step": 8416
+    },
+    {
+      "epoch": 2.525470778002851,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.397198373248984e-05,
+      "loss": 0.5665,
+      "step": 8417
+    },
+    {
+      "epoch": 2.5257708755345485,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.395692122307578e-05,
+      "loss": 0.5202,
+      "step": 8418
+    },
+    {
+      "epoch": 2.5260709730662465,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.39418587136617e-05,
+      "loss": 0.4842,
+      "step": 8419
+    },
+    {
+      "epoch": 2.5263710705979445,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.392679620424763e-05,
+      "loss": 0.4941,
+      "step": 8420
+    },
+    {
+      "epoch": 2.526671168129642,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.391173369483356e-05,
+      "loss": 0.5494,
+      "step": 8421
+    },
+    {
+      "epoch": 2.5269712656613397,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.38966711854195e-05,
+      "loss": 0.6081,
+      "step": 8422
+    },
+    {
+      "epoch": 2.5272713631930377,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.388160867600544e-05,
+      "loss": 0.6061,
+      "step": 8423
+    },
+    {
+      "epoch": 2.5275714607247357,
+      "grad_norm": 0.97265625,
+      "learning_rate": 7.386654616659135e-05,
+      "loss": 0.5499,
+      "step": 8424
+    },
+    {
+      "epoch": 2.5278715582564333,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.385148365717729e-05,
+      "loss": 0.6226,
+      "step": 8425
+    },
+    {
+      "epoch": 2.5281716557881313,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.383642114776322e-05,
+      "loss": 0.6379,
+      "step": 8426
+    },
+    {
+      "epoch": 2.528471753319829,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.382135863834914e-05,
+      "loss": 0.4759,
+      "step": 8427
+    },
+    {
+      "epoch": 2.528771850851527,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.380629612893509e-05,
+      "loss": 0.5455,
+      "step": 8428
+    },
+    {
+      "epoch": 2.5290719483832245,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.379123361952101e-05,
+      "loss": 0.5304,
+      "step": 8429
+    },
+    {
+      "epoch": 2.5293720459149225,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.377617111010695e-05,
+      "loss": 0.6429,
+      "step": 8430
+    },
+    {
+      "epoch": 2.52967214344662,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.376110860069288e-05,
+      "loss": 0.5887,
+      "step": 8431
+    },
+    {
+      "epoch": 2.529972240978318,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.37460460912788e-05,
+      "loss": 0.491,
+      "step": 8432
+    },
+    {
+      "epoch": 2.5302723385100157,
+      "grad_norm": 1.328125,
+      "learning_rate": 7.373098358186475e-05,
+      "loss": 0.6382,
+      "step": 8433
+    },
+    {
+      "epoch": 2.5305724360417137,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.371592107245067e-05,
+      "loss": 0.5803,
+      "step": 8434
+    },
+    {
+      "epoch": 2.5308725335734112,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.37008585630366e-05,
+      "loss": 0.542,
+      "step": 8435
+    },
+    {
+      "epoch": 2.5311726311051093,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.368579605362254e-05,
+      "loss": 0.5471,
+      "step": 8436
+    },
+    {
+      "epoch": 2.531472728636807,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.367073354420847e-05,
+      "loss": 0.497,
+      "step": 8437
+    },
+    {
+      "epoch": 2.531772826168505,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.36556710347944e-05,
+      "loss": 0.527,
+      "step": 8438
+    },
+    {
+      "epoch": 2.5320729237002024,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.364060852538033e-05,
+      "loss": 0.5366,
+      "step": 8439
+    },
+    {
+      "epoch": 2.5323730212319004,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.362554601596626e-05,
+      "loss": 0.5861,
+      "step": 8440
+    },
+    {
+      "epoch": 2.532673118763598,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.36104835065522e-05,
+      "loss": 0.5275,
+      "step": 8441
+    },
+    {
+      "epoch": 2.532973216295296,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.359542099713813e-05,
+      "loss": 0.5834,
+      "step": 8442
+    },
+    {
+      "epoch": 2.5332733138269936,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.358035848772405e-05,
+      "loss": 0.5927,
+      "step": 8443
+    },
+    {
+      "epoch": 2.5335734113586916,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.356529597831e-05,
+      "loss": 0.5188,
+      "step": 8444
+    },
+    {
+      "epoch": 2.533873508890389,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.355023346889592e-05,
+      "loss": 0.5325,
+      "step": 8445
+    },
+    {
+      "epoch": 2.534173606422087,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.353517095948186e-05,
+      "loss": 0.609,
+      "step": 8446
+    },
+    {
+      "epoch": 2.5344737039537852,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.352010845006779e-05,
+      "loss": 0.5318,
+      "step": 8447
+    },
+    {
+      "epoch": 2.534773801485483,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.350504594065371e-05,
+      "loss": 0.6169,
+      "step": 8448
+    },
+    {
+      "epoch": 2.5350738990171804,
+      "grad_norm": 0.87109375,
+      "learning_rate": 7.348998343123966e-05,
+      "loss": 0.4877,
+      "step": 8449
+    },
+    {
+      "epoch": 2.5353739965488784,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.347492092182558e-05,
+      "loss": 0.5906,
+      "step": 8450
+    },
+    {
+      "epoch": 2.5356740940805764,
+      "grad_norm": 0.953125,
+      "learning_rate": 7.345985841241151e-05,
+      "loss": 0.4843,
+      "step": 8451
+    },
+    {
+      "epoch": 2.535974191612274,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.344479590299745e-05,
+      "loss": 0.602,
+      "step": 8452
+    },
+    {
+      "epoch": 2.5362742891439716,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.342973339358338e-05,
+      "loss": 0.6044,
+      "step": 8453
+    },
+    {
+      "epoch": 2.5365743866756696,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.34146708841693e-05,
+      "loss": 0.5777,
+      "step": 8454
+    },
+    {
+      "epoch": 2.5368744842073676,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.339960837475523e-05,
+      "loss": 0.4934,
+      "step": 8455
+    },
+    {
+      "epoch": 2.537174581739065,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.338454586534117e-05,
+      "loss": 0.6163,
+      "step": 8456
+    },
+    {
+      "epoch": 2.5374746792707628,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.336948335592711e-05,
+      "loss": 0.5357,
+      "step": 8457
+    },
+    {
+      "epoch": 2.5377747768024608,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.335442084651302e-05,
+      "loss": 0.5554,
+      "step": 8458
+    },
+    {
+      "epoch": 2.538074874334159,
+      "grad_norm": 0.9765625,
+      "learning_rate": 7.333935833709896e-05,
+      "loss": 0.5495,
+      "step": 8459
+    },
+    {
+      "epoch": 2.5383749718658564,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.332429582768489e-05,
+      "loss": 0.436,
+      "step": 8460
+    },
+    {
+      "epoch": 2.538675069397554,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.330923331827083e-05,
+      "loss": 0.5673,
+      "step": 8461
+    },
+    {
+      "epoch": 2.538975166929252,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.329417080885676e-05,
+      "loss": 0.5845,
+      "step": 8462
+    },
+    {
+      "epoch": 2.53927526446095,
+      "grad_norm": 1.265625,
+      "learning_rate": 7.327910829944268e-05,
+      "loss": 0.6333,
+      "step": 8463
+    },
+    {
+      "epoch": 2.5395753619926476,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.326404579002862e-05,
+      "loss": 0.6241,
+      "step": 8464
+    },
+    {
+      "epoch": 2.5398754595243456,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.324898328061456e-05,
+      "loss": 0.4843,
+      "step": 8465
+    },
+    {
+      "epoch": 2.540175557056043,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.323392077120048e-05,
+      "loss": 0.5799,
+      "step": 8466
+    },
+    {
+      "epoch": 2.540475654587741,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.321885826178642e-05,
+      "loss": 0.6208,
+      "step": 8467
+    },
+    {
+      "epoch": 2.5407757521194387,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.320379575237234e-05,
+      "loss": 0.5746,
+      "step": 8468
+    },
+    {
+      "epoch": 2.5410758496511368,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.318873324295829e-05,
+      "loss": 0.4365,
+      "step": 8469
+    },
+    {
+      "epoch": 2.5413759471828343,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.317367073354421e-05,
+      "loss": 0.5553,
+      "step": 8470
+    },
+    {
+      "epoch": 2.5416760447145323,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.315860822413014e-05,
+      "loss": 0.5787,
+      "step": 8471
+    },
+    {
+      "epoch": 2.54197614224623,
+      "grad_norm": 1.2734375,
+      "learning_rate": 7.314354571471608e-05,
+      "loss": 0.6138,
+      "step": 8472
+    },
+    {
+      "epoch": 2.542276239777928,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.3128483205302e-05,
+      "loss": 0.4793,
+      "step": 8473
+    },
+    {
+      "epoch": 2.5425763373096255,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.311342069588793e-05,
+      "loss": 0.5642,
+      "step": 8474
+    },
+    {
+      "epoch": 2.5428764348413235,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.309835818647387e-05,
+      "loss": 0.6509,
+      "step": 8475
+    },
+    {
+      "epoch": 2.543176532373021,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.30832956770598e-05,
+      "loss": 0.628,
+      "step": 8476
+    },
+    {
+      "epoch": 2.543476629904719,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.306823316764574e-05,
+      "loss": 0.5594,
+      "step": 8477
+    },
+    {
+      "epoch": 2.5437767274364167,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.305317065823167e-05,
+      "loss": 0.5508,
+      "step": 8478
+    },
+    {
+      "epoch": 2.5440768249681147,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.30381081488176e-05,
+      "loss": 0.5252,
+      "step": 8479
+    },
+    {
+      "epoch": 2.5443769224998123,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.302304563940353e-05,
+      "loss": 0.514,
+      "step": 8480
+    },
+    {
+      "epoch": 2.5446770200315103,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.300798312998946e-05,
+      "loss": 0.5298,
+      "step": 8481
+    },
+    {
+      "epoch": 2.544977117563208,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.299292062057539e-05,
+      "loss": 0.4665,
+      "step": 8482
+    },
+    {
+      "epoch": 2.545277215094906,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.297785811116133e-05,
+      "loss": 0.5677,
+      "step": 8483
+    },
+    {
+      "epoch": 2.5455773126266035,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.296279560174725e-05,
+      "loss": 0.5338,
+      "step": 8484
+    },
+    {
+      "epoch": 2.5458774101583015,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.294773309233318e-05,
+      "loss": 0.549,
+      "step": 8485
+    },
+    {
+      "epoch": 2.5461775076899995,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.293267058291912e-05,
+      "loss": 0.5634,
+      "step": 8486
+    },
+    {
+      "epoch": 2.546477605221697,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.291760807350505e-05,
+      "loss": 0.5653,
+      "step": 8487
+    },
+    {
+      "epoch": 2.5467777027533947,
+      "grad_norm": 1.0,
+      "learning_rate": 7.290254556409099e-05,
+      "loss": 0.5152,
+      "step": 8488
+    },
+    {
+      "epoch": 2.5470778002850927,
+      "grad_norm": 1.125,
+      "learning_rate": 7.28874830546769e-05,
+      "loss": 0.5508,
+      "step": 8489
+    },
+    {
+      "epoch": 2.5473778978167907,
+      "grad_norm": 1.125,
+      "learning_rate": 7.287242054526284e-05,
+      "loss": 0.538,
+      "step": 8490
+    },
+    {
+      "epoch": 2.5476779953484883,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.285735803584878e-05,
+      "loss": 0.5214,
+      "step": 8491
+    },
+    {
+      "epoch": 2.547978092880186,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.284229552643471e-05,
+      "loss": 0.5006,
+      "step": 8492
+    },
+    {
+      "epoch": 2.548278190411884,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.282723301702064e-05,
+      "loss": 0.4546,
+      "step": 8493
+    },
+    {
+      "epoch": 2.548578287943582,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.281217050760658e-05,
+      "loss": 0.6353,
+      "step": 8494
+    },
+    {
+      "epoch": 2.5488783854752795,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.27971079981925e-05,
+      "loss": 0.6331,
+      "step": 8495
+    },
+    {
+      "epoch": 2.549178483006977,
+      "grad_norm": 1.1484375,
+      "learning_rate": 7.278204548877844e-05,
+      "loss": 0.5577,
+      "step": 8496
+    },
+    {
+      "epoch": 2.549478580538675,
+      "grad_norm": 1.125,
+      "learning_rate": 7.276698297936436e-05,
+      "loss": 0.6231,
+      "step": 8497
+    },
+    {
+      "epoch": 2.549778678070373,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.27519204699503e-05,
+      "loss": 0.4784,
+      "step": 8498
+    },
+    {
+      "epoch": 2.5500787756020706,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.273685796053624e-05,
+      "loss": 0.4647,
+      "step": 8499
+    },
+    {
+      "epoch": 2.550378873133768,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.272179545112216e-05,
+      "loss": 0.4761,
+      "step": 8500
+    },
+    {
+      "epoch": 2.5506789706654662,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.270673294170809e-05,
+      "loss": 0.5774,
+      "step": 8501
+    },
+    {
+      "epoch": 2.5509790681971642,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.269167043229402e-05,
+      "loss": 0.5518,
+      "step": 8502
+    },
+    {
+      "epoch": 2.551279165728862,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.267660792287996e-05,
+      "loss": 0.5008,
+      "step": 8503
+    },
+    {
+      "epoch": 2.55157926326056,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.26615454134659e-05,
+      "loss": 0.6536,
+      "step": 8504
+    },
+    {
+      "epoch": 2.5518793607922574,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.264648290405181e-05,
+      "loss": 0.5671,
+      "step": 8505
+    },
+    {
+      "epoch": 2.5521794583239554,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.263142039463775e-05,
+      "loss": 0.6441,
+      "step": 8506
+    },
+    {
+      "epoch": 2.552479555855653,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.261635788522369e-05,
+      "loss": 0.519,
+      "step": 8507
+    },
+    {
+      "epoch": 2.552779653387351,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.26012953758096e-05,
+      "loss": 0.5907,
+      "step": 8508
+    },
+    {
+      "epoch": 2.5530797509190486,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.258623286639555e-05,
+      "loss": 0.4801,
+      "step": 8509
+    },
+    {
+      "epoch": 2.5533798484507466,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.257117035698147e-05,
+      "loss": 0.6136,
+      "step": 8510
+    },
+    {
+      "epoch": 2.553679945982444,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.255610784756741e-05,
+      "loss": 0.5723,
+      "step": 8511
+    },
+    {
+      "epoch": 2.553980043514142,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.254104533815334e-05,
+      "loss": 0.5338,
+      "step": 8512
+    },
+    {
+      "epoch": 2.55428014104584,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.252598282873927e-05,
+      "loss": 0.5561,
+      "step": 8513
+    },
+    {
+      "epoch": 2.554580238577538,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.25109203193252e-05,
+      "loss": 0.5596,
+      "step": 8514
+    },
+    {
+      "epoch": 2.5548803361092354,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.249585780991113e-05,
+      "loss": 0.5588,
+      "step": 8515
+    },
+    {
+      "epoch": 2.5551804336409334,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.248079530049706e-05,
+      "loss": 0.5781,
+      "step": 8516
+    },
+    {
+      "epoch": 2.555480531172631,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.2465732791083e-05,
+      "loss": 0.4498,
+      "step": 8517
+    },
+    {
+      "epoch": 2.555780628704329,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.245067028166893e-05,
+      "loss": 0.4941,
+      "step": 8518
+    },
+    {
+      "epoch": 2.5560807262360266,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.243560777225487e-05,
+      "loss": 0.5374,
+      "step": 8519
+    },
+    {
+      "epoch": 2.5563808237677246,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.24205452628408e-05,
+      "loss": 0.6107,
+      "step": 8520
+    },
+    {
+      "epoch": 2.556680921299422,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.240548275342672e-05,
+      "loss": 0.6127,
+      "step": 8521
+    },
+    {
+      "epoch": 2.55698101883112,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.239042024401266e-05,
+      "loss": 0.5866,
+      "step": 8522
+    },
+    {
+      "epoch": 2.5572811163628177,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.237535773459859e-05,
+      "loss": 0.6535,
+      "step": 8523
+    },
+    {
+      "epoch": 2.5575812138945158,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.236029522518451e-05,
+      "loss": 0.5458,
+      "step": 8524
+    },
+    {
+      "epoch": 2.5578813114262138,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.234523271577045e-05,
+      "loss": 0.5642,
+      "step": 8525
+    },
+    {
+      "epoch": 2.5581814089579114,
+      "grad_norm": 1.125,
+      "learning_rate": 7.233017020635638e-05,
+      "loss": 0.5637,
+      "step": 8526
+    },
+    {
+      "epoch": 2.558481506489609,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.231510769694232e-05,
+      "loss": 0.5302,
+      "step": 8527
+    },
+    {
+      "epoch": 2.558781604021307,
+      "grad_norm": 1.234375,
+      "learning_rate": 7.230004518752825e-05,
+      "loss": 0.5004,
+      "step": 8528
+    },
+    {
+      "epoch": 2.559081701553005,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.228498267811417e-05,
+      "loss": 0.606,
+      "step": 8529
+    },
+    {
+      "epoch": 2.5593817990847025,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.226992016870012e-05,
+      "loss": 0.5814,
+      "step": 8530
+    },
+    {
+      "epoch": 2.5596818966164,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.225485765928604e-05,
+      "loss": 0.5689,
+      "step": 8531
+    },
+    {
+      "epoch": 2.559981994148098,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.223979514987197e-05,
+      "loss": 0.4757,
+      "step": 8532
+    },
+    {
+      "epoch": 2.560282091679796,
+      "grad_norm": 0.90234375,
+      "learning_rate": 7.222473264045791e-05,
+      "loss": 0.5218,
+      "step": 8533
+    },
+    {
+      "epoch": 2.5605821892114937,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.220967013104384e-05,
+      "loss": 0.4637,
+      "step": 8534
+    },
+    {
+      "epoch": 2.5608822867431913,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.219460762162976e-05,
+      "loss": 0.5314,
+      "step": 8535
+    },
+    {
+      "epoch": 2.5611823842748893,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.21795451122157e-05,
+      "loss": 0.6199,
+      "step": 8536
+    },
+    {
+      "epoch": 2.5614824818065873,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.216448260280163e-05,
+      "loss": 0.5988,
+      "step": 8537
+    },
+    {
+      "epoch": 2.561782579338285,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.214942009338757e-05,
+      "loss": 0.5071,
+      "step": 8538
+    },
+    {
+      "epoch": 2.5620826768699825,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.213435758397348e-05,
+      "loss": 0.5076,
+      "step": 8539
+    },
+    {
+      "epoch": 2.5623827744016805,
+      "grad_norm": 1.0,
+      "learning_rate": 7.211929507455942e-05,
+      "loss": 0.6217,
+      "step": 8540
+    },
+    {
+      "epoch": 2.5626828719333785,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.210423256514536e-05,
+      "loss": 0.4829,
+      "step": 8541
+    },
+    {
+      "epoch": 2.562982969465076,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.208917005573129e-05,
+      "loss": 0.5406,
+      "step": 8542
+    },
+    {
+      "epoch": 2.563283066996774,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.207410754631722e-05,
+      "loss": 0.6032,
+      "step": 8543
+    },
+    {
+      "epoch": 2.5635831645284717,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.205904503690314e-05,
+      "loss": 0.5368,
+      "step": 8544
+    },
+    {
+      "epoch": 2.5638832620601697,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.204398252748908e-05,
+      "loss": 0.6293,
+      "step": 8545
+    },
+    {
+      "epoch": 2.5641833595918673,
+      "grad_norm": 1.3046875,
+      "learning_rate": 7.202892001807502e-05,
+      "loss": 0.5834,
+      "step": 8546
+    },
+    {
+      "epoch": 2.5644834571235653,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.201385750866094e-05,
+      "loss": 0.5271,
+      "step": 8547
+    },
+    {
+      "epoch": 2.564783554655263,
+      "grad_norm": 1.125,
+      "learning_rate": 7.199879499924688e-05,
+      "loss": 0.5905,
+      "step": 8548
+    },
+    {
+      "epoch": 2.565083652186961,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.19837324898328e-05,
+      "loss": 0.5898,
+      "step": 8549
+    },
+    {
+      "epoch": 2.5653837497186585,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.196866998041875e-05,
+      "loss": 0.5147,
+      "step": 8550
+    },
+    {
+      "epoch": 2.5656838472503565,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.195360747100467e-05,
+      "loss": 0.5302,
+      "step": 8551
+    },
+    {
+      "epoch": 2.565983944782054,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.19385449615906e-05,
+      "loss": 0.5681,
+      "step": 8552
+    },
+    {
+      "epoch": 2.566284042313752,
+      "grad_norm": 1.2421875,
+      "learning_rate": 7.192348245217654e-05,
+      "loss": 0.5758,
+      "step": 8553
+    },
+    {
+      "epoch": 2.5665841398454496,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.190841994276248e-05,
+      "loss": 0.4797,
+      "step": 8554
+    },
+    {
+      "epoch": 2.5668842373771477,
+      "grad_norm": 1.25,
+      "learning_rate": 7.189335743334839e-05,
+      "loss": 0.5812,
+      "step": 8555
+    },
+    {
+      "epoch": 2.5671843349088452,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.187829492393433e-05,
+      "loss": 0.5504,
+      "step": 8556
+    },
+    {
+      "epoch": 2.5674844324405433,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.186323241452026e-05,
+      "loss": 0.4973,
+      "step": 8557
+    },
+    {
+      "epoch": 2.567784529972241,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.18481699051062e-05,
+      "loss": 0.5117,
+      "step": 8558
+    },
+    {
+      "epoch": 2.568084627503939,
+      "grad_norm": 0.88671875,
+      "learning_rate": 7.183310739569213e-05,
+      "loss": 0.5598,
+      "step": 8559
+    },
+    {
+      "epoch": 2.5683847250356364,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.181804488627805e-05,
+      "loss": 0.5285,
+      "step": 8560
+    },
+    {
+      "epoch": 2.5686848225673344,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.1802982376864e-05,
+      "loss": 0.5731,
+      "step": 8561
+    },
+    {
+      "epoch": 2.568984920099032,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.178791986744992e-05,
+      "loss": 0.5311,
+      "step": 8562
+    },
+    {
+      "epoch": 2.56928501763073,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.177285735803585e-05,
+      "loss": 0.4746,
+      "step": 8563
+    },
+    {
+      "epoch": 2.569585115162428,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.175779484862179e-05,
+      "loss": 0.5415,
+      "step": 8564
+    },
+    {
+      "epoch": 2.5698852126941256,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.174273233920771e-05,
+      "loss": 0.5834,
+      "step": 8565
+    },
+    {
+      "epoch": 2.570185310225823,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.172766982979364e-05,
+      "loss": 0.5367,
+      "step": 8566
+    },
+    {
+      "epoch": 2.570485407757521,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.171260732037958e-05,
+      "loss": 0.5735,
+      "step": 8567
+    },
+    {
+      "epoch": 2.5707855052892192,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.169754481096551e-05,
+      "loss": 0.5393,
+      "step": 8568
+    },
+    {
+      "epoch": 2.571085602820917,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.168248230155145e-05,
+      "loss": 0.5105,
+      "step": 8569
+    },
+    {
+      "epoch": 2.5713857003526144,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.166741979213738e-05,
+      "loss": 0.6001,
+      "step": 8570
+    },
+    {
+      "epoch": 2.5716857978843124,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.16523572827233e-05,
+      "loss": 0.5354,
+      "step": 8571
+    },
+    {
+      "epoch": 2.5719858954160104,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.163729477330924e-05,
+      "loss": 0.6086,
+      "step": 8572
+    },
+    {
+      "epoch": 2.572285992947708,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.162223226389517e-05,
+      "loss": 0.6049,
+      "step": 8573
+    },
+    {
+      "epoch": 2.5725860904794056,
+      "grad_norm": 1.09375,
+      "learning_rate": 7.16071697544811e-05,
+      "loss": 0.6561,
+      "step": 8574
+    },
+    {
+      "epoch": 2.5728861880111036,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.159210724506704e-05,
+      "loss": 0.59,
+      "step": 8575
+    },
+    {
+      "epoch": 2.5731862855428016,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.157704473565296e-05,
+      "loss": 0.4908,
+      "step": 8576
+    },
+    {
+      "epoch": 2.573486383074499,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.15619822262389e-05,
+      "loss": 0.4941,
+      "step": 8577
+    },
+    {
+      "epoch": 2.5737864806061967,
+      "grad_norm": 1.125,
+      "learning_rate": 7.154691971682482e-05,
+      "loss": 0.4881,
+      "step": 8578
+    },
+    {
+      "epoch": 2.5740865781378948,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.153185720741076e-05,
+      "loss": 0.5235,
+      "step": 8579
+    },
+    {
+      "epoch": 2.574386675669593,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.15167946979967e-05,
+      "loss": 0.5545,
+      "step": 8580
+    },
+    {
+      "epoch": 2.5746867732012904,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.150173218858262e-05,
+      "loss": 0.5383,
+      "step": 8581
+    },
+    {
+      "epoch": 2.5749868707329884,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.148666967916855e-05,
+      "loss": 0.5634,
+      "step": 8582
+    },
+    {
+      "epoch": 2.575286968264686,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.147160716975449e-05,
+      "loss": 0.6583,
+      "step": 8583
+    },
+    {
+      "epoch": 2.575587065796384,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.145654466034042e-05,
+      "loss": 0.4924,
+      "step": 8584
+    },
+    {
+      "epoch": 2.5758871633280815,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.144148215092636e-05,
+      "loss": 0.5212,
+      "step": 8585
+    },
+    {
+      "epoch": 2.5761872608597796,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.142641964151227e-05,
+      "loss": 0.5253,
+      "step": 8586
+    },
+    {
+      "epoch": 2.576487358391477,
+      "grad_norm": 1.21875,
+      "learning_rate": 7.141135713209821e-05,
+      "loss": 0.5779,
+      "step": 8587
+    },
+    {
+      "epoch": 2.576787455923175,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.139629462268415e-05,
+      "loss": 0.4934,
+      "step": 8588
+    },
+    {
+      "epoch": 2.5770875534548727,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.138123211327006e-05,
+      "loss": 0.5096,
+      "step": 8589
+    },
+    {
+      "epoch": 2.5773876509865707,
+      "grad_norm": 1.1953125,
+      "learning_rate": 7.1366169603856e-05,
+      "loss": 0.6214,
+      "step": 8590
+    },
+    {
+      "epoch": 2.5776877485182683,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.135110709444193e-05,
+      "loss": 0.4985,
+      "step": 8591
+    },
+    {
+      "epoch": 2.5779878460499663,
+      "grad_norm": 1.2265625,
+      "learning_rate": 7.133604458502787e-05,
+      "loss": 0.6485,
+      "step": 8592
+    },
+    {
+      "epoch": 2.578287943581664,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.13209820756138e-05,
+      "loss": 0.5367,
+      "step": 8593
+    },
+    {
+      "epoch": 2.578588041113362,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.130591956619973e-05,
+      "loss": 0.548,
+      "step": 8594
+    },
+    {
+      "epoch": 2.5788881386450595,
+      "grad_norm": 0.91796875,
+      "learning_rate": 7.129085705678567e-05,
+      "loss": 0.5245,
+      "step": 8595
+    },
+    {
+      "epoch": 2.5791882361767575,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.127579454737159e-05,
+      "loss": 0.5075,
+      "step": 8596
+    },
+    {
+      "epoch": 2.579488333708455,
+      "grad_norm": 1.015625,
+      "learning_rate": 7.126073203795752e-05,
+      "loss": 0.4757,
+      "step": 8597
+    },
+    {
+      "epoch": 2.579788431240153,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.124566952854346e-05,
+      "loss": 0.608,
+      "step": 8598
+    },
+    {
+      "epoch": 2.5800885287718507,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.123060701912939e-05,
+      "loss": 0.5166,
+      "step": 8599
+    },
+    {
+      "epoch": 2.5803886263035487,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.121554450971533e-05,
+      "loss": 0.5901,
+      "step": 8600
+    },
+    {
+      "epoch": 2.5806887238352463,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.120048200030125e-05,
+      "loss": 0.614,
+      "step": 8601
+    },
+    {
+      "epoch": 2.5809888213669443,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.118541949088718e-05,
+      "loss": 0.4755,
+      "step": 8602
+    },
+    {
+      "epoch": 2.5812889188986423,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.117035698147312e-05,
+      "loss": 0.6205,
+      "step": 8603
+    },
+    {
+      "epoch": 2.58158901643034,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.115529447205905e-05,
+      "loss": 0.5577,
+      "step": 8604
+    },
+    {
+      "epoch": 2.5818891139620375,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.114023196264497e-05,
+      "loss": 0.5202,
+      "step": 8605
+    },
+    {
+      "epoch": 2.5821892114937355,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.112516945323091e-05,
+      "loss": 0.4648,
+      "step": 8606
+    },
+    {
+      "epoch": 2.5824893090254335,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.111010694381684e-05,
+      "loss": 0.5855,
+      "step": 8607
+    },
+    {
+      "epoch": 2.582789406557131,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.109504443440278e-05,
+      "loss": 0.525,
+      "step": 8608
+    },
+    {
+      "epoch": 2.5830895040888286,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.107998192498871e-05,
+      "loss": 0.6916,
+      "step": 8609
+    },
+    {
+      "epoch": 2.5833896016205267,
+      "grad_norm": 0.93359375,
+      "learning_rate": 7.106491941557463e-05,
+      "loss": 0.4487,
+      "step": 8610
+    },
+    {
+      "epoch": 2.5836896991522247,
+      "grad_norm": 1.125,
+      "learning_rate": 7.104985690616058e-05,
+      "loss": 0.6854,
+      "step": 8611
+    },
+    {
+      "epoch": 2.5839897966839223,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.10347943967465e-05,
+      "loss": 0.49,
+      "step": 8612
+    },
+    {
+      "epoch": 2.58428989421562,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.101973188733243e-05,
+      "loss": 0.5477,
+      "step": 8613
+    },
+    {
+      "epoch": 2.584589991747318,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.100466937791837e-05,
+      "loss": 0.5495,
+      "step": 8614
+    },
+    {
+      "epoch": 2.584890089279016,
+      "grad_norm": 1.125,
+      "learning_rate": 7.09896068685043e-05,
+      "loss": 0.5776,
+      "step": 8615
+    },
+    {
+      "epoch": 2.5851901868107134,
+      "grad_norm": 1.125,
+      "learning_rate": 7.097454435909022e-05,
+      "loss": 0.6593,
+      "step": 8616
+    },
+    {
+      "epoch": 2.585490284342411,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.095948184967616e-05,
+      "loss": 0.4197,
+      "step": 8617
+    },
+    {
+      "epoch": 2.585790381874109,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.094441934026209e-05,
+      "loss": 0.4878,
+      "step": 8618
+    },
+    {
+      "epoch": 2.586090479405807,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.092935683084803e-05,
+      "loss": 0.4985,
+      "step": 8619
+    },
+    {
+      "epoch": 2.5863905769375046,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.091429432143394e-05,
+      "loss": 0.5554,
+      "step": 8620
+    },
+    {
+      "epoch": 2.5866906744692026,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.089923181201988e-05,
+      "loss": 0.4612,
+      "step": 8621
+    },
+    {
+      "epoch": 2.5869907720009,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.088416930260582e-05,
+      "loss": 0.6253,
+      "step": 8622
+    },
+    {
+      "epoch": 2.5872908695325982,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.086910679319175e-05,
+      "loss": 0.5029,
+      "step": 8623
+    },
+    {
+      "epoch": 2.587590967064296,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.085404428377768e-05,
+      "loss": 0.5241,
+      "step": 8624
+    },
+    {
+      "epoch": 2.587891064595994,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.083898177436362e-05,
+      "loss": 0.595,
+      "step": 8625
+    },
+    {
+      "epoch": 2.5881911621276914,
+      "grad_norm": 1.078125,
+      "learning_rate": 7.082391926494954e-05,
+      "loss": 0.5421,
+      "step": 8626
+    },
+    {
+      "epoch": 2.5884912596593894,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.080885675553548e-05,
+      "loss": 0.5843,
+      "step": 8627
+    },
+    {
+      "epoch": 2.588791357191087,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.07937942461214e-05,
+      "loss": 0.5245,
+      "step": 8628
+    },
+    {
+      "epoch": 2.589091454722785,
+      "grad_norm": 1.125,
+      "learning_rate": 7.077873173670734e-05,
+      "loss": 0.5292,
+      "step": 8629
+    },
+    {
+      "epoch": 2.5893915522544826,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.076366922729328e-05,
+      "loss": 0.5951,
+      "step": 8630
+    },
+    {
+      "epoch": 2.5896916497861806,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.07486067178792e-05,
+      "loss": 0.5266,
+      "step": 8631
+    },
+    {
+      "epoch": 2.589991747317878,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.073354420846513e-05,
+      "loss": 0.5267,
+      "step": 8632
+    },
+    {
+      "epoch": 2.590291844849576,
+      "grad_norm": 0.875,
+      "learning_rate": 7.071848169905106e-05,
+      "loss": 0.4335,
+      "step": 8633
+    },
+    {
+      "epoch": 2.5905919423812738,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.0703419189637e-05,
+      "loss": 0.5702,
+      "step": 8634
+    },
+    {
+      "epoch": 2.590892039912972,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.068835668022294e-05,
+      "loss": 0.5262,
+      "step": 8635
+    },
+    {
+      "epoch": 2.5911921374446694,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.067329417080885e-05,
+      "loss": 0.4971,
+      "step": 8636
+    },
+    {
+      "epoch": 2.5914922349763674,
+      "grad_norm": 1.2109375,
+      "learning_rate": 7.065823166139479e-05,
+      "loss": 0.5675,
+      "step": 8637
+    },
+    {
+      "epoch": 2.591792332508065,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.064316915198072e-05,
+      "loss": 0.5618,
+      "step": 8638
+    },
+    {
+      "epoch": 2.592092430039763,
+      "grad_norm": 0.9296875,
+      "learning_rate": 7.062810664256666e-05,
+      "loss": 0.5698,
+      "step": 8639
+    },
+    {
+      "epoch": 2.5923925275714605,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.061304413315259e-05,
+      "loss": 0.5457,
+      "step": 8640
+    },
+    {
+      "epoch": 2.5926926251031586,
+      "grad_norm": 1.140625,
+      "learning_rate": 7.059798162373851e-05,
+      "loss": 0.5634,
+      "step": 8641
+    },
+    {
+      "epoch": 2.5929927226348566,
+      "grad_norm": 1.1171875,
+      "learning_rate": 7.058291911432445e-05,
+      "loss": 0.6232,
+      "step": 8642
+    },
+    {
+      "epoch": 2.593292820166554,
+      "grad_norm": 1.0078125,
+      "learning_rate": 7.056785660491038e-05,
+      "loss": 0.5502,
+      "step": 8643
+    },
+    {
+      "epoch": 2.5935929176982517,
+      "grad_norm": 1.15625,
+      "learning_rate": 7.055279409549631e-05,
+      "loss": 0.5184,
+      "step": 8644
+    },
+    {
+      "epoch": 2.5938930152299498,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.053773158608225e-05,
+      "loss": 0.5164,
+      "step": 8645
+    },
+    {
+      "epoch": 2.5941931127616478,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.052266907666817e-05,
+      "loss": 0.6044,
+      "step": 8646
+    },
+    {
+      "epoch": 2.5944932102933453,
+      "grad_norm": 1.1796875,
+      "learning_rate": 7.05076065672541e-05,
+      "loss": 0.5308,
+      "step": 8647
+    },
+    {
+      "epoch": 2.594793307825043,
+      "grad_norm": 1.109375,
+      "learning_rate": 7.049254405784004e-05,
+      "loss": 0.5586,
+      "step": 8648
+    },
+    {
+      "epoch": 2.595093405356741,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.047748154842597e-05,
+      "loss": 0.6095,
+      "step": 8649
+    },
+    {
+      "epoch": 2.595393502888439,
+      "grad_norm": 1.1875,
+      "learning_rate": 7.046241903901191e-05,
+      "loss": 0.6769,
+      "step": 8650
+    },
+    {
+      "epoch": 2.5956936004201365,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.044735652959784e-05,
+      "loss": 0.5417,
+      "step": 8651
+    },
+    {
+      "epoch": 2.595993697951834,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.043229402018376e-05,
+      "loss": 0.5753,
+      "step": 8652
+    },
+    {
+      "epoch": 2.596293795483532,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.04172315107697e-05,
+      "loss": 0.5232,
+      "step": 8653
+    },
+    {
+      "epoch": 2.59659389301523,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.040216900135563e-05,
+      "loss": 0.4718,
+      "step": 8654
+    },
+    {
+      "epoch": 2.5968939905469277,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.038710649194156e-05,
+      "loss": 0.4829,
+      "step": 8655
+    },
+    {
+      "epoch": 2.5971940880786253,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.03720439825275e-05,
+      "loss": 0.4923,
+      "step": 8656
+    },
+    {
+      "epoch": 2.5974941856103233,
+      "grad_norm": 1.046875,
+      "learning_rate": 7.035698147311342e-05,
+      "loss": 0.526,
+      "step": 8657
+    },
+    {
+      "epoch": 2.5977942831420213,
+      "grad_norm": 1.125,
+      "learning_rate": 7.034191896369936e-05,
+      "loss": 0.5763,
+      "step": 8658
+    },
+    {
+      "epoch": 2.598094380673719,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.032685645428529e-05,
+      "loss": 0.5857,
+      "step": 8659
+    },
+    {
+      "epoch": 2.598394478205417,
+      "grad_norm": 1.0625,
+      "learning_rate": 7.031179394487122e-05,
+      "loss": 0.5408,
+      "step": 8660
+    },
+    {
+      "epoch": 2.5986945757371145,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.029673143545716e-05,
+      "loss": 0.5704,
+      "step": 8661
+    },
+    {
+      "epoch": 2.5989946732688125,
+      "grad_norm": 0.99609375,
+      "learning_rate": 7.028166892604308e-05,
+      "loss": 0.5323,
+      "step": 8662
+    },
+    {
+      "epoch": 2.59929477080051,
+      "grad_norm": 1.0,
+      "learning_rate": 7.026660641662901e-05,
+      "loss": 0.489,
+      "step": 8663
+    },
+    {
+      "epoch": 2.599594868332208,
+      "grad_norm": 1.03125,
+      "learning_rate": 7.025154390721495e-05,
+      "loss": 0.5391,
+      "step": 8664
+    },
+    {
+      "epoch": 2.5998949658639057,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.023648139780088e-05,
+      "loss": 0.5178,
+      "step": 8665
+    },
+    {
+      "epoch": 2.6001950633956037,
+      "grad_norm": 0.96875,
+      "learning_rate": 7.022141888838682e-05,
+      "loss": 0.5402,
+      "step": 8666
+    },
+    {
+      "epoch": 2.6004951609273013,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.020635637897273e-05,
+      "loss": 0.5541,
+      "step": 8667
+    },
+    {
+      "epoch": 2.6007952584589993,
+      "grad_norm": 1.125,
+      "learning_rate": 7.019129386955867e-05,
+      "loss": 0.54,
+      "step": 8668
+    },
+    {
+      "epoch": 2.601095355990697,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.017623136014461e-05,
+      "loss": 0.5852,
+      "step": 8669
+    },
+    {
+      "epoch": 2.601395453522395,
+      "grad_norm": 1.1015625,
+      "learning_rate": 7.016116885073052e-05,
+      "loss": 0.5514,
+      "step": 8670
+    },
+    {
+      "epoch": 2.6016955510540924,
+      "grad_norm": 0.98046875,
+      "learning_rate": 7.014610634131647e-05,
+      "loss": 0.6204,
+      "step": 8671
+    },
+    {
+      "epoch": 2.6019956485857905,
+      "grad_norm": 0.9609375,
+      "learning_rate": 7.01310438319024e-05,
+      "loss": 0.5779,
+      "step": 8672
+    },
+    {
+      "epoch": 2.602295746117488,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.011598132248833e-05,
+      "loss": 0.4943,
+      "step": 8673
+    },
+    {
+      "epoch": 2.602595843649186,
+      "grad_norm": 1.171875,
+      "learning_rate": 7.010091881307426e-05,
+      "loss": 0.6298,
+      "step": 8674
+    },
+    {
+      "epoch": 2.6028959411808836,
+      "grad_norm": 1.0859375,
+      "learning_rate": 7.008585630366019e-05,
+      "loss": 0.4789,
+      "step": 8675
+    },
+    {
+      "epoch": 2.6031960387125817,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.007079379424613e-05,
+      "loss": 0.5876,
+      "step": 8676
+    },
+    {
+      "epoch": 2.6034961362442792,
+      "grad_norm": 1.1328125,
+      "learning_rate": 7.005573128483207e-05,
+      "loss": 0.4932,
+      "step": 8677
+    },
+    {
+      "epoch": 2.6037962337759772,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.004066877541798e-05,
+      "loss": 0.4535,
+      "step": 8678
+    },
+    {
+      "epoch": 2.604096331307675,
+      "grad_norm": 1.0,
+      "learning_rate": 7.002560626600392e-05,
+      "loss": 0.5209,
+      "step": 8679
+    },
+    {
+      "epoch": 2.604396428839373,
+      "grad_norm": 1.0234375,
+      "learning_rate": 7.001054375658985e-05,
+      "loss": 0.4388,
+      "step": 8680
+    },
+    {
+      "epoch": 2.604696526371071,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.999548124717579e-05,
+      "loss": 0.6,
+      "step": 8681
+    },
+    {
+      "epoch": 2.6049966239027684,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.998041873776171e-05,
+      "loss": 0.5304,
+      "step": 8682
+    },
+    {
+      "epoch": 2.605296721434466,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.996535622834764e-05,
+      "loss": 0.5558,
+      "step": 8683
+    },
+    {
+      "epoch": 2.605596818966164,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.995029371893358e-05,
+      "loss": 0.5746,
+      "step": 8684
+    },
+    {
+      "epoch": 2.605896916497862,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.993523120951951e-05,
+      "loss": 0.6367,
+      "step": 8685
+    },
+    {
+      "epoch": 2.6061970140295596,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.992016870010543e-05,
+      "loss": 0.5643,
+      "step": 8686
+    },
+    {
+      "epoch": 2.606497111561257,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.990510619069137e-05,
+      "loss": 0.4909,
+      "step": 8687
+    },
+    {
+      "epoch": 2.606797209092955,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.98900436812773e-05,
+      "loss": 0.5577,
+      "step": 8688
+    },
+    {
+      "epoch": 2.607097306624653,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.987498117186324e-05,
+      "loss": 0.4898,
+      "step": 8689
+    },
+    {
+      "epoch": 2.607397404156351,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.985991866244917e-05,
+      "loss": 0.5468,
+      "step": 8690
+    },
+    {
+      "epoch": 2.6076975016880484,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.98448561530351e-05,
+      "loss": 0.6713,
+      "step": 8691
+    },
+    {
+      "epoch": 2.6079975992197464,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.982979364362104e-05,
+      "loss": 0.5327,
+      "step": 8692
+    },
+    {
+      "epoch": 2.6082976967514444,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.981473113420696e-05,
+      "loss": 0.5263,
+      "step": 8693
+    },
+    {
+      "epoch": 2.608597794283142,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.979966862479289e-05,
+      "loss": 0.5046,
+      "step": 8694
+    },
+    {
+      "epoch": 2.6088978918148396,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.978460611537883e-05,
+      "loss": 0.4785,
+      "step": 8695
+    },
+    {
+      "epoch": 2.6091979893465376,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.976954360596476e-05,
+      "loss": 0.5583,
+      "step": 8696
+    },
+    {
+      "epoch": 2.6094980868782356,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.975448109655068e-05,
+      "loss": 0.5663,
+      "step": 8697
+    },
+    {
+      "epoch": 2.609798184409933,
+      "grad_norm": 1.28125,
+      "learning_rate": 6.973941858713662e-05,
+      "loss": 0.6745,
+      "step": 8698
+    },
+    {
+      "epoch": 2.610098281941631,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.972435607772255e-05,
+      "loss": 0.5858,
+      "step": 8699
+    },
+    {
+      "epoch": 2.6103983794733288,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.970929356830849e-05,
+      "loss": 0.6457,
+      "step": 8700
+    },
+    {
+      "epoch": 2.6106984770050268,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.969423105889442e-05,
+      "loss": 0.5235,
+      "step": 8701
+    },
+    {
+      "epoch": 2.6109985745367243,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.967916854948034e-05,
+      "loss": 0.5308,
+      "step": 8702
+    },
+    {
+      "epoch": 2.6112986720684224,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.966410604006628e-05,
+      "loss": 0.4622,
+      "step": 8703
+    },
+    {
+      "epoch": 2.61159876960012,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.964904353065221e-05,
+      "loss": 0.6435,
+      "step": 8704
+    },
+    {
+      "epoch": 2.611898867131818,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.963398102123814e-05,
+      "loss": 0.64,
+      "step": 8705
+    },
+    {
+      "epoch": 2.6121989646635155,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.961891851182408e-05,
+      "loss": 0.599,
+      "step": 8706
+    },
+    {
+      "epoch": 2.6124990621952136,
+      "grad_norm": 1.0,
+      "learning_rate": 6.960385600241e-05,
+      "loss": 0.5312,
+      "step": 8707
+    },
+    {
+      "epoch": 2.612799159726911,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.958879349299594e-05,
+      "loss": 0.4872,
+      "step": 8708
+    },
+    {
+      "epoch": 2.613099257258609,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.957373098358186e-05,
+      "loss": 0.4413,
+      "step": 8709
+    },
+    {
+      "epoch": 2.6133993547903067,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.95586684741678e-05,
+      "loss": 0.4907,
+      "step": 8710
+    },
+    {
+      "epoch": 2.6136994523220047,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.954360596475374e-05,
+      "loss": 0.5341,
+      "step": 8711
+    },
+    {
+      "epoch": 2.6139995498537023,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.952854345533967e-05,
+      "loss": 0.5684,
+      "step": 8712
+    },
+    {
+      "epoch": 2.6142996473854003,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.951348094592559e-05,
+      "loss": 0.6071,
+      "step": 8713
+    },
+    {
+      "epoch": 2.614599744917098,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.949841843651152e-05,
+      "loss": 0.501,
+      "step": 8714
+    },
+    {
+      "epoch": 2.614899842448796,
+      "grad_norm": 1.125,
+      "learning_rate": 6.948335592709746e-05,
+      "loss": 0.5714,
+      "step": 8715
+    },
+    {
+      "epoch": 2.615199939980494,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.94682934176834e-05,
+      "loss": 0.4657,
+      "step": 8716
+    },
+    {
+      "epoch": 2.6155000375121915,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.945323090826931e-05,
+      "loss": 0.6028,
+      "step": 8717
+    },
+    {
+      "epoch": 2.615800135043889,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.943816839885525e-05,
+      "loss": 0.5124,
+      "step": 8718
+    },
+    {
+      "epoch": 2.616100232575587,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.942310588944119e-05,
+      "loss": 0.5707,
+      "step": 8719
+    },
+    {
+      "epoch": 2.616400330107285,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.940804338002712e-05,
+      "loss": 0.5969,
+      "step": 8720
+    },
+    {
+      "epoch": 2.6167004276389827,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.939298087061305e-05,
+      "loss": 0.5397,
+      "step": 8721
+    },
+    {
+      "epoch": 2.6170005251706803,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.937791836119897e-05,
+      "loss": 0.5625,
+      "step": 8722
+    },
+    {
+      "epoch": 2.6173006227023783,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.936285585178491e-05,
+      "loss": 0.558,
+      "step": 8723
+    },
+    {
+      "epoch": 2.6176007202340763,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.934779334237084e-05,
+      "loss": 0.6127,
+      "step": 8724
+    },
+    {
+      "epoch": 2.617900817765774,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.933273083295677e-05,
+      "loss": 0.5917,
+      "step": 8725
+    },
+    {
+      "epoch": 2.6182009152974715,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.931766832354271e-05,
+      "loss": 0.6333,
+      "step": 8726
+    },
+    {
+      "epoch": 2.6185010128291695,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.930260581412863e-05,
+      "loss": 0.5755,
+      "step": 8727
+    },
+    {
+      "epoch": 2.6188011103608675,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.928754330471456e-05,
+      "loss": 0.4963,
+      "step": 8728
+    },
+    {
+      "epoch": 2.619101207892565,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.92724807953005e-05,
+      "loss": 0.5163,
+      "step": 8729
+    },
+    {
+      "epoch": 2.6194013054242626,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.925741828588643e-05,
+      "loss": 0.546,
+      "step": 8730
+    },
+    {
+      "epoch": 2.6197014029559607,
+      "grad_norm": 1.125,
+      "learning_rate": 6.924235577647237e-05,
+      "loss": 0.5956,
+      "step": 8731
+    },
+    {
+      "epoch": 2.6200015004876587,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.92272932670583e-05,
+      "loss": 0.5306,
+      "step": 8732
+    },
+    {
+      "epoch": 2.6203015980193562,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.921223075764422e-05,
+      "loss": 0.5902,
+      "step": 8733
+    },
+    {
+      "epoch": 2.620601695551054,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.919716824823016e-05,
+      "loss": 0.4895,
+      "step": 8734
+    },
+    {
+      "epoch": 2.620901793082752,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.918210573881609e-05,
+      "loss": 0.5557,
+      "step": 8735
+    },
+    {
+      "epoch": 2.62120189061445,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.916704322940202e-05,
+      "loss": 0.6653,
+      "step": 8736
+    },
+    {
+      "epoch": 2.6215019881461474,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.915198071998796e-05,
+      "loss": 0.5036,
+      "step": 8737
+    },
+    {
+      "epoch": 2.6218020856778455,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.913691821057388e-05,
+      "loss": 0.5876,
+      "step": 8738
+    },
+    {
+      "epoch": 2.622102183209543,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.912185570115982e-05,
+      "loss": 0.5162,
+      "step": 8739
+    },
+    {
+      "epoch": 2.622402280741241,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.910679319174575e-05,
+      "loss": 0.4796,
+      "step": 8740
+    },
+    {
+      "epoch": 2.6227023782729386,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.909173068233168e-05,
+      "loss": 0.6701,
+      "step": 8741
+    },
+    {
+      "epoch": 2.6230024758046366,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.907666817291762e-05,
+      "loss": 0.4646,
+      "step": 8742
+    },
+    {
+      "epoch": 2.623302573336334,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.906160566350354e-05,
+      "loss": 0.5332,
+      "step": 8743
+    },
+    {
+      "epoch": 2.6236026708680322,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.904654315408947e-05,
+      "loss": 0.6338,
+      "step": 8744
+    },
+    {
+      "epoch": 2.62390276839973,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.903148064467541e-05,
+      "loss": 0.4745,
+      "step": 8745
+    },
+    {
+      "epoch": 2.624202865931428,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.901641813526134e-05,
+      "loss": 0.5639,
+      "step": 8746
+    },
+    {
+      "epoch": 2.6245029634631254,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.900135562584728e-05,
+      "loss": 0.5349,
+      "step": 8747
+    },
+    {
+      "epoch": 2.6248030609948234,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.89862931164332e-05,
+      "loss": 0.5415,
+      "step": 8748
+    },
+    {
+      "epoch": 2.625103158526521,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.897123060701913e-05,
+      "loss": 0.5434,
+      "step": 8749
+    },
+    {
+      "epoch": 2.625403256058219,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.895616809760507e-05,
+      "loss": 0.4523,
+      "step": 8750
+    },
+    {
+      "epoch": 2.6257033535899166,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.894110558819098e-05,
+      "loss": 0.5889,
+      "step": 8751
+    },
+    {
+      "epoch": 2.6260034511216146,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.892604307877693e-05,
+      "loss": 0.5547,
+      "step": 8752
+    },
+    {
+      "epoch": 2.626303548653312,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.891098056936287e-05,
+      "loss": 0.5922,
+      "step": 8753
+    },
+    {
+      "epoch": 2.62660364618501,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.889591805994879e-05,
+      "loss": 0.5354,
+      "step": 8754
+    },
+    {
+      "epoch": 2.626903743716708,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.888085555053472e-05,
+      "loss": 0.5555,
+      "step": 8755
+    },
+    {
+      "epoch": 2.627203841248406,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.886579304112065e-05,
+      "loss": 0.5547,
+      "step": 8756
+    },
+    {
+      "epoch": 2.6275039387801034,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.885073053170659e-05,
+      "loss": 0.5853,
+      "step": 8757
+    },
+    {
+      "epoch": 2.6278040363118014,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.883566802229253e-05,
+      "loss": 0.4332,
+      "step": 8758
+    },
+    {
+      "epoch": 2.6281041338434994,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.882060551287844e-05,
+      "loss": 0.5348,
+      "step": 8759
+    },
+    {
+      "epoch": 2.628404231375197,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.880554300346438e-05,
+      "loss": 0.5405,
+      "step": 8760
+    },
+    {
+      "epoch": 2.6287043289068945,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.879048049405032e-05,
+      "loss": 0.5824,
+      "step": 8761
+    },
+    {
+      "epoch": 2.6290044264385926,
+      "grad_norm": 1.0,
+      "learning_rate": 6.877541798463625e-05,
+      "loss": 0.4968,
+      "step": 8762
+    },
+    {
+      "epoch": 2.6293045239702906,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.876035547522217e-05,
+      "loss": 0.4635,
+      "step": 8763
+    },
+    {
+      "epoch": 2.629604621501988,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.87452929658081e-05,
+      "loss": 0.5626,
+      "step": 8764
+    },
+    {
+      "epoch": 2.6299047190336857,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.873023045639404e-05,
+      "loss": 0.537,
+      "step": 8765
+    },
+    {
+      "epoch": 2.6302048165653837,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.871516794697998e-05,
+      "loss": 0.4939,
+      "step": 8766
+    },
+    {
+      "epoch": 2.6305049140970818,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.87001054375659e-05,
+      "loss": 0.4241,
+      "step": 8767
+    },
+    {
+      "epoch": 2.6308050116287793,
+      "grad_norm": 1.0,
+      "learning_rate": 6.868504292815183e-05,
+      "loss": 0.5693,
+      "step": 8768
+    },
+    {
+      "epoch": 2.631105109160477,
+      "grad_norm": 1.0,
+      "learning_rate": 6.866998041873776e-05,
+      "loss": 0.5202,
+      "step": 8769
+    },
+    {
+      "epoch": 2.631405206692175,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.86549179093237e-05,
+      "loss": 0.5375,
+      "step": 8770
+    },
+    {
+      "epoch": 2.631705304223873,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.863985539990963e-05,
+      "loss": 0.5765,
+      "step": 8771
+    },
+    {
+      "epoch": 2.6320054017555705,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.862479289049556e-05,
+      "loss": 0.5106,
+      "step": 8772
+    },
+    {
+      "epoch": 2.632305499287268,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.86097303810815e-05,
+      "loss": 0.528,
+      "step": 8773
+    },
+    {
+      "epoch": 2.632605596818966,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.859466787166742e-05,
+      "loss": 0.5923,
+      "step": 8774
+    },
+    {
+      "epoch": 2.632905694350664,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.857960536225335e-05,
+      "loss": 0.4665,
+      "step": 8775
+    },
+    {
+      "epoch": 2.6332057918823617,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.856454285283929e-05,
+      "loss": 0.5881,
+      "step": 8776
+    },
+    {
+      "epoch": 2.6335058894140597,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.854948034342522e-05,
+      "loss": 0.4076,
+      "step": 8777
+    },
+    {
+      "epoch": 2.6338059869457573,
+      "grad_norm": 1.125,
+      "learning_rate": 6.853441783401114e-05,
+      "loss": 0.5641,
+      "step": 8778
+    },
+    {
+      "epoch": 2.6341060844774553,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.851935532459708e-05,
+      "loss": 0.5091,
+      "step": 8779
+    },
+    {
+      "epoch": 2.634406182009153,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.850429281518301e-05,
+      "loss": 0.5103,
+      "step": 8780
+    },
+    {
+      "epoch": 2.634706279540851,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.848923030576895e-05,
+      "loss": 0.5915,
+      "step": 8781
+    },
+    {
+      "epoch": 2.6350063770725485,
+      "grad_norm": 1.125,
+      "learning_rate": 6.847416779635488e-05,
+      "loss": 0.6009,
+      "step": 8782
+    },
+    {
+      "epoch": 2.6353064746042465,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.84591052869408e-05,
+      "loss": 0.5724,
+      "step": 8783
+    },
+    {
+      "epoch": 2.635606572135944,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.844404277752674e-05,
+      "loss": 0.5538,
+      "step": 8784
+    },
+    {
+      "epoch": 2.635906669667642,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.842898026811267e-05,
+      "loss": 0.5816,
+      "step": 8785
+    },
+    {
+      "epoch": 2.6362067671993397,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.84139177586986e-05,
+      "loss": 0.5725,
+      "step": 8786
+    },
+    {
+      "epoch": 2.6365068647310377,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.839885524928454e-05,
+      "loss": 0.5078,
+      "step": 8787
+    },
+    {
+      "epoch": 2.6368069622627353,
+      "grad_norm": 1.125,
+      "learning_rate": 6.838379273987046e-05,
+      "loss": 0.5955,
+      "step": 8788
+    },
+    {
+      "epoch": 2.6371070597944333,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.83687302304564e-05,
+      "loss": 0.6332,
+      "step": 8789
+    },
+    {
+      "epoch": 2.637407157326131,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.835366772104233e-05,
+      "loss": 0.5506,
+      "step": 8790
+    },
+    {
+      "epoch": 2.637707254857829,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.833860521162826e-05,
+      "loss": 0.5453,
+      "step": 8791
+    },
+    {
+      "epoch": 2.6380073523895264,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.83235427022142e-05,
+      "loss": 0.5378,
+      "step": 8792
+    },
+    {
+      "epoch": 2.6383074499212245,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.830848019280013e-05,
+      "loss": 0.5328,
+      "step": 8793
+    },
+    {
+      "epoch": 2.6386075474529225,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.829341768338605e-05,
+      "loss": 0.5768,
+      "step": 8794
+    },
+    {
+      "epoch": 2.63890764498462,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.827835517397199e-05,
+      "loss": 0.4618,
+      "step": 8795
+    },
+    {
+      "epoch": 2.6392077425163176,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.826329266455792e-05,
+      "loss": 0.5405,
+      "step": 8796
+    },
+    {
+      "epoch": 2.6395078400480156,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.824823015514386e-05,
+      "loss": 0.5886,
+      "step": 8797
+    },
+    {
+      "epoch": 2.6398079375797137,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.823316764572977e-05,
+      "loss": 0.5518,
+      "step": 8798
+    },
+    {
+      "epoch": 2.6401080351114112,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.821810513631571e-05,
+      "loss": 0.5663,
+      "step": 8799
+    },
+    {
+      "epoch": 2.640408132643109,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.820304262690165e-05,
+      "loss": 0.6024,
+      "step": 8800
+    },
+    {
+      "epoch": 2.640708230174807,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.818798011748758e-05,
+      "loss": 0.5877,
+      "step": 8801
+    },
+    {
+      "epoch": 2.641008327706505,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.81729176080735e-05,
+      "loss": 0.5365,
+      "step": 8802
+    },
+    {
+      "epoch": 2.6413084252382024,
+      "grad_norm": 1.0,
+      "learning_rate": 6.815785509865943e-05,
+      "loss": 0.5541,
+      "step": 8803
+    },
+    {
+      "epoch": 2.6416085227699,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.814279258924537e-05,
+      "loss": 0.5643,
+      "step": 8804
+    },
+    {
+      "epoch": 2.641908620301598,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.81277300798313e-05,
+      "loss": 0.4654,
+      "step": 8805
+    },
+    {
+      "epoch": 2.642208717833296,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.811266757041723e-05,
+      "loss": 0.5602,
+      "step": 8806
+    },
+    {
+      "epoch": 2.6425088153649936,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.809760506100317e-05,
+      "loss": 0.5337,
+      "step": 8807
+    },
+    {
+      "epoch": 2.642808912896691,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.808254255158911e-05,
+      "loss": 0.5059,
+      "step": 8808
+    },
+    {
+      "epoch": 2.643109010428389,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.806748004217502e-05,
+      "loss": 0.5689,
+      "step": 8809
+    },
+    {
+      "epoch": 2.643409107960087,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.805241753276096e-05,
+      "loss": 0.5146,
+      "step": 8810
+    },
+    {
+      "epoch": 2.643709205491785,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.803735502334689e-05,
+      "loss": 0.4773,
+      "step": 8811
+    },
+    {
+      "epoch": 2.6440093030234824,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.802229251393283e-05,
+      "loss": 0.6241,
+      "step": 8812
+    },
+    {
+      "epoch": 2.6443094005551804,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.800723000451876e-05,
+      "loss": 0.5061,
+      "step": 8813
+    },
+    {
+      "epoch": 2.6446094980868784,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.799216749510468e-05,
+      "loss": 0.5546,
+      "step": 8814
+    },
+    {
+      "epoch": 2.644909595618576,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.797710498569062e-05,
+      "loss": 0.591,
+      "step": 8815
+    },
+    {
+      "epoch": 2.645209693150274,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.796204247627655e-05,
+      "loss": 0.5792,
+      "step": 8816
+    },
+    {
+      "epoch": 2.6455097906819716,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.794697996686248e-05,
+      "loss": 0.6271,
+      "step": 8817
+    },
+    {
+      "epoch": 2.6458098882136696,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.793191745744842e-05,
+      "loss": 0.5023,
+      "step": 8818
+    },
+    {
+      "epoch": 2.646109985745367,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.791685494803434e-05,
+      "loss": 0.4555,
+      "step": 8819
+    },
+    {
+      "epoch": 2.646410083277065,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.790179243862028e-05,
+      "loss": 0.5306,
+      "step": 8820
+    },
+    {
+      "epoch": 2.6467101808087627,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.788672992920621e-05,
+      "loss": 0.5476,
+      "step": 8821
+    },
+    {
+      "epoch": 2.6470102783404608,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.787166741979214e-05,
+      "loss": 0.4569,
+      "step": 8822
+    },
+    {
+      "epoch": 2.6473103758721583,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.785660491037808e-05,
+      "loss": 0.6115,
+      "step": 8823
+    },
+    {
+      "epoch": 2.6476104734038564,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.7841542400964e-05,
+      "loss": 0.5749,
+      "step": 8824
+    },
+    {
+      "epoch": 2.647910570935554,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.782647989154993e-05,
+      "loss": 0.6021,
+      "step": 8825
+    },
+    {
+      "epoch": 2.648210668467252,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.781141738213587e-05,
+      "loss": 0.5945,
+      "step": 8826
+    },
+    {
+      "epoch": 2.6485107659989495,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.77963548727218e-05,
+      "loss": 0.5652,
+      "step": 8827
+    },
+    {
+      "epoch": 2.6488108635306475,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.778129236330774e-05,
+      "loss": 0.5819,
+      "step": 8828
+    },
+    {
+      "epoch": 2.649110961062345,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.776622985389366e-05,
+      "loss": 0.4838,
+      "step": 8829
+    },
+    {
+      "epoch": 2.649411058594043,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.775116734447959e-05,
+      "loss": 0.5511,
+      "step": 8830
+    },
+    {
+      "epoch": 2.6497111561257407,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.773610483506553e-05,
+      "loss": 0.5241,
+      "step": 8831
+    },
+    {
+      "epoch": 2.6500112536574387,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.772104232565144e-05,
+      "loss": 0.6186,
+      "step": 8832
+    },
+    {
+      "epoch": 2.6503113511891367,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.770597981623739e-05,
+      "loss": 0.5466,
+      "step": 8833
+    },
+    {
+      "epoch": 2.6506114487208343,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.769091730682333e-05,
+      "loss": 0.5096,
+      "step": 8834
+    },
+    {
+      "epoch": 2.650911546252532,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.767585479740925e-05,
+      "loss": 0.4191,
+      "step": 8835
+    },
+    {
+      "epoch": 2.65121164378423,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.766079228799518e-05,
+      "loss": 0.5331,
+      "step": 8836
+    },
+    {
+      "epoch": 2.651511741315928,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.764572977858112e-05,
+      "loss": 0.5423,
+      "step": 8837
+    },
+    {
+      "epoch": 2.6518118388476255,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.763066726916705e-05,
+      "loss": 0.5171,
+      "step": 8838
+    },
+    {
+      "epoch": 2.652111936379323,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.761560475975299e-05,
+      "loss": 0.4972,
+      "step": 8839
+    },
+    {
+      "epoch": 2.652412033911021,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.76005422503389e-05,
+      "loss": 0.5608,
+      "step": 8840
+    },
+    {
+      "epoch": 2.652712131442719,
+      "grad_norm": 1.2734375,
+      "learning_rate": 6.758547974092484e-05,
+      "loss": 0.5688,
+      "step": 8841
+    },
+    {
+      "epoch": 2.6530122289744167,
+      "grad_norm": 1.0,
+      "learning_rate": 6.757041723151078e-05,
+      "loss": 0.4667,
+      "step": 8842
+    },
+    {
+      "epoch": 2.6533123265061143,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.755535472209671e-05,
+      "loss": 0.499,
+      "step": 8843
+    },
+    {
+      "epoch": 2.6536124240378123,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.754029221268263e-05,
+      "loss": 0.5057,
+      "step": 8844
+    },
+    {
+      "epoch": 2.6539125215695103,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.752522970326856e-05,
+      "loss": 0.5179,
+      "step": 8845
+    },
+    {
+      "epoch": 2.654212619101208,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.75101671938545e-05,
+      "loss": 0.6785,
+      "step": 8846
+    },
+    {
+      "epoch": 2.6545127166329054,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.749510468444044e-05,
+      "loss": 0.5252,
+      "step": 8847
+    },
+    {
+      "epoch": 2.6548128141646035,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.748004217502635e-05,
+      "loss": 0.5423,
+      "step": 8848
+    },
+    {
+      "epoch": 2.6551129116963015,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.74649796656123e-05,
+      "loss": 0.5384,
+      "step": 8849
+    },
+    {
+      "epoch": 2.655413009227999,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.744991715619822e-05,
+      "loss": 0.5111,
+      "step": 8850
+    },
+    {
+      "epoch": 2.6557131067596966,
+      "grad_norm": 1.125,
+      "learning_rate": 6.743485464678416e-05,
+      "loss": 0.4805,
+      "step": 8851
+    },
+    {
+      "epoch": 2.6560132042913946,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.741979213737009e-05,
+      "loss": 0.4899,
+      "step": 8852
+    },
+    {
+      "epoch": 2.6563133018230927,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.740472962795602e-05,
+      "loss": 0.5694,
+      "step": 8853
+    },
+    {
+      "epoch": 2.6566133993547902,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.738966711854196e-05,
+      "loss": 0.5397,
+      "step": 8854
+    },
+    {
+      "epoch": 2.6569134968864883,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.73746046091279e-05,
+      "loss": 0.5336,
+      "step": 8855
+    },
+    {
+      "epoch": 2.657213594418186,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.735954209971381e-05,
+      "loss": 0.5437,
+      "step": 8856
+    },
+    {
+      "epoch": 2.657513691949884,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.734447959029975e-05,
+      "loss": 0.6358,
+      "step": 8857
+    },
+    {
+      "epoch": 2.6578137894815814,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.732941708088568e-05,
+      "loss": 0.5053,
+      "step": 8858
+    },
+    {
+      "epoch": 2.6581138870132794,
+      "grad_norm": 1.0,
+      "learning_rate": 6.73143545714716e-05,
+      "loss": 0.5058,
+      "step": 8859
+    },
+    {
+      "epoch": 2.658413984544977,
+      "grad_norm": 1.125,
+      "learning_rate": 6.729929206205754e-05,
+      "loss": 0.6006,
+      "step": 8860
+    },
+    {
+      "epoch": 2.658714082076675,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.728422955264347e-05,
+      "loss": 0.503,
+      "step": 8861
+    },
+    {
+      "epoch": 2.6590141796083726,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.726916704322941e-05,
+      "loss": 0.5241,
+      "step": 8862
+    },
+    {
+      "epoch": 2.6593142771400706,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.725410453381534e-05,
+      "loss": 0.5524,
+      "step": 8863
+    },
+    {
+      "epoch": 2.659614374671768,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.723904202440126e-05,
+      "loss": 0.5542,
+      "step": 8864
+    },
+    {
+      "epoch": 2.659914472203466,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.72239795149872e-05,
+      "loss": 0.6226,
+      "step": 8865
+    },
+    {
+      "epoch": 2.660214569735164,
+      "grad_norm": 1.0,
+      "learning_rate": 6.720891700557313e-05,
+      "loss": 0.5805,
+      "step": 8866
+    },
+    {
+      "epoch": 2.660514667266862,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.719385449615906e-05,
+      "loss": 0.572,
+      "step": 8867
+    },
+    {
+      "epoch": 2.6608147647985594,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.7178791986745e-05,
+      "loss": 0.5851,
+      "step": 8868
+    },
+    {
+      "epoch": 2.6611148623302574,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.716372947733092e-05,
+      "loss": 0.5041,
+      "step": 8869
+    },
+    {
+      "epoch": 2.661414959861955,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.714866696791686e-05,
+      "loss": 0.5561,
+      "step": 8870
+    },
+    {
+      "epoch": 2.661715057393653,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.713360445850279e-05,
+      "loss": 0.5367,
+      "step": 8871
+    },
+    {
+      "epoch": 2.662015154925351,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.711854194908872e-05,
+      "loss": 0.5318,
+      "step": 8872
+    },
+    {
+      "epoch": 2.6623152524570486,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.710347943967466e-05,
+      "loss": 0.6212,
+      "step": 8873
+    },
+    {
+      "epoch": 2.662615349988746,
+      "grad_norm": 1.0,
+      "learning_rate": 6.708841693026059e-05,
+      "loss": 0.5746,
+      "step": 8874
+    },
+    {
+      "epoch": 2.662915447520444,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.707335442084651e-05,
+      "loss": 0.6158,
+      "step": 8875
+    },
+    {
+      "epoch": 2.663215545052142,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.705829191143245e-05,
+      "loss": 0.4574,
+      "step": 8876
+    },
+    {
+      "epoch": 2.6635156425838398,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.704322940201838e-05,
+      "loss": 0.59,
+      "step": 8877
+    },
+    {
+      "epoch": 2.6638157401155373,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.702816689260432e-05,
+      "loss": 0.5359,
+      "step": 8878
+    },
+    {
+      "epoch": 2.6641158376472354,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.701310438319025e-05,
+      "loss": 0.6501,
+      "step": 8879
+    },
+    {
+      "epoch": 2.6644159351789334,
+      "grad_norm": 1.0,
+      "learning_rate": 6.699804187377617e-05,
+      "loss": 0.5111,
+      "step": 8880
+    },
+    {
+      "epoch": 2.664716032710631,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.698297936436211e-05,
+      "loss": 0.5408,
+      "step": 8881
+    },
+    {
+      "epoch": 2.6650161302423285,
+      "grad_norm": 1.0,
+      "learning_rate": 6.696791685494804e-05,
+      "loss": 0.5435,
+      "step": 8882
+    },
+    {
+      "epoch": 2.6653162277740265,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.695285434553397e-05,
+      "loss": 0.5987,
+      "step": 8883
+    },
+    {
+      "epoch": 2.6656163253057246,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.693779183611991e-05,
+      "loss": 0.5586,
+      "step": 8884
+    },
+    {
+      "epoch": 2.665916422837422,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.692272932670583e-05,
+      "loss": 0.5313,
+      "step": 8885
+    },
+    {
+      "epoch": 2.6662165203691197,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.690766681729176e-05,
+      "loss": 0.488,
+      "step": 8886
+    },
+    {
+      "epoch": 2.6665166179008177,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.689260430787769e-05,
+      "loss": 0.4628,
+      "step": 8887
+    },
+    {
+      "epoch": 2.6668167154325157,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.687754179846363e-05,
+      "loss": 0.5051,
+      "step": 8888
+    },
+    {
+      "epoch": 2.6671168129642133,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.686247928904957e-05,
+      "loss": 0.5569,
+      "step": 8889
+    },
+    {
+      "epoch": 2.667416910495911,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.684741677963548e-05,
+      "loss": 0.6013,
+      "step": 8890
+    },
+    {
+      "epoch": 2.667717008027609,
+      "grad_norm": 1.0,
+      "learning_rate": 6.683235427022142e-05,
+      "loss": 0.5538,
+      "step": 8891
+    },
+    {
+      "epoch": 2.668017105559307,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.681729176080735e-05,
+      "loss": 0.5737,
+      "step": 8892
+    },
+    {
+      "epoch": 2.6683172030910045,
+      "grad_norm": 1.0,
+      "learning_rate": 6.680222925139329e-05,
+      "loss": 0.5462,
+      "step": 8893
+    },
+    {
+      "epoch": 2.6686173006227025,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.678716674197922e-05,
+      "loss": 0.4955,
+      "step": 8894
+    },
+    {
+      "epoch": 2.6689173981544,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.677210423256514e-05,
+      "loss": 0.5506,
+      "step": 8895
+    },
+    {
+      "epoch": 2.669217495686098,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.675704172315108e-05,
+      "loss": 0.5025,
+      "step": 8896
+    },
+    {
+      "epoch": 2.6695175932177957,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.674197921373702e-05,
+      "loss": 0.5539,
+      "step": 8897
+    },
+    {
+      "epoch": 2.6698176907494937,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.672691670432294e-05,
+      "loss": 0.4761,
+      "step": 8898
+    },
+    {
+      "epoch": 2.6701177882811913,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.671185419490888e-05,
+      "loss": 0.5107,
+      "step": 8899
+    },
+    {
+      "epoch": 2.6704178858128893,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.66967916854948e-05,
+      "loss": 0.5406,
+      "step": 8900
+    },
+    {
+      "epoch": 2.670717983344587,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.668172917608074e-05,
+      "loss": 0.4935,
+      "step": 8901
+    },
+    {
+      "epoch": 2.671018080876285,
+      "grad_norm": 1.0,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.5105,
+      "step": 8902
+    },
+    {
+      "epoch": 2.6713181784079825,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.66516041572526e-05,
+      "loss": 0.5454,
+      "step": 8903
+    },
+    {
+      "epoch": 2.6716182759396805,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.663654164783854e-05,
+      "loss": 0.5273,
+      "step": 8904
+    },
+    {
+      "epoch": 2.671918373471378,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.662147913842446e-05,
+      "loss": 0.5338,
+      "step": 8905
+    },
+    {
+      "epoch": 2.672218471003076,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.660641662901039e-05,
+      "loss": 0.5779,
+      "step": 8906
+    },
+    {
+      "epoch": 2.6725185685347737,
+      "grad_norm": 1.125,
+      "learning_rate": 6.659135411959633e-05,
+      "loss": 0.6221,
+      "step": 8907
+    },
+    {
+      "epoch": 2.6728186660664717,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.657629161018226e-05,
+      "loss": 0.5756,
+      "step": 8908
+    },
+    {
+      "epoch": 2.6731187635981692,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.65612291007682e-05,
+      "loss": 0.5161,
+      "step": 8909
+    },
+    {
+      "epoch": 2.6734188611298673,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.654616659135412e-05,
+      "loss": 0.6321,
+      "step": 8910
+    },
+    {
+      "epoch": 2.6737189586615653,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.653110408194005e-05,
+      "loss": 0.4711,
+      "step": 8911
+    },
+    {
+      "epoch": 2.674019056193263,
+      "grad_norm": 1.0,
+      "learning_rate": 6.651604157252599e-05,
+      "loss": 0.4404,
+      "step": 8912
+    },
+    {
+      "epoch": 2.6743191537249604,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.650097906311192e-05,
+      "loss": 0.5286,
+      "step": 8913
+    },
+    {
+      "epoch": 2.6746192512566584,
+      "grad_norm": 1.0,
+      "learning_rate": 6.648591655369785e-05,
+      "loss": 0.5298,
+      "step": 8914
+    },
+    {
+      "epoch": 2.6749193487883565,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.647085404428379e-05,
+      "loss": 0.4782,
+      "step": 8915
+    },
+    {
+      "epoch": 2.675219446320054,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.645579153486971e-05,
+      "loss": 0.4892,
+      "step": 8916
+    },
+    {
+      "epoch": 2.6755195438517516,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.644072902545564e-05,
+      "loss": 0.5969,
+      "step": 8917
+    },
+    {
+      "epoch": 2.6758196413834496,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.642566651604158e-05,
+      "loss": 0.5702,
+      "step": 8918
+    },
+    {
+      "epoch": 2.6761197389151477,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.64106040066275e-05,
+      "loss": 0.5383,
+      "step": 8919
+    },
+    {
+      "epoch": 2.6764198364468452,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.639554149721345e-05,
+      "loss": 0.5483,
+      "step": 8920
+    },
+    {
+      "epoch": 2.676719933978543,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.638047898779936e-05,
+      "loss": 0.5134,
+      "step": 8921
+    },
+    {
+      "epoch": 2.677020031510241,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.63654164783853e-05,
+      "loss": 0.4893,
+      "step": 8922
+    },
+    {
+      "epoch": 2.677320129041939,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.635035396897124e-05,
+      "loss": 0.6017,
+      "step": 8923
+    },
+    {
+      "epoch": 2.6776202265736364,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.633529145955717e-05,
+      "loss": 0.4914,
+      "step": 8924
+    },
+    {
+      "epoch": 2.677920324105334,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.63202289501431e-05,
+      "loss": 0.4989,
+      "step": 8925
+    },
+    {
+      "epoch": 2.678220421637032,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.630516644072903e-05,
+      "loss": 0.6682,
+      "step": 8926
+    },
+    {
+      "epoch": 2.67852051916873,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.629010393131496e-05,
+      "loss": 0.5539,
+      "step": 8927
+    },
+    {
+      "epoch": 2.6788206167004276,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.62750414219009e-05,
+      "loss": 0.5118,
+      "step": 8928
+    },
+    {
+      "epoch": 2.679120714232125,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.625997891248681e-05,
+      "loss": 0.4996,
+      "step": 8929
+    },
+    {
+      "epoch": 2.679420811763823,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.624491640307275e-05,
+      "loss": 0.5851,
+      "step": 8930
+    },
+    {
+      "epoch": 2.679720909295521,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.62298538936587e-05,
+      "loss": 0.4424,
+      "step": 8931
+    },
+    {
+      "epoch": 2.6800210068272188,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.621479138424462e-05,
+      "loss": 0.5124,
+      "step": 8932
+    },
+    {
+      "epoch": 2.680321104358917,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.619972887483055e-05,
+      "loss": 0.5112,
+      "step": 8933
+    },
+    {
+      "epoch": 2.6806212018906144,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.618466636541648e-05,
+      "loss": 0.4941,
+      "step": 8934
+    },
+    {
+      "epoch": 2.6809212994223124,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.616960385600242e-05,
+      "loss": 0.5932,
+      "step": 8935
+    },
+    {
+      "epoch": 2.68122139695401,
+      "grad_norm": 1.3359375,
+      "learning_rate": 6.615454134658836e-05,
+      "loss": 0.5531,
+      "step": 8936
+    },
+    {
+      "epoch": 2.681521494485708,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.613947883717427e-05,
+      "loss": 0.5259,
+      "step": 8937
+    },
+    {
+      "epoch": 2.6818215920174056,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.612441632776021e-05,
+      "loss": 0.5205,
+      "step": 8938
+    },
+    {
+      "epoch": 2.6821216895491036,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.610935381834614e-05,
+      "loss": 0.5914,
+      "step": 8939
+    },
+    {
+      "epoch": 2.682421787080801,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.609429130893206e-05,
+      "loss": 0.5422,
+      "step": 8940
+    },
+    {
+      "epoch": 2.682721884612499,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.6079228799518e-05,
+      "loss": 0.4813,
+      "step": 8941
+    },
+    {
+      "epoch": 2.6830219821441967,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.606416629010393e-05,
+      "loss": 0.5574,
+      "step": 8942
+    },
+    {
+      "epoch": 2.6833220796758948,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.604910378068987e-05,
+      "loss": 0.4976,
+      "step": 8943
+    },
+    {
+      "epoch": 2.6836221772075923,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.60340412712758e-05,
+      "loss": 0.5413,
+      "step": 8944
+    },
+    {
+      "epoch": 2.6839222747392903,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.601897876186172e-05,
+      "loss": 0.4827,
+      "step": 8945
+    },
+    {
+      "epoch": 2.684222372270988,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.600391625244766e-05,
+      "loss": 0.594,
+      "step": 8946
+    },
+    {
+      "epoch": 2.684522469802686,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.598885374303359e-05,
+      "loss": 0.5031,
+      "step": 8947
+    },
+    {
+      "epoch": 2.6848225673343835,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.597379123361952e-05,
+      "loss": 0.5608,
+      "step": 8948
+    },
+    {
+      "epoch": 2.6851226648660815,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.595872872420546e-05,
+      "loss": 0.5511,
+      "step": 8949
+    },
+    {
+      "epoch": 2.6854227623977796,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.594366621479138e-05,
+      "loss": 0.5326,
+      "step": 8950
+    },
+    {
+      "epoch": 2.685722859929477,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.592860370537732e-05,
+      "loss": 0.5457,
+      "step": 8951
+    },
+    {
+      "epoch": 2.6860229574611747,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.591354119596325e-05,
+      "loss": 0.5402,
+      "step": 8952
+    },
+    {
+      "epoch": 2.6863230549928727,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.589847868654918e-05,
+      "loss": 0.494,
+      "step": 8953
+    },
+    {
+      "epoch": 2.6866231525245707,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.588341617713512e-05,
+      "loss": 0.5454,
+      "step": 8954
+    },
+    {
+      "epoch": 2.6869232500562683,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.586835366772105e-05,
+      "loss": 0.6376,
+      "step": 8955
+    },
+    {
+      "epoch": 2.687223347587966,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.585329115830697e-05,
+      "loss": 0.5612,
+      "step": 8956
+    },
+    {
+      "epoch": 2.687523445119664,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.583822864889291e-05,
+      "loss": 0.606,
+      "step": 8957
+    },
+    {
+      "epoch": 2.687823542651362,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.582316613947884e-05,
+      "loss": 0.5296,
+      "step": 8958
+    },
+    {
+      "epoch": 2.6881236401830595,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.580810363006478e-05,
+      "loss": 0.5509,
+      "step": 8959
+    },
+    {
+      "epoch": 2.688423737714757,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.57930411206507e-05,
+      "loss": 0.5607,
+      "step": 8960
+    },
+    {
+      "epoch": 2.688723835246455,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.577797861123663e-05,
+      "loss": 0.4868,
+      "step": 8961
+    },
+    {
+      "epoch": 2.689023932778153,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.576291610182257e-05,
+      "loss": 0.646,
+      "step": 8962
+    },
+    {
+      "epoch": 2.6893240303098507,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.57478535924085e-05,
+      "loss": 0.5746,
+      "step": 8963
+    },
+    {
+      "epoch": 2.6896241278415483,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.573279108299443e-05,
+      "loss": 0.5634,
+      "step": 8964
+    },
+    {
+      "epoch": 2.6899242253732463,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.571772857358037e-05,
+      "loss": 0.5578,
+      "step": 8965
+    },
+    {
+      "epoch": 2.6902243229049443,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.57026660641663e-05,
+      "loss": 0.518,
+      "step": 8966
+    },
+    {
+      "epoch": 2.690524420436642,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.568760355475222e-05,
+      "loss": 0.4704,
+      "step": 8967
+    },
+    {
+      "epoch": 2.6908245179683394,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.567254104533815e-05,
+      "loss": 0.5867,
+      "step": 8968
+    },
+    {
+      "epoch": 2.6911246155000375,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.565747853592409e-05,
+      "loss": 0.518,
+      "step": 8969
+    },
+    {
+      "epoch": 2.6914247130317355,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.564241602651003e-05,
+      "loss": 0.6116,
+      "step": 8970
+    },
+    {
+      "epoch": 2.691724810563433,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.562735351709594e-05,
+      "loss": 0.5385,
+      "step": 8971
+    },
+    {
+      "epoch": 2.692024908095131,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.561229100768188e-05,
+      "loss": 0.6044,
+      "step": 8972
+    },
+    {
+      "epoch": 2.6923250056268286,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.559722849826782e-05,
+      "loss": 0.5064,
+      "step": 8973
+    },
+    {
+      "epoch": 2.6926251031585267,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.558216598885375e-05,
+      "loss": 0.5427,
+      "step": 8974
+    },
+    {
+      "epoch": 2.6929252006902242,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.556710347943968e-05,
+      "loss": 0.5262,
+      "step": 8975
+    },
+    {
+      "epoch": 2.6932252982219222,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.55520409700256e-05,
+      "loss": 0.4852,
+      "step": 8976
+    },
+    {
+      "epoch": 2.69352539575362,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.553697846061154e-05,
+      "loss": 0.6047,
+      "step": 8977
+    },
+    {
+      "epoch": 2.693825493285318,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.552191595119748e-05,
+      "loss": 0.5501,
+      "step": 8978
+    },
+    {
+      "epoch": 2.6941255908170154,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.55068534417834e-05,
+      "loss": 0.5504,
+      "step": 8979
+    },
+    {
+      "epoch": 2.6944256883487134,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.549179093236934e-05,
+      "loss": 0.5377,
+      "step": 8980
+    },
+    {
+      "epoch": 2.694725785880411,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.547672842295526e-05,
+      "loss": 0.628,
+      "step": 8981
+    },
+    {
+      "epoch": 2.695025883412109,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.54616659135412e-05,
+      "loss": 0.5466,
+      "step": 8982
+    },
+    {
+      "epoch": 2.6953259809438066,
+      "grad_norm": 1.0,
+      "learning_rate": 6.544660340412713e-05,
+      "loss": 0.5019,
+      "step": 8983
+    },
+    {
+      "epoch": 2.6956260784755046,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.543154089471306e-05,
+      "loss": 0.5771,
+      "step": 8984
+    },
+    {
+      "epoch": 2.695926176007202,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.5416478385299e-05,
+      "loss": 0.5531,
+      "step": 8985
+    },
+    {
+      "epoch": 2.6962262735389,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.540141587588494e-05,
+      "loss": 0.6047,
+      "step": 8986
+    },
+    {
+      "epoch": 2.696526371070598,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.538635336647085e-05,
+      "loss": 0.5668,
+      "step": 8987
+    },
+    {
+      "epoch": 2.696826468602296,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.537129085705679e-05,
+      "loss": 0.5045,
+      "step": 8988
+    },
+    {
+      "epoch": 2.697126566133994,
+      "grad_norm": 1.125,
+      "learning_rate": 6.535622834764272e-05,
+      "loss": 0.5764,
+      "step": 8989
+    },
+    {
+      "epoch": 2.6974266636656914,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.534116583822866e-05,
+      "loss": 0.5086,
+      "step": 8990
+    },
+    {
+      "epoch": 2.697726761197389,
+      "grad_norm": 1.4453125,
+      "learning_rate": 6.532610332881458e-05,
+      "loss": 0.6539,
+      "step": 8991
+    },
+    {
+      "epoch": 2.698026858729087,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.531104081940051e-05,
+      "loss": 0.51,
+      "step": 8992
+    },
+    {
+      "epoch": 2.698326956260785,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.529597830998645e-05,
+      "loss": 0.54,
+      "step": 8993
+    },
+    {
+      "epoch": 2.6986270537924826,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.528091580057238e-05,
+      "loss": 0.5659,
+      "step": 8994
+    },
+    {
+      "epoch": 2.69892715132418,
+      "grad_norm": 1.234375,
+      "learning_rate": 6.52658532911583e-05,
+      "loss": 0.4881,
+      "step": 8995
+    },
+    {
+      "epoch": 2.699227248855878,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.525079078174425e-05,
+      "loss": 0.5503,
+      "step": 8996
+    },
+    {
+      "epoch": 2.699527346387576,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.523572827233017e-05,
+      "loss": 0.5192,
+      "step": 8997
+    },
+    {
+      "epoch": 2.6998274439192738,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.52206657629161e-05,
+      "loss": 0.5392,
+      "step": 8998
+    },
+    {
+      "epoch": 2.7001275414509713,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.520560325350204e-05,
+      "loss": 0.4894,
+      "step": 8999
+    },
+    {
+      "epoch": 2.7004276389826694,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.519054074408797e-05,
+      "loss": 0.5352,
+      "step": 9000
+    },
+    {
+      "epoch": 2.7007277365143674,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.51754782346739e-05,
+      "loss": 0.5538,
+      "step": 9001
+    },
+    {
+      "epoch": 2.701027834046065,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.516041572525983e-05,
+      "loss": 0.5527,
+      "step": 9002
+    },
+    {
+      "epoch": 2.7013279315777625,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.514535321584576e-05,
+      "loss": 0.4845,
+      "step": 9003
+    },
+    {
+      "epoch": 2.7016280291094605,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.51302907064317e-05,
+      "loss": 0.5344,
+      "step": 9004
+    },
+    {
+      "epoch": 2.7019281266411586,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.511522819701763e-05,
+      "loss": 0.5193,
+      "step": 9005
+    },
+    {
+      "epoch": 2.702228224172856,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.510016568760355e-05,
+      "loss": 0.6159,
+      "step": 9006
+    },
+    {
+      "epoch": 2.7025283217045537,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.50851031781895e-05,
+      "loss": 0.4314,
+      "step": 9007
+    },
+    {
+      "epoch": 2.7028284192362517,
+      "grad_norm": 0.98046875,
+      "learning_rate": 6.507004066877542e-05,
+      "loss": 0.5367,
+      "step": 9008
+    },
+    {
+      "epoch": 2.7031285167679497,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.505497815936136e-05,
+      "loss": 0.4446,
+      "step": 9009
+    },
+    {
+      "epoch": 2.7034286142996473,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.503991564994727e-05,
+      "loss": 0.5275,
+      "step": 9010
+    },
+    {
+      "epoch": 2.7037287118313453,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.502485314053321e-05,
+      "loss": 0.5547,
+      "step": 9011
+    },
+    {
+      "epoch": 2.704028809363043,
+      "grad_norm": 1.3125,
+      "learning_rate": 6.500979063111915e-05,
+      "loss": 0.6246,
+      "step": 9012
+    },
+    {
+      "epoch": 2.704328906894741,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.499472812170508e-05,
+      "loss": 0.557,
+      "step": 9013
+    },
+    {
+      "epoch": 2.7046290044264385,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.497966561229101e-05,
+      "loss": 0.5501,
+      "step": 9014
+    },
+    {
+      "epoch": 2.7049291019581365,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.496460310287695e-05,
+      "loss": 0.4818,
+      "step": 9015
+    },
+    {
+      "epoch": 2.705229199489834,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.494954059346288e-05,
+      "loss": 0.4912,
+      "step": 9016
+    },
+    {
+      "epoch": 2.705529297021532,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.493447808404882e-05,
+      "loss": 0.4756,
+      "step": 9017
+    },
+    {
+      "epoch": 2.7058293945532297,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.491941557463473e-05,
+      "loss": 0.4862,
+      "step": 9018
+    },
+    {
+      "epoch": 2.7061294920849277,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.490435306522067e-05,
+      "loss": 0.4936,
+      "step": 9019
+    },
+    {
+      "epoch": 2.7064295896166253,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.488929055580661e-05,
+      "loss": 0.5097,
+      "step": 9020
+    },
+    {
+      "epoch": 2.7067296871483233,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.487422804639252e-05,
+      "loss": 0.5429,
+      "step": 9021
+    },
+    {
+      "epoch": 2.707029784680021,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.485916553697846e-05,
+      "loss": 0.4904,
+      "step": 9022
+    },
+    {
+      "epoch": 2.707329882211719,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.484410302756439e-05,
+      "loss": 0.5781,
+      "step": 9023
+    },
+    {
+      "epoch": 2.7076299797434165,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.482904051815033e-05,
+      "loss": 0.4907,
+      "step": 9024
+    },
+    {
+      "epoch": 2.7079300772751145,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.481397800873626e-05,
+      "loss": 0.5016,
+      "step": 9025
+    },
+    {
+      "epoch": 2.708230174806812,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.479891549932218e-05,
+      "loss": 0.3837,
+      "step": 9026
+    },
+    {
+      "epoch": 2.70853027233851,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.478385298990812e-05,
+      "loss": 0.4953,
+      "step": 9027
+    },
+    {
+      "epoch": 2.708830369870208,
+      "grad_norm": 1.4140625,
+      "learning_rate": 6.476879048049405e-05,
+      "loss": 0.6085,
+      "step": 9028
+    },
+    {
+      "epoch": 2.7091304674019057,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.475372797107998e-05,
+      "loss": 0.525,
+      "step": 9029
+    },
+    {
+      "epoch": 2.7094305649336032,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.473866546166592e-05,
+      "loss": 0.5462,
+      "step": 9030
+    },
+    {
+      "epoch": 2.7097306624653013,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.472360295225184e-05,
+      "loss": 0.5171,
+      "step": 9031
+    },
+    {
+      "epoch": 2.7100307599969993,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.470854044283778e-05,
+      "loss": 0.598,
+      "step": 9032
+    },
+    {
+      "epoch": 2.710330857528697,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.469347793342371e-05,
+      "loss": 0.545,
+      "step": 9033
+    },
+    {
+      "epoch": 2.7106309550603944,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.467841542400964e-05,
+      "loss": 0.6265,
+      "step": 9034
+    },
+    {
+      "epoch": 2.7109310525920924,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.466335291459558e-05,
+      "loss": 0.4907,
+      "step": 9035
+    },
+    {
+      "epoch": 2.7112311501237905,
+      "grad_norm": 1.125,
+      "learning_rate": 6.46482904051815e-05,
+      "loss": 0.616,
+      "step": 9036
+    },
+    {
+      "epoch": 2.711531247655488,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.463322789576743e-05,
+      "loss": 0.5343,
+      "step": 9037
+    },
+    {
+      "epoch": 2.7118313451871856,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.461816538635337e-05,
+      "loss": 0.5248,
+      "step": 9038
+    },
+    {
+      "epoch": 2.7121314427188836,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.46031028769393e-05,
+      "loss": 0.5387,
+      "step": 9039
+    },
+    {
+      "epoch": 2.7124315402505816,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.458804036752524e-05,
+      "loss": 0.4933,
+      "step": 9040
+    },
+    {
+      "epoch": 2.712731637782279,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.457297785811117e-05,
+      "loss": 0.5033,
+      "step": 9041
+    },
+    {
+      "epoch": 2.713031735313977,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.455791534869709e-05,
+      "loss": 0.5328,
+      "step": 9042
+    },
+    {
+      "epoch": 2.713331832845675,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.454285283928303e-05,
+      "loss": 0.4917,
+      "step": 9043
+    },
+    {
+      "epoch": 2.713631930377373,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.452779032986896e-05,
+      "loss": 0.6129,
+      "step": 9044
+    },
+    {
+      "epoch": 2.7139320279090704,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.451272782045489e-05,
+      "loss": 0.583,
+      "step": 9045
+    },
+    {
+      "epoch": 2.714232125440768,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.449766531104083e-05,
+      "loss": 0.5557,
+      "step": 9046
+    },
+    {
+      "epoch": 2.714532222972466,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.448260280162675e-05,
+      "loss": 0.5015,
+      "step": 9047
+    },
+    {
+      "epoch": 2.714832320504164,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.446754029221268e-05,
+      "loss": 0.4782,
+      "step": 9048
+    },
+    {
+      "epoch": 2.7151324180358616,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.445247778279862e-05,
+      "loss": 0.5923,
+      "step": 9049
+    },
+    {
+      "epoch": 2.7154325155675596,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.443741527338455e-05,
+      "loss": 0.5737,
+      "step": 9050
+    },
+    {
+      "epoch": 2.715732613099257,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.442235276397049e-05,
+      "loss": 0.6009,
+      "step": 9051
+    },
+    {
+      "epoch": 2.716032710630955,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.44072902545564e-05,
+      "loss": 0.5737,
+      "step": 9052
+    },
+    {
+      "epoch": 2.7163328081626528,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.439222774514234e-05,
+      "loss": 0.5623,
+      "step": 9053
+    },
+    {
+      "epoch": 2.716632905694351,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.437716523572828e-05,
+      "loss": 0.5956,
+      "step": 9054
+    },
+    {
+      "epoch": 2.7169330032260484,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.436210272631421e-05,
+      "loss": 0.5547,
+      "step": 9055
+    },
+    {
+      "epoch": 2.7172331007577464,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.434704021690014e-05,
+      "loss": 0.5554,
+      "step": 9056
+    },
+    {
+      "epoch": 2.717533198289444,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.433197770748606e-05,
+      "loss": 0.5771,
+      "step": 9057
+    },
+    {
+      "epoch": 2.717833295821142,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.4316915198072e-05,
+      "loss": 0.5309,
+      "step": 9058
+    },
+    {
+      "epoch": 2.7181333933528395,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.430185268865794e-05,
+      "loss": 0.6521,
+      "step": 9059
+    },
+    {
+      "epoch": 2.7184334908845376,
+      "grad_norm": 1.0,
+      "learning_rate": 6.428679017924386e-05,
+      "loss": 0.5938,
+      "step": 9060
+    },
+    {
+      "epoch": 2.718733588416235,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.42717276698298e-05,
+      "loss": 0.5985,
+      "step": 9061
+    },
+    {
+      "epoch": 2.719033685947933,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.425666516041574e-05,
+      "loss": 0.4968,
+      "step": 9062
+    },
+    {
+      "epoch": 2.7193337834796307,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.424160265100166e-05,
+      "loss": 0.5119,
+      "step": 9063
+    },
+    {
+      "epoch": 2.7196338810113287,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.422654014158759e-05,
+      "loss": 0.4933,
+      "step": 9064
+    },
+    {
+      "epoch": 2.7199339785430263,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.421147763217352e-05,
+      "loss": 0.6031,
+      "step": 9065
+    },
+    {
+      "epoch": 2.7202340760747243,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.419641512275946e-05,
+      "loss": 0.5432,
+      "step": 9066
+    },
+    {
+      "epoch": 2.7205341736064224,
+      "grad_norm": 1.125,
+      "learning_rate": 6.41813526133454e-05,
+      "loss": 0.6033,
+      "step": 9067
+    },
+    {
+      "epoch": 2.72083427113812,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.416629010393131e-05,
+      "loss": 0.577,
+      "step": 9068
+    },
+    {
+      "epoch": 2.7211343686698175,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.415122759451725e-05,
+      "loss": 0.6306,
+      "step": 9069
+    },
+    {
+      "epoch": 2.7214344662015155,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.413616508510318e-05,
+      "loss": 0.485,
+      "step": 9070
+    },
+    {
+      "epoch": 2.7217345637332135,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.412110257568912e-05,
+      "loss": 0.5872,
+      "step": 9071
+    },
+    {
+      "epoch": 2.722034661264911,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.410604006627504e-05,
+      "loss": 0.6745,
+      "step": 9072
+    },
+    {
+      "epoch": 2.7223347587966087,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.409097755686097e-05,
+      "loss": 0.5706,
+      "step": 9073
+    },
+    {
+      "epoch": 2.7226348563283067,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.407591504744691e-05,
+      "loss": 0.5393,
+      "step": 9074
+    },
+    {
+      "epoch": 2.7229349538600047,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.406085253803284e-05,
+      "loss": 0.5267,
+      "step": 9075
+    },
+    {
+      "epoch": 2.7232350513917023,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.404579002861877e-05,
+      "loss": 0.5863,
+      "step": 9076
+    },
+    {
+      "epoch": 2.7235351489234,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.40307275192047e-05,
+      "loss": 0.5063,
+      "step": 9077
+    },
+    {
+      "epoch": 2.723835246455098,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.401566500979063e-05,
+      "loss": 0.5847,
+      "step": 9078
+    },
+    {
+      "epoch": 2.724135343986796,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.400060250037656e-05,
+      "loss": 0.5778,
+      "step": 9079
+    },
+    {
+      "epoch": 2.7244354415184935,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.39855399909625e-05,
+      "loss": 0.4815,
+      "step": 9080
+    },
+    {
+      "epoch": 2.724735539050191,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.397047748154843e-05,
+      "loss": 0.5492,
+      "step": 9081
+    },
+    {
+      "epoch": 2.725035636581889,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.395541497213437e-05,
+      "loss": 0.4937,
+      "step": 9082
+    },
+    {
+      "epoch": 2.725335734113587,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.394035246272029e-05,
+      "loss": 0.5487,
+      "step": 9083
+    },
+    {
+      "epoch": 2.7256358316452847,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.392528995330622e-05,
+      "loss": 0.5523,
+      "step": 9084
+    },
+    {
+      "epoch": 2.7259359291769822,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.391022744389216e-05,
+      "loss": 0.6072,
+      "step": 9085
+    },
+    {
+      "epoch": 2.7262360267086803,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.389516493447809e-05,
+      "loss": 0.4682,
+      "step": 9086
+    },
+    {
+      "epoch": 2.7265361242403783,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.388010242506401e-05,
+      "loss": 0.5802,
+      "step": 9087
+    },
+    {
+      "epoch": 2.726836221772076,
+      "grad_norm": 1.2890625,
+      "learning_rate": 6.386503991564995e-05,
+      "loss": 0.5619,
+      "step": 9088
+    },
+    {
+      "epoch": 2.727136319303774,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.384997740623588e-05,
+      "loss": 0.4789,
+      "step": 9089
+    },
+    {
+      "epoch": 2.7274364168354714,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.383491489682182e-05,
+      "loss": 0.5522,
+      "step": 9090
+    },
+    {
+      "epoch": 2.7277365143671695,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.381985238740775e-05,
+      "loss": 0.54,
+      "step": 9091
+    },
+    {
+      "epoch": 2.728036611898867,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.380478987799367e-05,
+      "loss": 0.5531,
+      "step": 9092
+    },
+    {
+      "epoch": 2.728336709430565,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.378972736857961e-05,
+      "loss": 0.5227,
+      "step": 9093
+    },
+    {
+      "epoch": 2.7286368069622626,
+      "grad_norm": 1.2265625,
+      "learning_rate": 6.377466485916554e-05,
+      "loss": 0.6812,
+      "step": 9094
+    },
+    {
+      "epoch": 2.7289369044939606,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.375960234975147e-05,
+      "loss": 0.4313,
+      "step": 9095
+    },
+    {
+      "epoch": 2.729237002025658,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.374453984033741e-05,
+      "loss": 0.4728,
+      "step": 9096
+    },
+    {
+      "epoch": 2.7295370995573562,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.372947733092334e-05,
+      "loss": 0.5966,
+      "step": 9097
+    },
+    {
+      "epoch": 2.729837197089054,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.371441482150928e-05,
+      "loss": 0.4793,
+      "step": 9098
+    },
+    {
+      "epoch": 2.730137294620752,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.369935231209519e-05,
+      "loss": 0.5089,
+      "step": 9099
+    },
+    {
+      "epoch": 2.7304373921524494,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.368428980268113e-05,
+      "loss": 0.4923,
+      "step": 9100
+    },
+    {
+      "epoch": 2.7307374896841474,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.366922729326707e-05,
+      "loss": 0.5209,
+      "step": 9101
+    },
+    {
+      "epoch": 2.731037587215845,
+      "grad_norm": 1.2578125,
+      "learning_rate": 6.365416478385298e-05,
+      "loss": 0.5802,
+      "step": 9102
+    },
+    {
+      "epoch": 2.731337684747543,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.363910227443892e-05,
+      "loss": 0.4805,
+      "step": 9103
+    },
+    {
+      "epoch": 2.7316377822792406,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.362403976502486e-05,
+      "loss": 0.5221,
+      "step": 9104
+    },
+    {
+      "epoch": 2.7319378798109386,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.360897725561079e-05,
+      "loss": 0.6427,
+      "step": 9105
+    },
+    {
+      "epoch": 2.7322379773426366,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.359391474619672e-05,
+      "loss": 0.5021,
+      "step": 9106
+    },
+    {
+      "epoch": 2.732538074874334,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.357885223678264e-05,
+      "loss": 0.5359,
+      "step": 9107
+    },
+    {
+      "epoch": 2.7328381724060318,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.356378972736858e-05,
+      "loss": 0.4795,
+      "step": 9108
+    },
+    {
+      "epoch": 2.73313826993773,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.354872721795452e-05,
+      "loss": 0.5979,
+      "step": 9109
+    },
+    {
+      "epoch": 2.733438367469428,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.353366470854044e-05,
+      "loss": 0.4911,
+      "step": 9110
+    },
+    {
+      "epoch": 2.7337384650011254,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.351860219912638e-05,
+      "loss": 0.6076,
+      "step": 9111
+    },
+    {
+      "epoch": 2.734038562532823,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.35035396897123e-05,
+      "loss": 0.5508,
+      "step": 9112
+    },
+    {
+      "epoch": 2.734338660064521,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.348847718029824e-05,
+      "loss": 0.542,
+      "step": 9113
+    },
+    {
+      "epoch": 2.734638757596219,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.347341467088417e-05,
+      "loss": 0.6371,
+      "step": 9114
+    },
+    {
+      "epoch": 2.7349388551279166,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.34583521614701e-05,
+      "loss": 0.5878,
+      "step": 9115
+    },
+    {
+      "epoch": 2.735238952659614,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.344328965205604e-05,
+      "loss": 0.4772,
+      "step": 9116
+    },
+    {
+      "epoch": 2.735539050191312,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.342822714264197e-05,
+      "loss": 0.5416,
+      "step": 9117
+    },
+    {
+      "epoch": 2.73583914772301,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.341316463322789e-05,
+      "loss": 0.5929,
+      "step": 9118
+    },
+    {
+      "epoch": 2.7361392452547078,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.339810212381383e-05,
+      "loss": 0.5958,
+      "step": 9119
+    },
+    {
+      "epoch": 2.7364393427864053,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.338303961439976e-05,
+      "loss": 0.5196,
+      "step": 9120
+    },
+    {
+      "epoch": 2.7367394403181033,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.33679771049857e-05,
+      "loss": 0.4434,
+      "step": 9121
+    },
+    {
+      "epoch": 2.7370395378498014,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.335291459557163e-05,
+      "loss": 0.513,
+      "step": 9122
+    },
+    {
+      "epoch": 2.737339635381499,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.333785208615755e-05,
+      "loss": 0.5057,
+      "step": 9123
+    },
+    {
+      "epoch": 2.737639732913197,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.33227895767435e-05,
+      "loss": 0.4226,
+      "step": 9124
+    },
+    {
+      "epoch": 2.7379398304448945,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.330772706732942e-05,
+      "loss": 0.5771,
+      "step": 9125
+    },
+    {
+      "epoch": 2.7382399279765925,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.329266455791535e-05,
+      "loss": 0.495,
+      "step": 9126
+    },
+    {
+      "epoch": 2.73854002550829,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.327760204850129e-05,
+      "loss": 0.6413,
+      "step": 9127
+    },
+    {
+      "epoch": 2.738840123039988,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.326253953908721e-05,
+      "loss": 0.5035,
+      "step": 9128
+    },
+    {
+      "epoch": 2.7391402205716857,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.324747702967314e-05,
+      "loss": 0.5024,
+      "step": 9129
+    },
+    {
+      "epoch": 2.7394403181033837,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.323241452025908e-05,
+      "loss": 0.5017,
+      "step": 9130
+    },
+    {
+      "epoch": 2.7397404156350813,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.321735201084501e-05,
+      "loss": 0.5649,
+      "step": 9131
+    },
+    {
+      "epoch": 2.7400405131667793,
+      "grad_norm": 0.92578125,
+      "learning_rate": 6.320228950143095e-05,
+      "loss": 0.6008,
+      "step": 9132
+    },
+    {
+      "epoch": 2.740340610698477,
+      "grad_norm": 1.3984375,
+      "learning_rate": 6.318722699201687e-05,
+      "loss": 0.6281,
+      "step": 9133
+    },
+    {
+      "epoch": 2.740640708230175,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.31721644826028e-05,
+      "loss": 0.4839,
+      "step": 9134
+    },
+    {
+      "epoch": 2.7409408057618725,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.315710197318874e-05,
+      "loss": 0.5339,
+      "step": 9135
+    },
+    {
+      "epoch": 2.7412409032935705,
+      "grad_norm": 1.125,
+      "learning_rate": 6.314203946377467e-05,
+      "loss": 0.5424,
+      "step": 9136
+    },
+    {
+      "epoch": 2.741541000825268,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.31269769543606e-05,
+      "loss": 0.5882,
+      "step": 9137
+    },
+    {
+      "epoch": 2.741841098356966,
+      "grad_norm": 1.125,
+      "learning_rate": 6.311191444494654e-05,
+      "loss": 0.5788,
+      "step": 9138
+    },
+    {
+      "epoch": 2.7421411958886637,
+      "grad_norm": 1.3671875,
+      "learning_rate": 6.309685193553246e-05,
+      "loss": 0.6252,
+      "step": 9139
+    },
+    {
+      "epoch": 2.7424412934203617,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.30817894261184e-05,
+      "loss": 0.69,
+      "step": 9140
+    },
+    {
+      "epoch": 2.7427413909520593,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.306672691670432e-05,
+      "loss": 0.5557,
+      "step": 9141
+    },
+    {
+      "epoch": 2.7430414884837573,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.305166440729026e-05,
+      "loss": 0.5095,
+      "step": 9142
+    },
+    {
+      "epoch": 2.743341586015455,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.30366018978762e-05,
+      "loss": 0.4509,
+      "step": 9143
+    },
+    {
+      "epoch": 2.743641683547153,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.302153938846212e-05,
+      "loss": 0.5771,
+      "step": 9144
+    },
+    {
+      "epoch": 2.743941781078851,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.300647687904805e-05,
+      "loss": 0.5779,
+      "step": 9145
+    },
+    {
+      "epoch": 2.7442418786105485,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.299141436963398e-05,
+      "loss": 0.6289,
+      "step": 9146
+    },
+    {
+      "epoch": 2.744541976142246,
+      "grad_norm": 0.9921875,
+      "learning_rate": 6.297635186021992e-05,
+      "loss": 0.5252,
+      "step": 9147
+    },
+    {
+      "epoch": 2.744842073673944,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.296128935080586e-05,
+      "loss": 0.5585,
+      "step": 9148
+    },
+    {
+      "epoch": 2.745142171205642,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.294622684139177e-05,
+      "loss": 0.5017,
+      "step": 9149
+    },
+    {
+      "epoch": 2.7454422687373397,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.293116433197771e-05,
+      "loss": 0.5896,
+      "step": 9150
+    },
+    {
+      "epoch": 2.7457423662690372,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.291610182256365e-05,
+      "loss": 0.5762,
+      "step": 9151
+    },
+    {
+      "epoch": 2.7460424638007352,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.290103931314958e-05,
+      "loss": 0.4935,
+      "step": 9152
+    },
+    {
+      "epoch": 2.7463425613324333,
+      "grad_norm": 1.125,
+      "learning_rate": 6.28859768037355e-05,
+      "loss": 0.5511,
+      "step": 9153
+    },
+    {
+      "epoch": 2.746642658864131,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.287091429432143e-05,
+      "loss": 0.5828,
+      "step": 9154
+    },
+    {
+      "epoch": 2.7469427563958284,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.285585178490737e-05,
+      "loss": 0.4927,
+      "step": 9155
+    },
+    {
+      "epoch": 2.7472428539275264,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.28407892754933e-05,
+      "loss": 0.5902,
+      "step": 9156
+    },
+    {
+      "epoch": 2.7475429514592244,
+      "grad_norm": 0.94921875,
+      "learning_rate": 6.282572676607923e-05,
+      "loss": 0.559,
+      "step": 9157
+    },
+    {
+      "epoch": 2.747843048990922,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.281066425666517e-05,
+      "loss": 0.5057,
+      "step": 9158
+    },
+    {
+      "epoch": 2.7481431465226196,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.279560174725109e-05,
+      "loss": 0.5404,
+      "step": 9159
+    },
+    {
+      "epoch": 2.7484432440543176,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.278053923783702e-05,
+      "loss": 0.4974,
+      "step": 9160
+    },
+    {
+      "epoch": 2.7487433415860156,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.276547672842296e-05,
+      "loss": 0.5214,
+      "step": 9161
+    },
+    {
+      "epoch": 2.749043439117713,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.275041421900889e-05,
+      "loss": 0.4553,
+      "step": 9162
+    },
+    {
+      "epoch": 2.7493435366494112,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.273535170959483e-05,
+      "loss": 0.4935,
+      "step": 9163
+    },
+    {
+      "epoch": 2.749643634181109,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.272028920018075e-05,
+      "loss": 0.5239,
+      "step": 9164
+    },
+    {
+      "epoch": 2.749943731712807,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.270522669076668e-05,
+      "loss": 0.485,
+      "step": 9165
+    },
+    {
+      "epoch": 2.7502438292445044,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.269016418135262e-05,
+      "loss": 0.4701,
+      "step": 9166
+    },
+    {
+      "epoch": 2.7505439267762024,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.267510167193855e-05,
+      "loss": 0.4696,
+      "step": 9167
+    },
+    {
+      "epoch": 2.7508440243079,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.266003916252447e-05,
+      "loss": 0.5478,
+      "step": 9168
+    },
+    {
+      "epoch": 2.751144121839598,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.264497665311041e-05,
+      "loss": 0.4997,
+      "step": 9169
+    },
+    {
+      "epoch": 2.7514442193712956,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.262991414369634e-05,
+      "loss": 0.5149,
+      "step": 9170
+    },
+    {
+      "epoch": 2.7517443169029936,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.261485163428228e-05,
+      "loss": 0.5313,
+      "step": 9171
+    },
+    {
+      "epoch": 2.752044414434691,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.259978912486821e-05,
+      "loss": 0.596,
+      "step": 9172
+    },
+    {
+      "epoch": 2.752344511966389,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.258472661545413e-05,
+      "loss": 0.5209,
+      "step": 9173
+    },
+    {
+      "epoch": 2.7526446094980868,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.256966410604007e-05,
+      "loss": 0.5668,
+      "step": 9174
+    },
+    {
+      "epoch": 2.7529447070297848,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.2554601596626e-05,
+      "loss": 0.5667,
+      "step": 9175
+    },
+    {
+      "epoch": 2.7532448045614824,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.253953908721193e-05,
+      "loss": 0.5817,
+      "step": 9176
+    },
+    {
+      "epoch": 2.7535449020931804,
+      "grad_norm": 0.93359375,
+      "learning_rate": 6.252447657779787e-05,
+      "loss": 0.516,
+      "step": 9177
+    },
+    {
+      "epoch": 2.753844999624878,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.25094140683838e-05,
+      "loss": 0.5017,
+      "step": 9178
+    },
+    {
+      "epoch": 2.754145097156576,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.249435155896974e-05,
+      "loss": 0.5591,
+      "step": 9179
+    },
+    {
+      "epoch": 2.7544451946882735,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.247928904955566e-05,
+      "loss": 0.4845,
+      "step": 9180
+    },
+    {
+      "epoch": 2.7547452922199716,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.246422654014159e-05,
+      "loss": 0.5859,
+      "step": 9181
+    },
+    {
+      "epoch": 2.755045389751669,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.244916403072753e-05,
+      "loss": 0.6288,
+      "step": 9182
+    },
+    {
+      "epoch": 2.755345487283367,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.243410152131344e-05,
+      "loss": 0.4817,
+      "step": 9183
+    },
+    {
+      "epoch": 2.755645584815065,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.241903901189938e-05,
+      "loss": 0.5743,
+      "step": 9184
+    },
+    {
+      "epoch": 2.7559456823467627,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.240397650248532e-05,
+      "loss": 0.4717,
+      "step": 9185
+    },
+    {
+      "epoch": 2.7562457798784603,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.238891399307125e-05,
+      "loss": 0.5683,
+      "step": 9186
+    },
+    {
+      "epoch": 2.7565458774101583,
+      "grad_norm": 1.1953125,
+      "learning_rate": 6.237385148365718e-05,
+      "loss": 0.4661,
+      "step": 9187
+    },
+    {
+      "epoch": 2.7568459749418563,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.23587889742431e-05,
+      "loss": 0.5956,
+      "step": 9188
+    },
+    {
+      "epoch": 2.757146072473554,
+      "grad_norm": 0.97265625,
+      "learning_rate": 6.234372646482904e-05,
+      "loss": 0.5816,
+      "step": 9189
+    },
+    {
+      "epoch": 2.7574461700052515,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.232866395541498e-05,
+      "loss": 0.6308,
+      "step": 9190
+    },
+    {
+      "epoch": 2.7577462675369495,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.23136014460009e-05,
+      "loss": 0.6107,
+      "step": 9191
+    },
+    {
+      "epoch": 2.7580463650686475,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.229853893658684e-05,
+      "loss": 0.6073,
+      "step": 9192
+    },
+    {
+      "epoch": 2.758346462600345,
+      "grad_norm": 1.0,
+      "learning_rate": 6.228347642717276e-05,
+      "loss": 0.5012,
+      "step": 9193
+    },
+    {
+      "epoch": 2.7586465601320427,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.22684139177587e-05,
+      "loss": 0.4767,
+      "step": 9194
+    },
+    {
+      "epoch": 2.7589466576637407,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.225335140834463e-05,
+      "loss": 0.4646,
+      "step": 9195
+    },
+    {
+      "epoch": 2.7592467551954387,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.223828889893056e-05,
+      "loss": 0.5304,
+      "step": 9196
+    },
+    {
+      "epoch": 2.7595468527271363,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.22232263895165e-05,
+      "loss": 0.4956,
+      "step": 9197
+    },
+    {
+      "epoch": 2.759846950258834,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.220816388010244e-05,
+      "loss": 0.4412,
+      "step": 9198
+    },
+    {
+      "epoch": 2.760147047790532,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.219310137068835e-05,
+      "loss": 0.5028,
+      "step": 9199
+    },
+    {
+      "epoch": 2.76044714532223,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.217803886127429e-05,
+      "loss": 0.6019,
+      "step": 9200
+    },
+    {
+      "epoch": 2.7607472428539275,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.216297635186022e-05,
+      "loss": 0.5692,
+      "step": 9201
+    },
+    {
+      "epoch": 2.7610473403856255,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.214791384244616e-05,
+      "loss": 0.5538,
+      "step": 9202
+    },
+    {
+      "epoch": 2.761347437917323,
+      "grad_norm": 1.25,
+      "learning_rate": 6.213285133303209e-05,
+      "loss": 0.5385,
+      "step": 9203
+    },
+    {
+      "epoch": 2.761647535449021,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.211778882361801e-05,
+      "loss": 0.4991,
+      "step": 9204
+    },
+    {
+      "epoch": 2.7619476329807187,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.210272631420395e-05,
+      "loss": 0.5964,
+      "step": 9205
+    },
+    {
+      "epoch": 2.7622477305124167,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.208766380478988e-05,
+      "loss": 0.443,
+      "step": 9206
+    },
+    {
+      "epoch": 2.7625478280441143,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.20726012953758e-05,
+      "loss": 0.494,
+      "step": 9207
+    },
+    {
+      "epoch": 2.7628479255758123,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.205753878596175e-05,
+      "loss": 0.5357,
+      "step": 9208
+    },
+    {
+      "epoch": 2.76314802310751,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.204247627654767e-05,
+      "loss": 0.6417,
+      "step": 9209
+    },
+    {
+      "epoch": 2.763448120639208,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.20274137671336e-05,
+      "loss": 0.4521,
+      "step": 9210
+    },
+    {
+      "epoch": 2.7637482181709054,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.201235125771954e-05,
+      "loss": 0.4986,
+      "step": 9211
+    },
+    {
+      "epoch": 2.7640483157026035,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.199728874830547e-05,
+      "loss": 0.4955,
+      "step": 9212
+    },
+    {
+      "epoch": 2.764348413234301,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.198222623889141e-05,
+      "loss": 0.4853,
+      "step": 9213
+    },
+    {
+      "epoch": 2.764648510765999,
+      "grad_norm": 1.203125,
+      "learning_rate": 6.196716372947733e-05,
+      "loss": 0.5418,
+      "step": 9214
+    },
+    {
+      "epoch": 2.7649486082976966,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.195210122006326e-05,
+      "loss": 0.497,
+      "step": 9215
+    },
+    {
+      "epoch": 2.7652487058293946,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.19370387106492e-05,
+      "loss": 0.449,
+      "step": 9216
+    },
+    {
+      "epoch": 2.765548803361092,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.192197620123513e-05,
+      "loss": 0.5579,
+      "step": 9217
+    },
+    {
+      "epoch": 2.7658489008927902,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.190691369182106e-05,
+      "loss": 0.4749,
+      "step": 9218
+    },
+    {
+      "epoch": 2.766148998424488,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.1891851182407e-05,
+      "loss": 0.4969,
+      "step": 9219
+    },
+    {
+      "epoch": 2.766449095956186,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.187678867299292e-05,
+      "loss": 0.5497,
+      "step": 9220
+    },
+    {
+      "epoch": 2.7667491934878834,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.186172616357886e-05,
+      "loss": 0.4591,
+      "step": 9221
+    },
+    {
+      "epoch": 2.7670492910195814,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.184666365416479e-05,
+      "loss": 0.5347,
+      "step": 9222
+    },
+    {
+      "epoch": 2.7673493885512794,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.183160114475072e-05,
+      "loss": 0.5095,
+      "step": 9223
+    },
+    {
+      "epoch": 2.767649486082977,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.181653863533666e-05,
+      "loss": 0.5225,
+      "step": 9224
+    },
+    {
+      "epoch": 2.7679495836146746,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.180147612592258e-05,
+      "loss": 0.6135,
+      "step": 9225
+    },
+    {
+      "epoch": 2.7682496811463726,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.178641361650851e-05,
+      "loss": 0.5403,
+      "step": 9226
+    },
+    {
+      "epoch": 2.7685497786780706,
+      "grad_norm": 1.2421875,
+      "learning_rate": 6.177135110709445e-05,
+      "loss": 0.4998,
+      "step": 9227
+    },
+    {
+      "epoch": 2.768849876209768,
+      "grad_norm": 1.296875,
+      "learning_rate": 6.175628859768038e-05,
+      "loss": 0.5888,
+      "step": 9228
+    },
+    {
+      "epoch": 2.7691499737414658,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.174122608826632e-05,
+      "loss": 0.5623,
+      "step": 9229
+    },
+    {
+      "epoch": 2.769450071273164,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.172616357885223e-05,
+      "loss": 0.6211,
+      "step": 9230
+    },
+    {
+      "epoch": 2.769750168804862,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.171110106943817e-05,
+      "loss": 0.5155,
+      "step": 9231
+    },
+    {
+      "epoch": 2.7700502663365594,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.169603856002411e-05,
+      "loss": 0.5157,
+      "step": 9232
+    },
+    {
+      "epoch": 2.770350363868257,
+      "grad_norm": 1.25,
+      "learning_rate": 6.168097605061004e-05,
+      "loss": 0.554,
+      "step": 9233
+    },
+    {
+      "epoch": 2.770650461399955,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.166591354119596e-05,
+      "loss": 0.521,
+      "step": 9234
+    },
+    {
+      "epoch": 2.770950558931653,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.165085103178189e-05,
+      "loss": 0.5028,
+      "step": 9235
+    },
+    {
+      "epoch": 2.7712506564633506,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.163578852236783e-05,
+      "loss": 0.563,
+      "step": 9236
+    },
+    {
+      "epoch": 2.771550753995048,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.162072601295376e-05,
+      "loss": 0.456,
+      "step": 9237
+    },
+    {
+      "epoch": 2.771850851526746,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.160566350353969e-05,
+      "loss": 0.4003,
+      "step": 9238
+    },
+    {
+      "epoch": 2.772150949058444,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.159060099412563e-05,
+      "loss": 0.528,
+      "step": 9239
+    },
+    {
+      "epoch": 2.7724510465901417,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.157553848471157e-05,
+      "loss": 0.4812,
+      "step": 9240
+    },
+    {
+      "epoch": 2.7727511441218398,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.156047597529748e-05,
+      "loss": 0.6076,
+      "step": 9241
+    },
+    {
+      "epoch": 2.7730512416535373,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.154541346588342e-05,
+      "loss": 0.5083,
+      "step": 9242
+    },
+    {
+      "epoch": 2.7733513391852354,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.153035095646935e-05,
+      "loss": 0.5055,
+      "step": 9243
+    },
+    {
+      "epoch": 2.773651436716933,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.151528844705529e-05,
+      "loss": 0.5357,
+      "step": 9244
+    },
+    {
+      "epoch": 2.773951534248631,
+      "grad_norm": 1.3515625,
+      "learning_rate": 6.150022593764121e-05,
+      "loss": 0.5622,
+      "step": 9245
+    },
+    {
+      "epoch": 2.7742516317803285,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.148516342822714e-05,
+      "loss": 0.5195,
+      "step": 9246
+    },
+    {
+      "epoch": 2.7745517293120265,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.147010091881308e-05,
+      "loss": 0.6346,
+      "step": 9247
+    },
+    {
+      "epoch": 2.774851826843724,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.145503840939901e-05,
+      "loss": 0.5438,
+      "step": 9248
+    },
+    {
+      "epoch": 2.775151924375422,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.143997589998493e-05,
+      "loss": 0.4787,
+      "step": 9249
+    },
+    {
+      "epoch": 2.7754520219071197,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.142491339057087e-05,
+      "loss": 0.522,
+      "step": 9250
+    },
+    {
+      "epoch": 2.7757521194388177,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.14098508811568e-05,
+      "loss": 0.6077,
+      "step": 9251
+    },
+    {
+      "epoch": 2.7760522169705153,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.139478837174274e-05,
+      "loss": 0.5761,
+      "step": 9252
+    },
+    {
+      "epoch": 2.7763523145022133,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.137972586232867e-05,
+      "loss": 0.5173,
+      "step": 9253
+    },
+    {
+      "epoch": 2.776652412033911,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.13646633529146e-05,
+      "loss": 0.6104,
+      "step": 9254
+    },
+    {
+      "epoch": 2.776952509565609,
+      "grad_norm": 0.8984375,
+      "learning_rate": 6.134960084350053e-05,
+      "loss": 0.4349,
+      "step": 9255
+    },
+    {
+      "epoch": 2.7772526070973065,
+      "grad_norm": 1.0,
+      "learning_rate": 6.133453833408646e-05,
+      "loss": 0.6098,
+      "step": 9256
+    },
+    {
+      "epoch": 2.7775527046290045,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.131947582467239e-05,
+      "loss": 0.5541,
+      "step": 9257
+    },
+    {
+      "epoch": 2.777852802160702,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.130441331525833e-05,
+      "loss": 0.4467,
+      "step": 9258
+    },
+    {
+      "epoch": 2.7781528996924,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.128935080584426e-05,
+      "loss": 0.5591,
+      "step": 9259
+    },
+    {
+      "epoch": 2.7784529972240977,
+      "grad_norm": 0.88671875,
+      "learning_rate": 6.12742882964302e-05,
+      "loss": 0.4688,
+      "step": 9260
+    },
+    {
+      "epoch": 2.7787530947557957,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.125922578701612e-05,
+      "loss": 0.5361,
+      "step": 9261
+    },
+    {
+      "epoch": 2.7790531922874937,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.124416327760205e-05,
+      "loss": 0.5045,
+      "step": 9262
+    },
+    {
+      "epoch": 2.7793532898191913,
+      "grad_norm": 1.0,
+      "learning_rate": 6.122910076818799e-05,
+      "loss": 0.5684,
+      "step": 9263
+    },
+    {
+      "epoch": 2.779653387350889,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.12140382587739e-05,
+      "loss": 0.5744,
+      "step": 9264
+    },
+    {
+      "epoch": 2.779953484882587,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.119897574935984e-05,
+      "loss": 0.4462,
+      "step": 9265
+    },
+    {
+      "epoch": 2.780253582414285,
+      "grad_norm": 1.25,
+      "learning_rate": 6.118391323994578e-05,
+      "loss": 0.5135,
+      "step": 9266
+    },
+    {
+      "epoch": 2.7805536799459825,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.116885073053171e-05,
+      "loss": 0.5514,
+      "step": 9267
+    },
+    {
+      "epoch": 2.78085377747768,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.115378822111764e-05,
+      "loss": 0.4779,
+      "step": 9268
+    },
+    {
+      "epoch": 2.781153875009378,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.113872571170358e-05,
+      "loss": 0.5139,
+      "step": 9269
+    },
+    {
+      "epoch": 2.781453972541076,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.11236632022895e-05,
+      "loss": 0.5491,
+      "step": 9270
+    },
+    {
+      "epoch": 2.7817540700727736,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.110860069287544e-05,
+      "loss": 0.6184,
+      "step": 9271
+    },
+    {
+      "epoch": 2.782054167604471,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.109353818346136e-05,
+      "loss": 0.5361,
+      "step": 9272
+    },
+    {
+      "epoch": 2.7823542651361692,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.10784756740473e-05,
+      "loss": 0.5258,
+      "step": 9273
+    },
+    {
+      "epoch": 2.7826543626678673,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.106341316463324e-05,
+      "loss": 0.5175,
+      "step": 9274
+    },
+    {
+      "epoch": 2.782954460199565,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.104835065521916e-05,
+      "loss": 0.4822,
+      "step": 9275
+    },
+    {
+      "epoch": 2.7832545577312624,
+      "grad_norm": 1.1015625,
+      "learning_rate": 6.103328814580509e-05,
+      "loss": 0.487,
+      "step": 9276
+    },
+    {
+      "epoch": 2.7835546552629604,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.1018225636391025e-05,
+      "loss": 0.5561,
+      "step": 9277
+    },
+    {
+      "epoch": 2.7838547527946584,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.100316312697696e-05,
+      "loss": 0.5575,
+      "step": 9278
+    },
+    {
+      "epoch": 2.784154850326356,
+      "grad_norm": 0.9609375,
+      "learning_rate": 6.098810061756289e-05,
+      "loss": 0.539,
+      "step": 9279
+    },
+    {
+      "epoch": 2.784454947858054,
+      "grad_norm": 0.99609375,
+      "learning_rate": 6.097303810814882e-05,
+      "loss": 0.5473,
+      "step": 9280
+    },
+    {
+      "epoch": 2.7847550453897516,
+      "grad_norm": 1.1875,
+      "learning_rate": 6.095797559873475e-05,
+      "loss": 0.5416,
+      "step": 9281
+    },
+    {
+      "epoch": 2.7850551429214496,
+      "grad_norm": 1.265625,
+      "learning_rate": 6.0942913089320686e-05,
+      "loss": 0.5591,
+      "step": 9282
+    },
+    {
+      "epoch": 2.785355240453147,
+      "grad_norm": 1.25,
+      "learning_rate": 6.092785057990662e-05,
+      "loss": 0.4956,
+      "step": 9283
+    },
+    {
+      "epoch": 2.785655337984845,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.0912788070492546e-05,
+      "loss": 0.6459,
+      "step": 9284
+    },
+    {
+      "epoch": 2.785955435516543,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.089772556107848e-05,
+      "loss": 0.6463,
+      "step": 9285
+    },
+    {
+      "epoch": 2.786255533048241,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.088266305166441e-05,
+      "loss": 0.5822,
+      "step": 9286
+    },
+    {
+      "epoch": 2.7865556305799384,
+      "grad_norm": 1.125,
+      "learning_rate": 6.086760054225035e-05,
+      "loss": 0.4933,
+      "step": 9287
+    },
+    {
+      "epoch": 2.7868557281116364,
+      "grad_norm": 1.0546875,
+      "learning_rate": 6.0852538032836274e-05,
+      "loss": 0.4616,
+      "step": 9288
+    },
+    {
+      "epoch": 2.787155825643334,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.083747552342221e-05,
+      "loss": 0.5279,
+      "step": 9289
+    },
+    {
+      "epoch": 2.787455923175032,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.082241301400814e-05,
+      "loss": 0.4906,
+      "step": 9290
+    },
+    {
+      "epoch": 2.7877560207067296,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.080735050459406e-05,
+      "loss": 0.5866,
+      "step": 9291
+    },
+    {
+      "epoch": 2.7880561182384276,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.0792287995179994e-05,
+      "loss": 0.6122,
+      "step": 9292
+    },
+    {
+      "epoch": 2.788356215770125,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.0777225485765934e-05,
+      "loss": 0.5252,
+      "step": 9293
+    },
+    {
+      "epoch": 2.788656313301823,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.076216297635187e-05,
+      "loss": 0.599,
+      "step": 9294
+    },
+    {
+      "epoch": 2.7889564108335207,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.074710046693779e-05,
+      "loss": 0.5333,
+      "step": 9295
+    },
+    {
+      "epoch": 2.7892565083652188,
+      "grad_norm": 1.21875,
+      "learning_rate": 6.073203795752372e-05,
+      "loss": 0.6351,
+      "step": 9296
+    },
+    {
+      "epoch": 2.7895566058969163,
+      "grad_norm": 0.89453125,
+      "learning_rate": 6.071697544810966e-05,
+      "loss": 0.5036,
+      "step": 9297
+    },
+    {
+      "epoch": 2.7898567034286144,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.0701912938695595e-05,
+      "loss": 0.5974,
+      "step": 9298
+    },
+    {
+      "epoch": 2.790156800960312,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.0686850429281515e-05,
+      "loss": 0.6417,
+      "step": 9299
+    },
+    {
+      "epoch": 2.79045689849201,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.067178791986745e-05,
+      "loss": 0.445,
+      "step": 9300
+    },
+    {
+      "epoch": 2.790756996023708,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.065672541045338e-05,
+      "loss": 0.5388,
+      "step": 9301
+    },
+    {
+      "epoch": 2.7910570935554055,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.064166290103932e-05,
+      "loss": 0.5999,
+      "step": 9302
+    },
+    {
+      "epoch": 2.791357191087103,
+      "grad_norm": 1.0,
+      "learning_rate": 6.062660039162524e-05,
+      "loss": 0.5113,
+      "step": 9303
+    },
+    {
+      "epoch": 2.791657288618801,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.0611537882211176e-05,
+      "loss": 0.5037,
+      "step": 9304
+    },
+    {
+      "epoch": 2.791957386150499,
+      "grad_norm": 1.140625,
+      "learning_rate": 6.059647537279711e-05,
+      "loss": 0.5299,
+      "step": 9305
+    },
+    {
+      "epoch": 2.7922574836821967,
+      "grad_norm": 1.0078125,
+      "learning_rate": 6.058141286338305e-05,
+      "loss": 0.4499,
+      "step": 9306
+    },
+    {
+      "epoch": 2.7925575812138943,
+      "grad_norm": 1.125,
+      "learning_rate": 6.056635035396897e-05,
+      "loss": 0.5283,
+      "step": 9307
+    },
+    {
+      "epoch": 2.7928576787455923,
+      "grad_norm": 0.96875,
+      "learning_rate": 6.0551287844554903e-05,
+      "loss": 0.4279,
+      "step": 9308
+    },
+    {
+      "epoch": 2.7931577762772903,
+      "grad_norm": 0.98828125,
+      "learning_rate": 6.053622533514084e-05,
+      "loss": 0.4966,
+      "step": 9309
+    },
+    {
+      "epoch": 2.793457873808988,
+      "grad_norm": 1.0,
+      "learning_rate": 6.052116282572677e-05,
+      "loss": 0.4442,
+      "step": 9310
+    },
+    {
+      "epoch": 2.7937579713406855,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.05061003163127e-05,
+      "loss": 0.4494,
+      "step": 9311
+    },
+    {
+      "epoch": 2.7940580688723835,
+      "grad_norm": 1.0859375,
+      "learning_rate": 6.049103780689863e-05,
+      "loss": 0.607,
+      "step": 9312
+    },
+    {
+      "epoch": 2.7943581664040815,
+      "grad_norm": 1.15625,
+      "learning_rate": 6.0475975297484564e-05,
+      "loss": 0.6837,
+      "step": 9313
+    },
+    {
+      "epoch": 2.794658263935779,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.04609127880705e-05,
+      "loss": 0.4935,
+      "step": 9314
+    },
+    {
+      "epoch": 2.7949583614674767,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.0445850278656425e-05,
+      "loss": 0.4422,
+      "step": 9315
+    },
+    {
+      "epoch": 2.7952584589991747,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.043078776924236e-05,
+      "loss": 0.5247,
+      "step": 9316
+    },
+    {
+      "epoch": 2.7955585565308727,
+      "grad_norm": 1.078125,
+      "learning_rate": 6.041572525982829e-05,
+      "loss": 0.4551,
+      "step": 9317
+    },
+    {
+      "epoch": 2.7958586540625703,
+      "grad_norm": 1.25,
+      "learning_rate": 6.040066275041422e-05,
+      "loss": 0.5598,
+      "step": 9318
+    },
+    {
+      "epoch": 2.7961587515942683,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.038560024100015e-05,
+      "loss": 0.542,
+      "step": 9319
+    },
+    {
+      "epoch": 2.796458849125966,
+      "grad_norm": 1.0625,
+      "learning_rate": 6.0370537731586085e-05,
+      "loss": 0.5499,
+      "step": 9320
+    },
+    {
+      "epoch": 2.796758946657664,
+      "grad_norm": 1.109375,
+      "learning_rate": 6.035547522217202e-05,
+      "loss": 0.6477,
+      "step": 9321
+    },
+    {
+      "epoch": 2.7970590441893615,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.0340412712757946e-05,
+      "loss": 0.527,
+      "step": 9322
+    },
+    {
+      "epoch": 2.7973591417210595,
+      "grad_norm": 1.1171875,
+      "learning_rate": 6.032535020334388e-05,
+      "loss": 0.615,
+      "step": 9323
+    },
+    {
+      "epoch": 2.797659239252757,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.031028769392981e-05,
+      "loss": 0.4679,
+      "step": 9324
+    },
+    {
+      "epoch": 2.797959336784455,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.0295225184515746e-05,
+      "loss": 0.5025,
+      "step": 9325
+    },
+    {
+      "epoch": 2.7982594343161526,
+      "grad_norm": 1.1484375,
+      "learning_rate": 6.028016267510167e-05,
+      "loss": 0.6033,
+      "step": 9326
+    },
+    {
+      "epoch": 2.7985595318478507,
+      "grad_norm": 1.1328125,
+      "learning_rate": 6.026510016568761e-05,
+      "loss": 0.4932,
+      "step": 9327
+    },
+    {
+      "epoch": 2.7988596293795482,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.025003765627354e-05,
+      "loss": 0.5291,
+      "step": 9328
+    },
+    {
+      "epoch": 2.7991597269112463,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.0234975146859474e-05,
+      "loss": 0.5869,
+      "step": 9329
+    },
+    {
+      "epoch": 2.799459824442944,
+      "grad_norm": 0.9453125,
+      "learning_rate": 6.0219912637445394e-05,
+      "loss": 0.5195,
+      "step": 9330
+    },
+    {
+      "epoch": 2.799759921974642,
+      "grad_norm": 1.0234375,
+      "learning_rate": 6.0204850128031334e-05,
+      "loss": 0.5224,
+      "step": 9331
+    },
+    {
+      "epoch": 2.8000600195063394,
+      "grad_norm": 0.953125,
+      "learning_rate": 6.018978761861727e-05,
+      "loss": 0.5016,
+      "step": 9332
+    },
+    {
+      "epoch": 2.8003601170380374,
+      "grad_norm": 1.25,
+      "learning_rate": 6.01747251092032e-05,
+      "loss": 0.571,
+      "step": 9333
+    },
+    {
+      "epoch": 2.800660214569735,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.015966259978912e-05,
+      "loss": 0.4914,
+      "step": 9334
+    },
+    {
+      "epoch": 2.800960312101433,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.014460009037506e-05,
+      "loss": 0.502,
+      "step": 9335
+    },
+    {
+      "epoch": 2.8012604096331306,
+      "grad_norm": 1.171875,
+      "learning_rate": 6.0129537580960995e-05,
+      "loss": 0.4649,
+      "step": 9336
+    },
+    {
+      "epoch": 2.8015605071648286,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.011447507154693e-05,
+      "loss": 0.5397,
+      "step": 9337
+    },
+    {
+      "epoch": 2.801860604696526,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.009941256213285e-05,
+      "loss": 0.4932,
+      "step": 9338
+    },
+    {
+      "epoch": 2.802160702228224,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.008435005271878e-05,
+      "loss": 0.5554,
+      "step": 9339
+    },
+    {
+      "epoch": 2.8024607997599222,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.006928754330472e-05,
+      "loss": 0.5565,
+      "step": 9340
+    },
+    {
+      "epoch": 2.80276089729162,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.0054225033890656e-05,
+      "loss": 0.5221,
+      "step": 9341
+    },
+    {
+      "epoch": 2.8030609948233174,
+      "grad_norm": 1.1796875,
+      "learning_rate": 6.0039162524476576e-05,
+      "loss": 0.4763,
+      "step": 9342
+    },
+    {
+      "epoch": 2.8033610923550154,
+      "grad_norm": 1.1640625,
+      "learning_rate": 6.002410001506251e-05,
+      "loss": 0.5698,
+      "step": 9343
+    },
+    {
+      "epoch": 2.8036611898867134,
+      "grad_norm": 1.0703125,
+      "learning_rate": 6.000903750564845e-05,
+      "loss": 0.4981,
+      "step": 9344
+    },
+    {
+      "epoch": 2.803961287418411,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.999397499623437e-05,
+      "loss": 0.4802,
+      "step": 9345
+    },
+    {
+      "epoch": 2.8042613849501086,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.99789124868203e-05,
+      "loss": 0.5847,
+      "step": 9346
+    },
+    {
+      "epoch": 2.8045614824818066,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.9963849977406237e-05,
+      "loss": 0.52,
+      "step": 9347
+    },
+    {
+      "epoch": 2.8048615800135046,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.994878746799218e-05,
+      "loss": 0.4705,
+      "step": 9348
+    },
+    {
+      "epoch": 2.805161677545202,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.99337249585781e-05,
+      "loss": 0.5486,
+      "step": 9349
+    },
+    {
+      "epoch": 2.8054617750768998,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.991866244916403e-05,
+      "loss": 0.491,
+      "step": 9350
+    },
+    {
+      "epoch": 2.8057618726085978,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.9903599939749964e-05,
+      "loss": 0.5471,
+      "step": 9351
+    },
+    {
+      "epoch": 2.806061970140296,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.98885374303359e-05,
+      "loss": 0.4065,
+      "step": 9352
+    },
+    {
+      "epoch": 2.8063620676719934,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.9873474920921824e-05,
+      "loss": 0.5679,
+      "step": 9353
+    },
+    {
+      "epoch": 2.806662165203691,
+      "grad_norm": 1.0,
+      "learning_rate": 5.985841241150776e-05,
+      "loss": 0.4403,
+      "step": 9354
+    },
+    {
+      "epoch": 2.806962262735389,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.984334990209369e-05,
+      "loss": 0.454,
+      "step": 9355
+    },
+    {
+      "epoch": 2.807262360267087,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.9828287392679625e-05,
+      "loss": 0.5827,
+      "step": 9356
+    },
+    {
+      "epoch": 2.8075624577987845,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.981322488326555e-05,
+      "loss": 0.5792,
+      "step": 9357
+    },
+    {
+      "epoch": 2.8078625553304826,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.9798162373851485e-05,
+      "loss": 0.5378,
+      "step": 9358
+    },
+    {
+      "epoch": 2.80816265286218,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.978309986443742e-05,
+      "loss": 0.5631,
+      "step": 9359
+    },
+    {
+      "epoch": 2.808462750393878,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.976803735502335e-05,
+      "loss": 0.5601,
+      "step": 9360
+    },
+    {
+      "epoch": 2.8087628479255757,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.975297484560928e-05,
+      "loss": 0.4396,
+      "step": 9361
+    },
+    {
+      "epoch": 2.8090629454572738,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.973791233619521e-05,
+      "loss": 0.5337,
+      "step": 9362
+    },
+    {
+      "epoch": 2.8093630429889713,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.9722849826781146e-05,
+      "loss": 0.5067,
+      "step": 9363
+    },
+    {
+      "epoch": 2.8096631405206693,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.970778731736708e-05,
+      "loss": 0.5547,
+      "step": 9364
+    },
+    {
+      "epoch": 2.809963238052367,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.9692724807953006e-05,
+      "loss": 0.5582,
+      "step": 9365
+    },
+    {
+      "epoch": 2.810263335584065,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.967766229853894e-05,
+      "loss": 0.5046,
+      "step": 9366
+    },
+    {
+      "epoch": 2.8105634331157625,
+      "grad_norm": 1.125,
+      "learning_rate": 5.966259978912487e-05,
+      "loss": 0.5892,
+      "step": 9367
+    },
+    {
+      "epoch": 2.8108635306474605,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.964753727971081e-05,
+      "loss": 0.5441,
+      "step": 9368
+    },
+    {
+      "epoch": 2.811163628179158,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.9632474770296734e-05,
+      "loss": 0.6123,
+      "step": 9369
+    },
+    {
+      "epoch": 2.811463725710856,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.961741226088267e-05,
+      "loss": 0.4803,
+      "step": 9370
+    },
+    {
+      "epoch": 2.8117638232425537,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.96023497514686e-05,
+      "loss": 0.5934,
+      "step": 9371
+    },
+    {
+      "epoch": 2.8120639207742517,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.958728724205452e-05,
+      "loss": 0.5357,
+      "step": 9372
+    },
+    {
+      "epoch": 2.8123640183059493,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.957222473264046e-05,
+      "loss": 0.5309,
+      "step": 9373
+    },
+    {
+      "epoch": 2.8126641158376473,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.9557162223226394e-05,
+      "loss": 0.4791,
+      "step": 9374
+    },
+    {
+      "epoch": 2.812964213369345,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.954209971381233e-05,
+      "loss": 0.5554,
+      "step": 9375
+    },
+    {
+      "epoch": 2.813264310901043,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.952703720439825e-05,
+      "loss": 0.5458,
+      "step": 9376
+    },
+    {
+      "epoch": 2.8135644084327405,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.951197469498419e-05,
+      "loss": 0.4268,
+      "step": 9377
+    },
+    {
+      "epoch": 2.8138645059644385,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.949691218557012e-05,
+      "loss": 0.4815,
+      "step": 9378
+    },
+    {
+      "epoch": 2.8141646034961365,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.9481849676156055e-05,
+      "loss": 0.5002,
+      "step": 9379
+    },
+    {
+      "epoch": 2.814464701027834,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.9466787166741975e-05,
+      "loss": 0.4936,
+      "step": 9380
+    },
+    {
+      "epoch": 2.8147647985595317,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.945172465732791e-05,
+      "loss": 0.519,
+      "step": 9381
+    },
+    {
+      "epoch": 2.8150648960912297,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.943666214791385e-05,
+      "loss": 0.5747,
+      "step": 9382
+    },
+    {
+      "epoch": 2.8153649936229277,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.942159963849978e-05,
+      "loss": 0.4861,
+      "step": 9383
+    },
+    {
+      "epoch": 2.8156650911546253,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.94065371290857e-05,
+      "loss": 0.5141,
+      "step": 9384
+    },
+    {
+      "epoch": 2.815965188686323,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.9391474619671636e-05,
+      "loss": 0.4844,
+      "step": 9385
+    },
+    {
+      "epoch": 2.816265286218021,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.9376412110257576e-05,
+      "loss": 0.5998,
+      "step": 9386
+    },
+    {
+      "epoch": 2.816565383749719,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.936134960084351e-05,
+      "loss": 0.6053,
+      "step": 9387
+    },
+    {
+      "epoch": 2.8168654812814165,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.934628709142943e-05,
+      "loss": 0.5313,
+      "step": 9388
+    },
+    {
+      "epoch": 2.817165578813114,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.9331224582015363e-05,
+      "loss": 0.4849,
+      "step": 9389
+    },
+    {
+      "epoch": 2.817465676344812,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.93161620726013e-05,
+      "loss": 0.4157,
+      "step": 9390
+    },
+    {
+      "epoch": 2.81776577387651,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.930109956318724e-05,
+      "loss": 0.5798,
+      "step": 9391
+    },
+    {
+      "epoch": 2.8180658714082076,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.928603705377316e-05,
+      "loss": 0.5843,
+      "step": 9392
+    },
+    {
+      "epoch": 2.818365968939905,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.927097454435909e-05,
+      "loss": 0.5105,
+      "step": 9393
+    },
+    {
+      "epoch": 2.8186660664716032,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.9255912034945024e-05,
+      "loss": 0.5525,
+      "step": 9394
+    },
+    {
+      "epoch": 2.8189661640033012,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.9240849525530965e-05,
+      "loss": 0.5086,
+      "step": 9395
+    },
+    {
+      "epoch": 2.819266261534999,
+      "grad_norm": 1.28125,
+      "learning_rate": 5.9225787016116885e-05,
+      "loss": 0.5586,
+      "step": 9396
+    },
+    {
+      "epoch": 2.819566359066697,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.921072450670282e-05,
+      "loss": 0.4744,
+      "step": 9397
+    },
+    {
+      "epoch": 2.8198664565983944,
+      "grad_norm": 1.125,
+      "learning_rate": 5.919566199728875e-05,
+      "loss": 0.6222,
+      "step": 9398
+    },
+    {
+      "epoch": 2.8201665541300924,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.918059948787468e-05,
+      "loss": 0.5216,
+      "step": 9399
+    },
+    {
+      "epoch": 2.82046665166179,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.916553697846061e-05,
+      "loss": 0.5629,
+      "step": 9400
+    },
+    {
+      "epoch": 2.820766749193488,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.9150474469046546e-05,
+      "loss": 0.3941,
+      "step": 9401
+    },
+    {
+      "epoch": 2.8210668467251856,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.913541195963248e-05,
+      "loss": 0.6369,
+      "step": 9402
+    },
+    {
+      "epoch": 2.8213669442568836,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.9120349450218406e-05,
+      "loss": 0.5413,
+      "step": 9403
+    },
+    {
+      "epoch": 2.821667041788581,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.910528694080434e-05,
+      "loss": 0.4508,
+      "step": 9404
+    },
+    {
+      "epoch": 2.821967139320279,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.909022443139027e-05,
+      "loss": 0.4949,
+      "step": 9405
+    },
+    {
+      "epoch": 2.822267236851977,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.9075161921976206e-05,
+      "loss": 0.5573,
+      "step": 9406
+    },
+    {
+      "epoch": 2.822567334383675,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.906009941256213e-05,
+      "loss": 0.5936,
+      "step": 9407
+    },
+    {
+      "epoch": 2.8228674319153724,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.904503690314807e-05,
+      "loss": 0.5125,
+      "step": 9408
+    },
+    {
+      "epoch": 2.8231675294470704,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.9029974393734e-05,
+      "loss": 0.4685,
+      "step": 9409
+    },
+    {
+      "epoch": 2.823467626978768,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.9014911884319934e-05,
+      "loss": 0.5677,
+      "step": 9410
+    },
+    {
+      "epoch": 2.823767724510466,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.899984937490586e-05,
+      "loss": 0.5449,
+      "step": 9411
+    },
+    {
+      "epoch": 2.8240678220421636,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.8984786865491794e-05,
+      "loss": 0.5049,
+      "step": 9412
+    },
+    {
+      "epoch": 2.8243679195738616,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.896972435607773e-05,
+      "loss": 0.4753,
+      "step": 9413
+    },
+    {
+      "epoch": 2.824668017105559,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.895466184666366e-05,
+      "loss": 0.53,
+      "step": 9414
+    },
+    {
+      "epoch": 2.824968114637257,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.893959933724959e-05,
+      "loss": 0.5211,
+      "step": 9415
+    },
+    {
+      "epoch": 2.8252682121689547,
+      "grad_norm": 1.2578125,
+      "learning_rate": 5.892453682783552e-05,
+      "loss": 0.6273,
+      "step": 9416
+    },
+    {
+      "epoch": 2.8255683097006528,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.8909474318421455e-05,
+      "loss": 0.5052,
+      "step": 9417
+    },
+    {
+      "epoch": 2.8258684072323508,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.889441180900739e-05,
+      "loss": 0.4382,
+      "step": 9418
+    },
+    {
+      "epoch": 2.8261685047640484,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.887934929959331e-05,
+      "loss": 0.4885,
+      "step": 9419
+    },
+    {
+      "epoch": 2.826468602295746,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.886428679017925e-05,
+      "loss": 0.596,
+      "step": 9420
+    },
+    {
+      "epoch": 2.826768699827444,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.884922428076518e-05,
+      "loss": 0.5264,
+      "step": 9421
+    },
+    {
+      "epoch": 2.827068797359142,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.8834161771351116e-05,
+      "loss": 0.5908,
+      "step": 9422
+    },
+    {
+      "epoch": 2.8273688948908395,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.8819099261937036e-05,
+      "loss": 0.5997,
+      "step": 9423
+    },
+    {
+      "epoch": 2.827668992422537,
+      "grad_norm": 1.0,
+      "learning_rate": 5.8804036752522976e-05,
+      "loss": 0.5509,
+      "step": 9424
+    },
+    {
+      "epoch": 2.827969089954235,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.878897424310891e-05,
+      "loss": 0.5387,
+      "step": 9425
+    },
+    {
+      "epoch": 2.828269187485933,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.877391173369483e-05,
+      "loss": 0.4988,
+      "step": 9426
+    },
+    {
+      "epoch": 2.8285692850176307,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.875884922428076e-05,
+      "loss": 0.5605,
+      "step": 9427
+    },
+    {
+      "epoch": 2.8288693825493283,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.8743786714866697e-05,
+      "loss": 0.5266,
+      "step": 9428
+    },
+    {
+      "epoch": 2.8291694800810263,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.872872420545264e-05,
+      "loss": 0.4961,
+      "step": 9429
+    },
+    {
+      "epoch": 2.8294695776127243,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.871366169603856e-05,
+      "loss": 0.5472,
+      "step": 9430
+    },
+    {
+      "epoch": 2.829769675144422,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.869859918662449e-05,
+      "loss": 0.5184,
+      "step": 9431
+    },
+    {
+      "epoch": 2.8300697726761195,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.8683536677210424e-05,
+      "loss": 0.5592,
+      "step": 9432
+    },
+    {
+      "epoch": 2.8303698702078175,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.8668474167796364e-05,
+      "loss": 0.564,
+      "step": 9433
+    },
+    {
+      "epoch": 2.8306699677395155,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.8653411658382284e-05,
+      "loss": 0.5771,
+      "step": 9434
+    },
+    {
+      "epoch": 2.830970065271213,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.863834914896822e-05,
+      "loss": 0.4428,
+      "step": 9435
+    },
+    {
+      "epoch": 2.831270162802911,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.862328663955415e-05,
+      "loss": 0.4807,
+      "step": 9436
+    },
+    {
+      "epoch": 2.8315702603346087,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.8608224130140085e-05,
+      "loss": 0.5016,
+      "step": 9437
+    },
+    {
+      "epoch": 2.8318703578663067,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.859316162072601e-05,
+      "loss": 0.5368,
+      "step": 9438
+    },
+    {
+      "epoch": 2.8321704553980043,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.8578099111311945e-05,
+      "loss": 0.4702,
+      "step": 9439
+    },
+    {
+      "epoch": 2.8324705529297023,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.856303660189788e-05,
+      "loss": 0.6637,
+      "step": 9440
+    },
+    {
+      "epoch": 2.8327706504614,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.854797409248381e-05,
+      "loss": 0.5504,
+      "step": 9441
+    },
+    {
+      "epoch": 2.833070747993098,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.853291158306974e-05,
+      "loss": 0.4443,
+      "step": 9442
+    },
+    {
+      "epoch": 2.8333708455247955,
+      "grad_norm": 1.28125,
+      "learning_rate": 5.851784907365567e-05,
+      "loss": 0.5896,
+      "step": 9443
+    },
+    {
+      "epoch": 2.8336709430564935,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.8502786564241606e-05,
+      "loss": 0.5472,
+      "step": 9444
+    },
+    {
+      "epoch": 2.833971040588191,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.848772405482754e-05,
+      "loss": 0.5241,
+      "step": 9445
+    },
+    {
+      "epoch": 2.834271138119889,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.8472661545413466e-05,
+      "loss": 0.4877,
+      "step": 9446
+    },
+    {
+      "epoch": 2.8345712356515866,
+      "grad_norm": 1.25,
+      "learning_rate": 5.84575990359994e-05,
+      "loss": 0.4939,
+      "step": 9447
+    },
+    {
+      "epoch": 2.8348713331832847,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.844253652658533e-05,
+      "loss": 0.49,
+      "step": 9448
+    },
+    {
+      "epoch": 2.8351714307149822,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.842747401717127e-05,
+      "loss": 0.5412,
+      "step": 9449
+    },
+    {
+      "epoch": 2.8354715282466803,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.8412411507757194e-05,
+      "loss": 0.4355,
+      "step": 9450
+    },
+    {
+      "epoch": 2.835771625778378,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.839734899834313e-05,
+      "loss": 0.4963,
+      "step": 9451
+    },
+    {
+      "epoch": 2.836071723310076,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.838228648892906e-05,
+      "loss": 0.4649,
+      "step": 9452
+    },
+    {
+      "epoch": 2.8363718208417734,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.836722397951499e-05,
+      "loss": 0.5476,
+      "step": 9453
+    },
+    {
+      "epoch": 2.8366719183734714,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.835216147010092e-05,
+      "loss": 0.5946,
+      "step": 9454
+    },
+    {
+      "epoch": 2.836972015905169,
+      "grad_norm": 1.2109375,
+      "learning_rate": 5.8337098960686854e-05,
+      "loss": 0.5776,
+      "step": 9455
+    },
+    {
+      "epoch": 2.837272113436867,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.832203645127279e-05,
+      "loss": 0.538,
+      "step": 9456
+    },
+    {
+      "epoch": 2.837572210968565,
+      "grad_norm": 1.296875,
+      "learning_rate": 5.830697394185871e-05,
+      "loss": 0.5688,
+      "step": 9457
+    },
+    {
+      "epoch": 2.8378723085002626,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.829191143244465e-05,
+      "loss": 0.4446,
+      "step": 9458
+    },
+    {
+      "epoch": 2.83817240603196,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.827684892303058e-05,
+      "loss": 0.5549,
+      "step": 9459
+    },
+    {
+      "epoch": 2.838472503563658,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.8261786413616515e-05,
+      "loss": 0.5205,
+      "step": 9460
+    },
+    {
+      "epoch": 2.8387726010953562,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.8246723904202435e-05,
+      "loss": 0.5553,
+      "step": 9461
+    },
+    {
+      "epoch": 2.839072698627054,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.8231661394788376e-05,
+      "loss": 0.5442,
+      "step": 9462
+    },
+    {
+      "epoch": 2.8393727961587514,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.821659888537431e-05,
+      "loss": 0.5311,
+      "step": 9463
+    },
+    {
+      "epoch": 2.8396728936904494,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.820153637596024e-05,
+      "loss": 0.4406,
+      "step": 9464
+    },
+    {
+      "epoch": 2.8399729912221474,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.818647386654616e-05,
+      "loss": 0.4244,
+      "step": 9465
+    },
+    {
+      "epoch": 2.840273088753845,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.81714113571321e-05,
+      "loss": 0.454,
+      "step": 9466
+    },
+    {
+      "epoch": 2.8405731862855426,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.8156348847718036e-05,
+      "loss": 0.5198,
+      "step": 9467
+    },
+    {
+      "epoch": 2.8408732838172406,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.814128633830397e-05,
+      "loss": 0.4583,
+      "step": 9468
+    },
+    {
+      "epoch": 2.8411733813489386,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.812622382888989e-05,
+      "loss": 0.5147,
+      "step": 9469
+    },
+    {
+      "epoch": 2.841473478880636,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.8111161319475823e-05,
+      "loss": 0.5,
+      "step": 9470
+    },
+    {
+      "epoch": 2.8417735764123337,
+      "grad_norm": 1.0,
+      "learning_rate": 5.8096098810061764e-05,
+      "loss": 0.5039,
+      "step": 9471
+    },
+    {
+      "epoch": 2.8420736739440318,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.80810363006477e-05,
+      "loss": 0.5341,
+      "step": 9472
+    },
+    {
+      "epoch": 2.84237377147573,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.806597379123362e-05,
+      "loss": 0.5035,
+      "step": 9473
+    },
+    {
+      "epoch": 2.8426738690074274,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.805091128181955e-05,
+      "loss": 0.5255,
+      "step": 9474
+    },
+    {
+      "epoch": 2.8429739665391254,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.803584877240549e-05,
+      "loss": 0.5213,
+      "step": 9475
+    },
+    {
+      "epoch": 2.843274064070823,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.8020786262991425e-05,
+      "loss": 0.555,
+      "step": 9476
+    },
+    {
+      "epoch": 2.843574161602521,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.8005723753577345e-05,
+      "loss": 0.4952,
+      "step": 9477
+    },
+    {
+      "epoch": 2.8438742591342185,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.799066124416328e-05,
+      "loss": 0.5206,
+      "step": 9478
+    },
+    {
+      "epoch": 2.8441743566659166,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.797559873474921e-05,
+      "loss": 0.5026,
+      "step": 9479
+    },
+    {
+      "epoch": 2.844474454197614,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.796053622533515e-05,
+      "loss": 0.5635,
+      "step": 9480
+    },
+    {
+      "epoch": 2.844774551729312,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.794547371592107e-05,
+      "loss": 0.5259,
+      "step": 9481
+    },
+    {
+      "epoch": 2.8450746492610097,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.7930411206507006e-05,
+      "loss": 0.5218,
+      "step": 9482
+    },
+    {
+      "epoch": 2.8453747467927077,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.791534869709294e-05,
+      "loss": 0.5199,
+      "step": 9483
+    },
+    {
+      "epoch": 2.8456748443244053,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.7900286187678866e-05,
+      "loss": 0.5326,
+      "step": 9484
+    },
+    {
+      "epoch": 2.8459749418561033,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.78852236782648e-05,
+      "loss": 0.5719,
+      "step": 9485
+    },
+    {
+      "epoch": 2.846275039387801,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.787016116885073e-05,
+      "loss": 0.5298,
+      "step": 9486
+    },
+    {
+      "epoch": 2.846575136919499,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.7855098659436666e-05,
+      "loss": 0.5404,
+      "step": 9487
+    },
+    {
+      "epoch": 2.8468752344511965,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.784003615002259e-05,
+      "loss": 0.5627,
+      "step": 9488
+    },
+    {
+      "epoch": 2.8471753319828945,
+      "grad_norm": 1.0,
+      "learning_rate": 5.782497364060853e-05,
+      "loss": 0.5407,
+      "step": 9489
+    },
+    {
+      "epoch": 2.847475429514592,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.780991113119446e-05,
+      "loss": 0.4657,
+      "step": 9490
+    },
+    {
+      "epoch": 2.84777552704629,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.7794848621780394e-05,
+      "loss": 0.5367,
+      "step": 9491
+    },
+    {
+      "epoch": 2.8480756245779877,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.777978611236632e-05,
+      "loss": 0.597,
+      "step": 9492
+    },
+    {
+      "epoch": 2.8483757221096857,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.7764723602952254e-05,
+      "loss": 0.492,
+      "step": 9493
+    },
+    {
+      "epoch": 2.8486758196413833,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.774966109353819e-05,
+      "loss": 0.5064,
+      "step": 9494
+    },
+    {
+      "epoch": 2.8489759171730813,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.773459858412412e-05,
+      "loss": 0.5458,
+      "step": 9495
+    },
+    {
+      "epoch": 2.8492760147047793,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.771953607471005e-05,
+      "loss": 0.527,
+      "step": 9496
+    },
+    {
+      "epoch": 2.849576112236477,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.770447356529598e-05,
+      "loss": 0.4579,
+      "step": 9497
+    },
+    {
+      "epoch": 2.8498762097681745,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.7689411055881915e-05,
+      "loss": 0.559,
+      "step": 9498
+    },
+    {
+      "epoch": 2.8501763072998725,
+      "grad_norm": 0.84765625,
+      "learning_rate": 5.767434854646785e-05,
+      "loss": 0.4437,
+      "step": 9499
+    },
+    {
+      "epoch": 2.8504764048315705,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.7659286037053775e-05,
+      "loss": 0.5316,
+      "step": 9500
+    },
+    {
+      "epoch": 2.850776502363268,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.764422352763971e-05,
+      "loss": 0.4249,
+      "step": 9501
+    },
+    {
+      "epoch": 2.8510765998949656,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.762916101822564e-05,
+      "loss": 0.4389,
+      "step": 9502
+    },
+    {
+      "epoch": 2.8513766974266637,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.7614098508811576e-05,
+      "loss": 0.4875,
+      "step": 9503
+    },
+    {
+      "epoch": 2.8516767949583617,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.75990359993975e-05,
+      "loss": 0.4956,
+      "step": 9504
+    },
+    {
+      "epoch": 2.8519768924900593,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.7583973489983436e-05,
+      "loss": 0.5671,
+      "step": 9505
+    },
+    {
+      "epoch": 2.852276990021757,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.756891098056937e-05,
+      "loss": 0.4625,
+      "step": 9506
+    },
+    {
+      "epoch": 2.852577087553455,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.75538484711553e-05,
+      "loss": 0.4553,
+      "step": 9507
+    },
+    {
+      "epoch": 2.852877185085153,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.753878596174122e-05,
+      "loss": 0.4826,
+      "step": 9508
+    },
+    {
+      "epoch": 2.8531772826168504,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.752372345232716e-05,
+      "loss": 0.4057,
+      "step": 9509
+    },
+    {
+      "epoch": 2.853477380148548,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.75086609429131e-05,
+      "loss": 0.5385,
+      "step": 9510
+    },
+    {
+      "epoch": 2.853777477680246,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.749359843349902e-05,
+      "loss": 0.5685,
+      "step": 9511
+    },
+    {
+      "epoch": 2.854077575211944,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.747853592408495e-05,
+      "loss": 0.4512,
+      "step": 9512
+    },
+    {
+      "epoch": 2.8543776727436416,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.746347341467089e-05,
+      "loss": 0.4697,
+      "step": 9513
+    },
+    {
+      "epoch": 2.8546777702753396,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.7448410905256824e-05,
+      "loss": 0.4592,
+      "step": 9514
+    },
+    {
+      "epoch": 2.854977867807037,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.7433348395842744e-05,
+      "loss": 0.4801,
+      "step": 9515
+    },
+    {
+      "epoch": 2.8552779653387352,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.741828588642868e-05,
+      "loss": 0.5486,
+      "step": 9516
+    },
+    {
+      "epoch": 2.855578062870433,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.740322337701461e-05,
+      "loss": 0.5344,
+      "step": 9517
+    },
+    {
+      "epoch": 2.855878160402131,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.738816086760055e-05,
+      "loss": 0.4911,
+      "step": 9518
+    },
+    {
+      "epoch": 2.8561782579338284,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.737309835818647e-05,
+      "loss": 0.5661,
+      "step": 9519
+    },
+    {
+      "epoch": 2.8564783554655264,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.7358035848772405e-05,
+      "loss": 0.623,
+      "step": 9520
+    },
+    {
+      "epoch": 2.856778452997224,
+      "grad_norm": 0.875,
+      "learning_rate": 5.734297333935834e-05,
+      "loss": 0.49,
+      "step": 9521
+    },
+    {
+      "epoch": 2.857078550528922,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.732791082994428e-05,
+      "loss": 0.5699,
+      "step": 9522
+    },
+    {
+      "epoch": 2.8573786480606196,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.73128483205302e-05,
+      "loss": 0.5146,
+      "step": 9523
+    },
+    {
+      "epoch": 2.8576787455923176,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.729778581111613e-05,
+      "loss": 0.5257,
+      "step": 9524
+    },
+    {
+      "epoch": 2.857978843124015,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.7282723301702066e-05,
+      "loss": 0.5167,
+      "step": 9525
+    },
+    {
+      "epoch": 2.858278940655713,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.7267660792288e-05,
+      "loss": 0.5327,
+      "step": 9526
+    },
+    {
+      "epoch": 2.8585790381874108,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.7252598282873926e-05,
+      "loss": 0.5775,
+      "step": 9527
+    },
+    {
+      "epoch": 2.858879135719109,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.723753577345986e-05,
+      "loss": 0.5229,
+      "step": 9528
+    },
+    {
+      "epoch": 2.8591792332508064,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.722247326404579e-05,
+      "loss": 0.562,
+      "step": 9529
+    },
+    {
+      "epoch": 2.8594793307825044,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.720741075463173e-05,
+      "loss": 0.5531,
+      "step": 9530
+    },
+    {
+      "epoch": 2.859779428314202,
+      "grad_norm": 1.46875,
+      "learning_rate": 5.7192348245217654e-05,
+      "loss": 0.6058,
+      "step": 9531
+    },
+    {
+      "epoch": 2.8600795258459,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.717728573580359e-05,
+      "loss": 0.4685,
+      "step": 9532
+    },
+    {
+      "epoch": 2.8603796233775975,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.716222322638952e-05,
+      "loss": 0.485,
+      "step": 9533
+    },
+    {
+      "epoch": 2.8606797209092956,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.7147160716975454e-05,
+      "loss": 0.5391,
+      "step": 9534
+    },
+    {
+      "epoch": 2.8609798184409936,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.713209820756138e-05,
+      "loss": 0.4388,
+      "step": 9535
+    },
+    {
+      "epoch": 2.861279915972691,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.7117035698147314e-05,
+      "loss": 0.5504,
+      "step": 9536
+    },
+    {
+      "epoch": 2.8615800135043887,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.710197318873325e-05,
+      "loss": 0.5034,
+      "step": 9537
+    },
+    {
+      "epoch": 2.8618801110360867,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.7086910679319175e-05,
+      "loss": 0.4859,
+      "step": 9538
+    },
+    {
+      "epoch": 2.8621802085677848,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.707184816990511e-05,
+      "loss": 0.5108,
+      "step": 9539
+    },
+    {
+      "epoch": 2.8624803060994823,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.705678566049104e-05,
+      "loss": 0.5464,
+      "step": 9540
+    },
+    {
+      "epoch": 2.86278040363118,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.7041723151076975e-05,
+      "loss": 0.5482,
+      "step": 9541
+    },
+    {
+      "epoch": 2.863080501162878,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.70266606416629e-05,
+      "loss": 0.575,
+      "step": 9542
+    },
+    {
+      "epoch": 2.863380598694576,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.7011598132248836e-05,
+      "loss": 0.4192,
+      "step": 9543
+    },
+    {
+      "epoch": 2.8636806962262735,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.699653562283477e-05,
+      "loss": 0.5119,
+      "step": 9544
+    },
+    {
+      "epoch": 2.863980793757971,
+      "grad_norm": 1.125,
+      "learning_rate": 5.69814731134207e-05,
+      "loss": 0.4532,
+      "step": 9545
+    },
+    {
+      "epoch": 2.864280891289669,
+      "grad_norm": 1.0,
+      "learning_rate": 5.696641060400662e-05,
+      "loss": 0.5379,
+      "step": 9546
+    },
+    {
+      "epoch": 2.864580988821367,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.695134809459256e-05,
+      "loss": 0.4904,
+      "step": 9547
+    },
+    {
+      "epoch": 2.8648810863530647,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.6936285585178496e-05,
+      "loss": 0.6077,
+      "step": 9548
+    },
+    {
+      "epoch": 2.8651811838847623,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.692122307576443e-05,
+      "loss": 0.5994,
+      "step": 9549
+    },
+    {
+      "epoch": 2.8654812814164603,
+      "grad_norm": 1.296875,
+      "learning_rate": 5.690616056635035e-05,
+      "loss": 0.5168,
+      "step": 9550
+    },
+    {
+      "epoch": 2.8657813789481583,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.689109805693629e-05,
+      "loss": 0.5978,
+      "step": 9551
+    },
+    {
+      "epoch": 2.866081476479856,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.6876035547522224e-05,
+      "loss": 0.4675,
+      "step": 9552
+    },
+    {
+      "epoch": 2.866381574011554,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.686097303810816e-05,
+      "loss": 0.4312,
+      "step": 9553
+    },
+    {
+      "epoch": 2.8666816715432515,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.684591052869408e-05,
+      "loss": 0.5822,
+      "step": 9554
+    },
+    {
+      "epoch": 2.8669817690749495,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.683084801928001e-05,
+      "loss": 0.5092,
+      "step": 9555
+    },
+    {
+      "epoch": 2.867281866606647,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.681578550986595e-05,
+      "loss": 0.5287,
+      "step": 9556
+    },
+    {
+      "epoch": 2.867581964138345,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.6800723000451885e-05,
+      "loss": 0.5799,
+      "step": 9557
+    },
+    {
+      "epoch": 2.8678820616700427,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.6785660491037805e-05,
+      "loss": 0.5155,
+      "step": 9558
+    },
+    {
+      "epoch": 2.8681821592017407,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.677059798162374e-05,
+      "loss": 0.5118,
+      "step": 9559
+    },
+    {
+      "epoch": 2.8684822567334383,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.675553547220968e-05,
+      "loss": 0.5263,
+      "step": 9560
+    },
+    {
+      "epoch": 2.8687823542651363,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.674047296279561e-05,
+      "loss": 0.5864,
+      "step": 9561
+    },
+    {
+      "epoch": 2.869082451796834,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.672541045338153e-05,
+      "loss": 0.546,
+      "step": 9562
+    },
+    {
+      "epoch": 2.869382549328532,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.6710347943967466e-05,
+      "loss": 0.5038,
+      "step": 9563
+    },
+    {
+      "epoch": 2.8696826468602294,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.66952854345534e-05,
+      "loss": 0.4517,
+      "step": 9564
+    },
+    {
+      "epoch": 2.8699827443919275,
+      "grad_norm": 1.125,
+      "learning_rate": 5.6680222925139326e-05,
+      "loss": 0.5882,
+      "step": 9565
+    },
+    {
+      "epoch": 2.870282841923625,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.666516041572526e-05,
+      "loss": 0.398,
+      "step": 9566
+    },
+    {
+      "epoch": 2.870582939455323,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.665009790631119e-05,
+      "loss": 0.6269,
+      "step": 9567
+    },
+    {
+      "epoch": 2.8708830369870206,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.6635035396897126e-05,
+      "loss": 0.4482,
+      "step": 9568
+    },
+    {
+      "epoch": 2.8711831345187186,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.661997288748305e-05,
+      "loss": 0.5188,
+      "step": 9569
+    },
+    {
+      "epoch": 2.8714832320504162,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.660491037806899e-05,
+      "loss": 0.5785,
+      "step": 9570
+    },
+    {
+      "epoch": 2.8717833295821142,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.658984786865492e-05,
+      "loss": 0.5098,
+      "step": 9571
+    },
+    {
+      "epoch": 2.872083427113812,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.6574785359240854e-05,
+      "loss": 0.4829,
+      "step": 9572
+    },
+    {
+      "epoch": 2.87238352464551,
+      "grad_norm": 1.25,
+      "learning_rate": 5.655972284982678e-05,
+      "loss": 0.4638,
+      "step": 9573
+    },
+    {
+      "epoch": 2.872683622177208,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.6544660340412714e-05,
+      "loss": 0.5118,
+      "step": 9574
+    },
+    {
+      "epoch": 2.8729837197089054,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.652959783099865e-05,
+      "loss": 0.5499,
+      "step": 9575
+    },
+    {
+      "epoch": 2.873283817240603,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.651453532158458e-05,
+      "loss": 0.5233,
+      "step": 9576
+    },
+    {
+      "epoch": 2.873583914772301,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.649947281217051e-05,
+      "loss": 0.5243,
+      "step": 9577
+    },
+    {
+      "epoch": 2.873884012303999,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.648441030275644e-05,
+      "loss": 0.524,
+      "step": 9578
+    },
+    {
+      "epoch": 2.8741841098356966,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.6469347793342375e-05,
+      "loss": 0.5868,
+      "step": 9579
+    },
+    {
+      "epoch": 2.874484207367394,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.645428528392831e-05,
+      "loss": 0.5457,
+      "step": 9580
+    },
+    {
+      "epoch": 2.874784304899092,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.6439222774514235e-05,
+      "loss": 0.5384,
+      "step": 9581
+    },
+    {
+      "epoch": 2.87508440243079,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.642416026510017e-05,
+      "loss": 0.4495,
+      "step": 9582
+    },
+    {
+      "epoch": 2.875384499962488,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.64090977556861e-05,
+      "loss": 0.5634,
+      "step": 9583
+    },
+    {
+      "epoch": 2.8756845974941854,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.6394035246272036e-05,
+      "loss": 0.5081,
+      "step": 9584
+    },
+    {
+      "epoch": 2.8759846950258834,
+      "grad_norm": 1.125,
+      "learning_rate": 5.637897273685796e-05,
+      "loss": 0.5495,
+      "step": 9585
+    },
+    {
+      "epoch": 2.8762847925575814,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.6363910227443896e-05,
+      "loss": 0.4721,
+      "step": 9586
+    },
+    {
+      "epoch": 2.876584890089279,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.634884771802983e-05,
+      "loss": 0.5897,
+      "step": 9587
+    },
+    {
+      "epoch": 2.8768849876209766,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.633378520861576e-05,
+      "loss": 0.5041,
+      "step": 9588
+    },
+    {
+      "epoch": 2.8771850851526746,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.631872269920169e-05,
+      "loss": 0.5176,
+      "step": 9589
+    },
+    {
+      "epoch": 2.8774851826843726,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.6303660189787623e-05,
+      "loss": 0.5552,
+      "step": 9590
+    },
+    {
+      "epoch": 2.87778528021607,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.628859768037356e-05,
+      "loss": 0.4783,
+      "step": 9591
+    },
+    {
+      "epoch": 2.878085377747768,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.627353517095948e-05,
+      "loss": 0.4854,
+      "step": 9592
+    },
+    {
+      "epoch": 2.8783854752794658,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.625847266154542e-05,
+      "loss": 0.5715,
+      "step": 9593
+    },
+    {
+      "epoch": 2.8786855728111638,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.624341015213135e-05,
+      "loss": 0.5643,
+      "step": 9594
+    },
+    {
+      "epoch": 2.8789856703428613,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.6228347642717284e-05,
+      "loss": 0.5208,
+      "step": 9595
+    },
+    {
+      "epoch": 2.8792857678745594,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.6213285133303204e-05,
+      "loss": 0.492,
+      "step": 9596
+    },
+    {
+      "epoch": 2.879585865406257,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.619822262388914e-05,
+      "loss": 0.5544,
+      "step": 9597
+    },
+    {
+      "epoch": 2.879885962937955,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.618316011447508e-05,
+      "loss": 0.564,
+      "step": 9598
+    },
+    {
+      "epoch": 2.8801860604696525,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.616809760506101e-05,
+      "loss": 0.5725,
+      "step": 9599
+    },
+    {
+      "epoch": 2.8804861580013505,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.615303509564693e-05,
+      "loss": 0.5089,
+      "step": 9600
+    },
+    {
+      "epoch": 2.880786255533048,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.6137972586232865e-05,
+      "loss": 0.5601,
+      "step": 9601
+    },
+    {
+      "epoch": 2.881086353064746,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.6122910076818805e-05,
+      "loss": 0.6164,
+      "step": 9602
+    },
+    {
+      "epoch": 2.8813864505964437,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.610784756740474e-05,
+      "loss": 0.5152,
+      "step": 9603
+    },
+    {
+      "epoch": 2.8816865481281417,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.609278505799066e-05,
+      "loss": 0.5936,
+      "step": 9604
+    },
+    {
+      "epoch": 2.8819866456598393,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.607772254857659e-05,
+      "loss": 0.4909,
+      "step": 9605
+    },
+    {
+      "epoch": 2.8822867431915373,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.6062660039162526e-05,
+      "loss": 0.5001,
+      "step": 9606
+    },
+    {
+      "epoch": 2.882586840723235,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.6047597529748466e-05,
+      "loss": 0.5577,
+      "step": 9607
+    },
+    {
+      "epoch": 2.882886938254933,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.6032535020334386e-05,
+      "loss": 0.5073,
+      "step": 9608
+    },
+    {
+      "epoch": 2.8831870357866305,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.601747251092032e-05,
+      "loss": 0.5529,
+      "step": 9609
+    },
+    {
+      "epoch": 2.8834871333183285,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.600241000150625e-05,
+      "loss": 0.5621,
+      "step": 9610
+    },
+    {
+      "epoch": 2.883787230850026,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.5987347492092194e-05,
+      "loss": 0.5284,
+      "step": 9611
+    },
+    {
+      "epoch": 2.884087328381724,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.5972284982678114e-05,
+      "loss": 0.4646,
+      "step": 9612
+    },
+    {
+      "epoch": 2.884387425913422,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.595722247326405e-05,
+      "loss": 0.5372,
+      "step": 9613
+    },
+    {
+      "epoch": 2.8846875234451197,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.594215996384998e-05,
+      "loss": 0.4937,
+      "step": 9614
+    },
+    {
+      "epoch": 2.8849876209768173,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.5927097454435914e-05,
+      "loss": 0.4619,
+      "step": 9615
+    },
+    {
+      "epoch": 2.8852877185085153,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.591203494502184e-05,
+      "loss": 0.5173,
+      "step": 9616
+    },
+    {
+      "epoch": 2.8855878160402133,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.5896972435607774e-05,
+      "loss": 0.426,
+      "step": 9617
+    },
+    {
+      "epoch": 2.885887913571911,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.588190992619371e-05,
+      "loss": 0.5639,
+      "step": 9618
+    },
+    {
+      "epoch": 2.8861880111036085,
+      "grad_norm": 1.0,
+      "learning_rate": 5.5866847416779635e-05,
+      "loss": 0.4545,
+      "step": 9619
+    },
+    {
+      "epoch": 2.8864881086353065,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.585178490736557e-05,
+      "loss": 0.4942,
+      "step": 9620
+    },
+    {
+      "epoch": 2.8867882061670045,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.58367223979515e-05,
+      "loss": 0.6024,
+      "step": 9621
+    },
+    {
+      "epoch": 2.887088303698702,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.5821659888537435e-05,
+      "loss": 0.5607,
+      "step": 9622
+    },
+    {
+      "epoch": 2.8873884012303996,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.580659737912336e-05,
+      "loss": 0.4971,
+      "step": 9623
+    },
+    {
+      "epoch": 2.8876884987620977,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.5791534869709296e-05,
+      "loss": 0.4761,
+      "step": 9624
+    },
+    {
+      "epoch": 2.8879885962937957,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.577647236029523e-05,
+      "loss": 0.4956,
+      "step": 9625
+    },
+    {
+      "epoch": 2.8882886938254932,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.576140985088116e-05,
+      "loss": 0.4961,
+      "step": 9626
+    },
+    {
+      "epoch": 2.888588791357191,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.574634734146709e-05,
+      "loss": 0.5081,
+      "step": 9627
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.573128483205302e-05,
+      "loss": 0.5037,
+      "step": 9628
+    },
+    {
+      "epoch": 2.889188986420587,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.5716222322638956e-05,
+      "loss": 0.4972,
+      "step": 9629
+    },
+    {
+      "epoch": 2.8894890839522844,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.570115981322489e-05,
+      "loss": 0.5486,
+      "step": 9630
+    },
+    {
+      "epoch": 2.8897891814839824,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.568609730381082e-05,
+      "loss": 0.5065,
+      "step": 9631
+    },
+    {
+      "epoch": 2.89008927901568,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.567103479439675e-05,
+      "loss": 0.4386,
+      "step": 9632
+    },
+    {
+      "epoch": 2.890389376547378,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.5655972284982684e-05,
+      "loss": 0.5665,
+      "step": 9633
+    },
+    {
+      "epoch": 2.8906894740790756,
+      "grad_norm": 1.125,
+      "learning_rate": 5.564090977556862e-05,
+      "loss": 0.5094,
+      "step": 9634
+    },
+    {
+      "epoch": 2.8909895716107736,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.562584726615454e-05,
+      "loss": 0.5174,
+      "step": 9635
+    },
+    {
+      "epoch": 2.891289669142471,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.561078475674048e-05,
+      "loss": 0.5781,
+      "step": 9636
+    },
+    {
+      "epoch": 2.8915897666741692,
+      "grad_norm": 1.125,
+      "learning_rate": 5.559572224732641e-05,
+      "loss": 0.5731,
+      "step": 9637
+    },
+    {
+      "epoch": 2.891889864205867,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.5580659737912345e-05,
+      "loss": 0.5929,
+      "step": 9638
+    },
+    {
+      "epoch": 2.892189961737565,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.5565597228498265e-05,
+      "loss": 0.6242,
+      "step": 9639
+    },
+    {
+      "epoch": 2.8924900592692624,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.5550534719084205e-05,
+      "loss": 0.5743,
+      "step": 9640
+    },
+    {
+      "epoch": 2.8927901568009604,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.553547220967014e-05,
+      "loss": 0.5035,
+      "step": 9641
+    },
+    {
+      "epoch": 2.893090254332658,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.552040970025607e-05,
+      "loss": 0.5857,
+      "step": 9642
+    },
+    {
+      "epoch": 2.893390351864356,
+      "grad_norm": 1.0,
+      "learning_rate": 5.550534719084199e-05,
+      "loss": 0.5343,
+      "step": 9643
+    },
+    {
+      "epoch": 2.8936904493960536,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.5490284681427926e-05,
+      "loss": 0.4575,
+      "step": 9644
+    },
+    {
+      "epoch": 2.8939905469277516,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.5475222172013866e-05,
+      "loss": 0.5367,
+      "step": 9645
+    },
+    {
+      "epoch": 2.894290644459449,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.5460159662599786e-05,
+      "loss": 0.4824,
+      "step": 9646
+    },
+    {
+      "epoch": 2.894590741991147,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.544509715318572e-05,
+      "loss": 0.5369,
+      "step": 9647
+    },
+    {
+      "epoch": 2.8948908395228448,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.543003464377165e-05,
+      "loss": 0.5478,
+      "step": 9648
+    },
+    {
+      "epoch": 2.895190937054543,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.541497213435759e-05,
+      "loss": 0.463,
+      "step": 9649
+    },
+    {
+      "epoch": 2.8954910345862404,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.539990962494351e-05,
+      "loss": 0.504,
+      "step": 9650
+    },
+    {
+      "epoch": 2.8957911321179384,
+      "grad_norm": 1.0,
+      "learning_rate": 5.538484711552945e-05,
+      "loss": 0.4829,
+      "step": 9651
+    },
+    {
+      "epoch": 2.8960912296496364,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.536978460611538e-05,
+      "loss": 0.5142,
+      "step": 9652
+    },
+    {
+      "epoch": 2.896391327181334,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.5354722096701314e-05,
+      "loss": 0.4732,
+      "step": 9653
+    },
+    {
+      "epoch": 2.8966914247130315,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.533965958728724e-05,
+      "loss": 0.4804,
+      "step": 9654
+    },
+    {
+      "epoch": 2.8969915222447296,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.5324597077873174e-05,
+      "loss": 0.5375,
+      "step": 9655
+    },
+    {
+      "epoch": 2.8972916197764276,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.530953456845911e-05,
+      "loss": 0.4629,
+      "step": 9656
+    },
+    {
+      "epoch": 2.897591717308125,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.529447205904504e-05,
+      "loss": 0.4816,
+      "step": 9657
+    },
+    {
+      "epoch": 2.8978918148398227,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.527940954963097e-05,
+      "loss": 0.4177,
+      "step": 9658
+    },
+    {
+      "epoch": 2.8981919123715207,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.52643470402169e-05,
+      "loss": 0.4853,
+      "step": 9659
+    },
+    {
+      "epoch": 2.8984920099032188,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.5249284530802835e-05,
+      "loss": 0.5206,
+      "step": 9660
+    },
+    {
+      "epoch": 2.8987921074349163,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.523422202138877e-05,
+      "loss": 0.5281,
+      "step": 9661
+    },
+    {
+      "epoch": 2.899092204966614,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.5219159511974695e-05,
+      "loss": 0.4875,
+      "step": 9662
+    },
+    {
+      "epoch": 2.899392302498312,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.520409700256063e-05,
+      "loss": 0.5085,
+      "step": 9663
+    },
+    {
+      "epoch": 2.89969240003001,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.518903449314656e-05,
+      "loss": 0.5328,
+      "step": 9664
+    },
+    {
+      "epoch": 2.8999924975617075,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.5173971983732496e-05,
+      "loss": 0.4923,
+      "step": 9665
+    },
+    {
+      "epoch": 2.900292595093405,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.515890947431842e-05,
+      "loss": 0.5333,
+      "step": 9666
+    },
+    {
+      "epoch": 2.900592692625103,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.5143846964904356e-05,
+      "loss": 0.5667,
+      "step": 9667
+    },
+    {
+      "epoch": 2.900892790156801,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.512878445549029e-05,
+      "loss": 0.5224,
+      "step": 9668
+    },
+    {
+      "epoch": 2.9011928876884987,
+      "grad_norm": 1.359375,
+      "learning_rate": 5.511372194607622e-05,
+      "loss": 0.5288,
+      "step": 9669
+    },
+    {
+      "epoch": 2.9014929852201967,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.509865943666215e-05,
+      "loss": 0.5269,
+      "step": 9670
+    },
+    {
+      "epoch": 2.9017930827518943,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.5083596927248083e-05,
+      "loss": 0.5407,
+      "step": 9671
+    },
+    {
+      "epoch": 2.9020931802835923,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.506853441783402e-05,
+      "loss": 0.5629,
+      "step": 9672
+    },
+    {
+      "epoch": 2.90239327781529,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.505347190841994e-05,
+      "loss": 0.5018,
+      "step": 9673
+    },
+    {
+      "epoch": 2.902693375346988,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.503840939900588e-05,
+      "loss": 0.4922,
+      "step": 9674
+    },
+    {
+      "epoch": 2.9029934728786855,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.502334688959181e-05,
+      "loss": 0.579,
+      "step": 9675
+    },
+    {
+      "epoch": 2.9032935704103835,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.5008284380177744e-05,
+      "loss": 0.4894,
+      "step": 9676
+    },
+    {
+      "epoch": 2.903593667942081,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.4993221870763664e-05,
+      "loss": 0.5545,
+      "step": 9677
+    },
+    {
+      "epoch": 2.903893765473779,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.4978159361349605e-05,
+      "loss": 0.534,
+      "step": 9678
+    },
+    {
+      "epoch": 2.9041938630054767,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.496309685193554e-05,
+      "loss": 0.5359,
+      "step": 9679
+    },
+    {
+      "epoch": 2.9044939605371747,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.494803434252147e-05,
+      "loss": 0.518,
+      "step": 9680
+    },
+    {
+      "epoch": 2.9047940580688723,
+      "grad_norm": 1.125,
+      "learning_rate": 5.493297183310739e-05,
+      "loss": 0.5676,
+      "step": 9681
+    },
+    {
+      "epoch": 2.9050941556005703,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.4917909323693325e-05,
+      "loss": 0.6307,
+      "step": 9682
+    },
+    {
+      "epoch": 2.905394253132268,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.4902846814279265e-05,
+      "loss": 0.461,
+      "step": 9683
+    },
+    {
+      "epoch": 2.905694350663966,
+      "grad_norm": 1.3203125,
+      "learning_rate": 5.48877843048652e-05,
+      "loss": 0.4967,
+      "step": 9684
+    },
+    {
+      "epoch": 2.9059944481956634,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.487272179545112e-05,
+      "loss": 0.451,
+      "step": 9685
+    },
+    {
+      "epoch": 2.9062945457273615,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.485765928603705e-05,
+      "loss": 0.5266,
+      "step": 9686
+    },
+    {
+      "epoch": 2.906594643259059,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.484259677662299e-05,
+      "loss": 0.7337,
+      "step": 9687
+    },
+    {
+      "epoch": 2.906894740790757,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.4827534267208926e-05,
+      "loss": 0.5269,
+      "step": 9688
+    },
+    {
+      "epoch": 2.9071948383224546,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.4812471757794846e-05,
+      "loss": 0.5271,
+      "step": 9689
+    },
+    {
+      "epoch": 2.9074949358541526,
+      "grad_norm": 1.125,
+      "learning_rate": 5.479740924838078e-05,
+      "loss": 0.5407,
+      "step": 9690
+    },
+    {
+      "epoch": 2.9077950333858507,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.478234673896671e-05,
+      "loss": 0.461,
+      "step": 9691
+    },
+    {
+      "epoch": 2.9080951309175482,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.4767284229552654e-05,
+      "loss": 0.5711,
+      "step": 9692
+    },
+    {
+      "epoch": 2.908395228449246,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.4752221720138574e-05,
+      "loss": 0.4966,
+      "step": 9693
+    },
+    {
+      "epoch": 2.908695325980944,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.473715921072451e-05,
+      "loss": 0.5308,
+      "step": 9694
+    },
+    {
+      "epoch": 2.908995423512642,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.472209670131044e-05,
+      "loss": 0.5955,
+      "step": 9695
+    },
+    {
+      "epoch": 2.9092955210443394,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.470703419189638e-05,
+      "loss": 0.5218,
+      "step": 9696
+    },
+    {
+      "epoch": 2.909595618576037,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.46919716824823e-05,
+      "loss": 0.423,
+      "step": 9697
+    },
+    {
+      "epoch": 2.909895716107735,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.4676909173068234e-05,
+      "loss": 0.4851,
+      "step": 9698
+    },
+    {
+      "epoch": 2.910195813639433,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.466184666365417e-05,
+      "loss": 0.5765,
+      "step": 9699
+    },
+    {
+      "epoch": 2.9104959111711306,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.4646784154240095e-05,
+      "loss": 0.5897,
+      "step": 9700
+    },
+    {
+      "epoch": 2.910796008702828,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.463172164482603e-05,
+      "loss": 0.5508,
+      "step": 9701
+    },
+    {
+      "epoch": 2.911096106234526,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.461665913541196e-05,
+      "loss": 0.4985,
+      "step": 9702
+    },
+    {
+      "epoch": 2.911396203766224,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.4601596625997895e-05,
+      "loss": 0.5738,
+      "step": 9703
+    },
+    {
+      "epoch": 2.911696301297922,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.458653411658382e-05,
+      "loss": 0.4994,
+      "step": 9704
+    },
+    {
+      "epoch": 2.9119963988296194,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.4571471607169756e-05,
+      "loss": 0.5159,
+      "step": 9705
+    },
+    {
+      "epoch": 2.9122964963613174,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.455640909775569e-05,
+      "loss": 0.6022,
+      "step": 9706
+    },
+    {
+      "epoch": 2.9125965938930154,
+      "grad_norm": 0.92578125,
+      "learning_rate": 5.454134658834162e-05,
+      "loss": 0.4509,
+      "step": 9707
+    },
+    {
+      "epoch": 2.912896691424713,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.452628407892755e-05,
+      "loss": 0.5333,
+      "step": 9708
+    },
+    {
+      "epoch": 2.913196788956411,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.451122156951348e-05,
+      "loss": 0.4097,
+      "step": 9709
+    },
+    {
+      "epoch": 2.9134968864881086,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.4496159060099417e-05,
+      "loss": 0.5841,
+      "step": 9710
+    },
+    {
+      "epoch": 2.9137969840198066,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.448109655068535e-05,
+      "loss": 0.4964,
+      "step": 9711
+    },
+    {
+      "epoch": 2.914097081551504,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.446603404127128e-05,
+      "loss": 0.4469,
+      "step": 9712
+    },
+    {
+      "epoch": 2.914397179083202,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.445097153185721e-05,
+      "loss": 0.4253,
+      "step": 9713
+    },
+    {
+      "epoch": 2.9146972766148997,
+      "grad_norm": 0.96484375,
+      "learning_rate": 5.4435909022443144e-05,
+      "loss": 0.4682,
+      "step": 9714
+    },
+    {
+      "epoch": 2.9149973741465978,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.442084651302908e-05,
+      "loss": 0.5884,
+      "step": 9715
+    },
+    {
+      "epoch": 2.9152974716782953,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.4405784003615004e-05,
+      "loss": 0.5166,
+      "step": 9716
+    },
+    {
+      "epoch": 2.9155975692099934,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.439072149420094e-05,
+      "loss": 0.5305,
+      "step": 9717
+    },
+    {
+      "epoch": 2.915897666741691,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.437565898478687e-05,
+      "loss": 0.4595,
+      "step": 9718
+    },
+    {
+      "epoch": 2.916197764273389,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.4360596475372805e-05,
+      "loss": 0.5073,
+      "step": 9719
+    },
+    {
+      "epoch": 2.9164978618050865,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.434553396595873e-05,
+      "loss": 0.6084,
+      "step": 9720
+    },
+    {
+      "epoch": 2.9167979593367845,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.4330471456544665e-05,
+      "loss": 0.4457,
+      "step": 9721
+    },
+    {
+      "epoch": 2.917098056868482,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.43154089471306e-05,
+      "loss": 0.5263,
+      "step": 9722
+    },
+    {
+      "epoch": 2.91739815440018,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.430034643771653e-05,
+      "loss": 0.5153,
+      "step": 9723
+    },
+    {
+      "epoch": 2.9176982519318777,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.428528392830245e-05,
+      "loss": 0.5612,
+      "step": 9724
+    },
+    {
+      "epoch": 2.9179983494635757,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.427022141888839e-05,
+      "loss": 0.5058,
+      "step": 9725
+    },
+    {
+      "epoch": 2.9182984469952733,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.4255158909474326e-05,
+      "loss": 0.5043,
+      "step": 9726
+    },
+    {
+      "epoch": 2.9185985445269713,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.4240096400060246e-05,
+      "loss": 0.6082,
+      "step": 9727
+    },
+    {
+      "epoch": 2.918898642058669,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.422503389064618e-05,
+      "loss": 0.5475,
+      "step": 9728
+    },
+    {
+      "epoch": 2.919198739590367,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.420997138123212e-05,
+      "loss": 0.5799,
+      "step": 9729
+    },
+    {
+      "epoch": 2.919498837122065,
+      "grad_norm": 0.9609375,
+      "learning_rate": 5.419490887181805e-05,
+      "loss": 0.4204,
+      "step": 9730
+    },
+    {
+      "epoch": 2.9197989346537625,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.417984636240397e-05,
+      "loss": 0.5571,
+      "step": 9731
+    },
+    {
+      "epoch": 2.92009903218546,
+      "grad_norm": 1.0,
+      "learning_rate": 5.416478385298991e-05,
+      "loss": 0.5342,
+      "step": 9732
+    },
+    {
+      "epoch": 2.920399129717158,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.414972134357584e-05,
+      "loss": 0.4599,
+      "step": 9733
+    },
+    {
+      "epoch": 2.920699227248856,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.413465883416178e-05,
+      "loss": 0.5479,
+      "step": 9734
+    },
+    {
+      "epoch": 2.9209993247805537,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.41195963247477e-05,
+      "loss": 0.5045,
+      "step": 9735
+    },
+    {
+      "epoch": 2.9212994223122513,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.4104533815333634e-05,
+      "loss": 0.5058,
+      "step": 9736
+    },
+    {
+      "epoch": 2.9215995198439493,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.408947130591957e-05,
+      "loss": 0.5561,
+      "step": 9737
+    },
+    {
+      "epoch": 2.9218996173756473,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.407440879650551e-05,
+      "loss": 0.5407,
+      "step": 9738
+    },
+    {
+      "epoch": 2.922199714907345,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.405934628709143e-05,
+      "loss": 0.6311,
+      "step": 9739
+    },
+    {
+      "epoch": 2.9224998124390424,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.404428377767736e-05,
+      "loss": 0.4579,
+      "step": 9740
+    },
+    {
+      "epoch": 2.9227999099707405,
+      "grad_norm": 1.0,
+      "learning_rate": 5.4029221268263295e-05,
+      "loss": 0.431,
+      "step": 9741
+    },
+    {
+      "epoch": 2.9231000075024385,
+      "grad_norm": 1.125,
+      "learning_rate": 5.401415875884923e-05,
+      "loss": 0.5443,
+      "step": 9742
+    },
+    {
+      "epoch": 2.923400105034136,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.3999096249435155e-05,
+      "loss": 0.4958,
+      "step": 9743
+    },
+    {
+      "epoch": 2.9237002025658336,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.398403374002109e-05,
+      "loss": 0.5498,
+      "step": 9744
+    },
+    {
+      "epoch": 2.9240003000975316,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.396897123060702e-05,
+      "loss": 0.4504,
+      "step": 9745
+    },
+    {
+      "epoch": 2.9243003976292297,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.3953908721192956e-05,
+      "loss": 0.5614,
+      "step": 9746
+    },
+    {
+      "epoch": 2.9246004951609272,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.393884621177888e-05,
+      "loss": 0.583,
+      "step": 9747
+    },
+    {
+      "epoch": 2.9249005926926253,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.3923783702364816e-05,
+      "loss": 0.5699,
+      "step": 9748
+    },
+    {
+      "epoch": 2.925200690224323,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.390872119295075e-05,
+      "loss": 0.5342,
+      "step": 9749
+    },
+    {
+      "epoch": 2.925500787756021,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.389365868353668e-05,
+      "loss": 0.5495,
+      "step": 9750
+    },
+    {
+      "epoch": 2.9258008852877184,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.387859617412261e-05,
+      "loss": 0.5283,
+      "step": 9751
+    },
+    {
+      "epoch": 2.9261009828194164,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.3863533664708543e-05,
+      "loss": 0.5544,
+      "step": 9752
+    },
+    {
+      "epoch": 2.926401080351114,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.384847115529448e-05,
+      "loss": 0.5601,
+      "step": 9753
+    },
+    {
+      "epoch": 2.926701177882812,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.3833408645880404e-05,
+      "loss": 0.4602,
+      "step": 9754
+    },
+    {
+      "epoch": 2.9270012754145096,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.381834613646634e-05,
+      "loss": 0.5578,
+      "step": 9755
+    },
+    {
+      "epoch": 2.9273013729462076,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.380328362705227e-05,
+      "loss": 0.3989,
+      "step": 9756
+    },
+    {
+      "epoch": 2.927601470477905,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.3788221117638204e-05,
+      "loss": 0.5371,
+      "step": 9757
+    },
+    {
+      "epoch": 2.927901568009603,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.377315860822413e-05,
+      "loss": 0.4812,
+      "step": 9758
+    },
+    {
+      "epoch": 2.928201665541301,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.3758096098810065e-05,
+      "loss": 0.5593,
+      "step": 9759
+    },
+    {
+      "epoch": 2.928501763072999,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.3743033589396e-05,
+      "loss": 0.5557,
+      "step": 9760
+    },
+    {
+      "epoch": 2.9288018606046964,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.372797107998193e-05,
+      "loss": 0.537,
+      "step": 9761
+    },
+    {
+      "epoch": 2.9291019581363944,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.371290857056785e-05,
+      "loss": 0.4255,
+      "step": 9762
+    },
+    {
+      "epoch": 2.929402055668092,
+      "grad_norm": 1.296875,
+      "learning_rate": 5.369784606115379e-05,
+      "loss": 0.6258,
+      "step": 9763
+    },
+    {
+      "epoch": 2.92970215319979,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.3682783551739725e-05,
+      "loss": 0.4837,
+      "step": 9764
+    },
+    {
+      "epoch": 2.9300022507314876,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.366772104232566e-05,
+      "loss": 0.5032,
+      "step": 9765
+    },
+    {
+      "epoch": 2.9303023482631856,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.365265853291158e-05,
+      "loss": 0.532,
+      "step": 9766
+    },
+    {
+      "epoch": 2.930602445794883,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.363759602349752e-05,
+      "loss": 0.5305,
+      "step": 9767
+    },
+    {
+      "epoch": 2.930902543326581,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.362253351408345e-05,
+      "loss": 0.4509,
+      "step": 9768
+    },
+    {
+      "epoch": 2.931202640858279,
+      "grad_norm": 0.93359375,
+      "learning_rate": 5.3607471004669386e-05,
+      "loss": 0.5013,
+      "step": 9769
+    },
+    {
+      "epoch": 2.9315027383899768,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.3592408495255306e-05,
+      "loss": 0.4606,
+      "step": 9770
+    },
+    {
+      "epoch": 2.9318028359216743,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.357734598584124e-05,
+      "loss": 0.5082,
+      "step": 9771
+    },
+    {
+      "epoch": 2.9321029334533724,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.356228347642718e-05,
+      "loss": 0.4781,
+      "step": 9772
+    },
+    {
+      "epoch": 2.9324030309850704,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.3547220967013114e-05,
+      "loss": 0.4632,
+      "step": 9773
+    },
+    {
+      "epoch": 2.932703128516768,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.3532158457599034e-05,
+      "loss": 0.5593,
+      "step": 9774
+    },
+    {
+      "epoch": 2.9330032260484655,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.351709594818497e-05,
+      "loss": 0.5714,
+      "step": 9775
+    },
+    {
+      "epoch": 2.9333033235801635,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.350203343877091e-05,
+      "loss": 0.4549,
+      "step": 9776
+    },
+    {
+      "epoch": 2.9336034211118616,
+      "grad_norm": 1.125,
+      "learning_rate": 5.348697092935684e-05,
+      "loss": 0.5254,
+      "step": 9777
+    },
+    {
+      "epoch": 2.933903518643559,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.347190841994276e-05,
+      "loss": 0.5049,
+      "step": 9778
+    },
+    {
+      "epoch": 2.9342036161752567,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.3456845910528695e-05,
+      "loss": 0.5303,
+      "step": 9779
+    },
+    {
+      "epoch": 2.9345037137069547,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.344178340111463e-05,
+      "loss": 0.572,
+      "step": 9780
+    },
+    {
+      "epoch": 2.9348038112386527,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.3426720891700555e-05,
+      "loss": 0.463,
+      "step": 9781
+    },
+    {
+      "epoch": 2.9351039087703503,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.341165838228649e-05,
+      "loss": 0.4756,
+      "step": 9782
+    },
+    {
+      "epoch": 2.935404006302048,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.339659587287242e-05,
+      "loss": 0.5175,
+      "step": 9783
+    },
+    {
+      "epoch": 2.935704103833746,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.3381533363458355e-05,
+      "loss": 0.5615,
+      "step": 9784
+    },
+    {
+      "epoch": 2.936004201365444,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.336647085404428e-05,
+      "loss": 0.5661,
+      "step": 9785
+    },
+    {
+      "epoch": 2.9363042988971415,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.3351408344630216e-05,
+      "loss": 0.4591,
+      "step": 9786
+    },
+    {
+      "epoch": 2.9366043964288395,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.333634583521615e-05,
+      "loss": 0.4997,
+      "step": 9787
+    },
+    {
+      "epoch": 2.936904493960537,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.332128332580208e-05,
+      "loss": 0.4578,
+      "step": 9788
+    },
+    {
+      "epoch": 2.937204591492235,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.330622081638801e-05,
+      "loss": 0.5754,
+      "step": 9789
+    },
+    {
+      "epoch": 2.9375046890239327,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.329115830697394e-05,
+      "loss": 0.5093,
+      "step": 9790
+    },
+    {
+      "epoch": 2.9378047865556307,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.3276095797559877e-05,
+      "loss": 0.5515,
+      "step": 9791
+    },
+    {
+      "epoch": 2.9381048840873283,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.326103328814581e-05,
+      "loss": 0.5464,
+      "step": 9792
+    },
+    {
+      "epoch": 2.9384049816190263,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.324597077873174e-05,
+      "loss": 0.4224,
+      "step": 9793
+    },
+    {
+      "epoch": 2.938705079150724,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.323090826931767e-05,
+      "loss": 0.4687,
+      "step": 9794
+    },
+    {
+      "epoch": 2.939005176682422,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.3215845759903604e-05,
+      "loss": 0.4941,
+      "step": 9795
+    },
+    {
+      "epoch": 2.9393052742141195,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.320078325048954e-05,
+      "loss": 0.5401,
+      "step": 9796
+    },
+    {
+      "epoch": 2.9396053717458175,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.3185720741075464e-05,
+      "loss": 0.4047,
+      "step": 9797
+    },
+    {
+      "epoch": 2.939905469277515,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.31706582316614e-05,
+      "loss": 0.5594,
+      "step": 9798
+    },
+    {
+      "epoch": 2.940205566809213,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.315559572224733e-05,
+      "loss": 0.5292,
+      "step": 9799
+    },
+    {
+      "epoch": 2.9405056643409107,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.3140533212833265e-05,
+      "loss": 0.4676,
+      "step": 9800
+    },
+    {
+      "epoch": 2.9408057618726087,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.312547070341919e-05,
+      "loss": 0.6035,
+      "step": 9801
+    },
+    {
+      "epoch": 2.9411058594043062,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.3110408194005125e-05,
+      "loss": 0.5006,
+      "step": 9802
+    },
+    {
+      "epoch": 2.9414059569360043,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.309534568459106e-05,
+      "loss": 0.5779,
+      "step": 9803
+    },
+    {
+      "epoch": 2.9417060544677023,
+      "grad_norm": 1.0,
+      "learning_rate": 5.308028317517699e-05,
+      "loss": 0.5299,
+      "step": 9804
+    },
+    {
+      "epoch": 2.9420061519994,
+      "grad_norm": 1.125,
+      "learning_rate": 5.306522066576292e-05,
+      "loss": 0.5771,
+      "step": 9805
+    },
+    {
+      "epoch": 2.9423062495310974,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.305015815634885e-05,
+      "loss": 0.4773,
+      "step": 9806
+    },
+    {
+      "epoch": 2.9426063470627954,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.3035095646934786e-05,
+      "loss": 0.6363,
+      "step": 9807
+    },
+    {
+      "epoch": 2.9429064445944935,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.3020033137520706e-05,
+      "loss": 0.5808,
+      "step": 9808
+    },
+    {
+      "epoch": 2.943206542126191,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.300497062810664e-05,
+      "loss": 0.5328,
+      "step": 9809
+    },
+    {
+      "epoch": 2.9435066396578886,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.298990811869258e-05,
+      "loss": 0.4948,
+      "step": 9810
+    },
+    {
+      "epoch": 2.9438067371895866,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.297484560927851e-05,
+      "loss": 0.4966,
+      "step": 9811
+    },
+    {
+      "epoch": 2.9441068347212846,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.295978309986443e-05,
+      "loss": 0.4975,
+      "step": 9812
+    },
+    {
+      "epoch": 2.9444069322529822,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.294472059045037e-05,
+      "loss": 0.5422,
+      "step": 9813
+    },
+    {
+      "epoch": 2.94470702978468,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.292965808103631e-05,
+      "loss": 0.5465,
+      "step": 9814
+    },
+    {
+      "epoch": 2.945007127316378,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.291459557162224e-05,
+      "loss": 0.6137,
+      "step": 9815
+    },
+    {
+      "epoch": 2.945307224848076,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.289953306220816e-05,
+      "loss": 0.5694,
+      "step": 9816
+    },
+    {
+      "epoch": 2.9456073223797734,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.2884470552794094e-05,
+      "loss": 0.6079,
+      "step": 9817
+    },
+    {
+      "epoch": 2.945907419911471,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.286940804338003e-05,
+      "loss": 0.5892,
+      "step": 9818
+    },
+    {
+      "epoch": 2.946207517443169,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.285434553396597e-05,
+      "loss": 0.517,
+      "step": 9819
+    },
+    {
+      "epoch": 2.946507614974867,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.283928302455189e-05,
+      "loss": 0.4589,
+      "step": 9820
+    },
+    {
+      "epoch": 2.9468077125065646,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.282422051513782e-05,
+      "loss": 0.54,
+      "step": 9821
+    },
+    {
+      "epoch": 2.947107810038262,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.2809158005723755e-05,
+      "loss": 0.4976,
+      "step": 9822
+    },
+    {
+      "epoch": 2.94740790756996,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.2794095496309695e-05,
+      "loss": 0.3874,
+      "step": 9823
+    },
+    {
+      "epoch": 2.947708005101658,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.2779032986895615e-05,
+      "loss": 0.4532,
+      "step": 9824
+    },
+    {
+      "epoch": 2.9480081026333558,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.276397047748155e-05,
+      "loss": 0.5241,
+      "step": 9825
+    },
+    {
+      "epoch": 2.948308200165054,
+      "grad_norm": 0.9296875,
+      "learning_rate": 5.274890796806748e-05,
+      "loss": 0.4649,
+      "step": 9826
+    },
+    {
+      "epoch": 2.9486082976967514,
+      "grad_norm": 0.95703125,
+      "learning_rate": 5.2733845458653416e-05,
+      "loss": 0.446,
+      "step": 9827
+    },
+    {
+      "epoch": 2.9489083952284494,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.271878294923934e-05,
+      "loss": 0.4599,
+      "step": 9828
+    },
+    {
+      "epoch": 2.949208492760147,
+      "grad_norm": 1.0,
+      "learning_rate": 5.2703720439825276e-05,
+      "loss": 0.4791,
+      "step": 9829
+    },
+    {
+      "epoch": 2.949508590291845,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.268865793041121e-05,
+      "loss": 0.4904,
+      "step": 9830
+    },
+    {
+      "epoch": 2.9498086878235426,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.267359542099714e-05,
+      "loss": 0.4993,
+      "step": 9831
+    },
+    {
+      "epoch": 2.9501087853552406,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.265853291158307e-05,
+      "loss": 0.5839,
+      "step": 9832
+    },
+    {
+      "epoch": 2.950408882886938,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.2643470402169003e-05,
+      "loss": 0.5357,
+      "step": 9833
+    },
+    {
+      "epoch": 2.950708980418636,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.262840789275494e-05,
+      "loss": 0.4728,
+      "step": 9834
+    },
+    {
+      "epoch": 2.9510090779503337,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.2613345383340864e-05,
+      "loss": 0.5333,
+      "step": 9835
+    },
+    {
+      "epoch": 2.9513091754820318,
+      "grad_norm": 0.94921875,
+      "learning_rate": 5.25982828739268e-05,
+      "loss": 0.518,
+      "step": 9836
+    },
+    {
+      "epoch": 2.9516092730137293,
+      "grad_norm": 1.0,
+      "learning_rate": 5.258322036451273e-05,
+      "loss": 0.4505,
+      "step": 9837
+    },
+    {
+      "epoch": 2.9519093705454273,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.2568157855098664e-05,
+      "loss": 0.4187,
+      "step": 9838
+    },
+    {
+      "epoch": 2.952209468077125,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.255309534568459e-05,
+      "loss": 0.5316,
+      "step": 9839
+    },
+    {
+      "epoch": 2.952509565608823,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.2538032836270525e-05,
+      "loss": 0.4915,
+      "step": 9840
+    },
+    {
+      "epoch": 2.9528096631405205,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.252297032685646e-05,
+      "loss": 0.4918,
+      "step": 9841
+    },
+    {
+      "epoch": 2.9531097606722185,
+      "grad_norm": 1.0,
+      "learning_rate": 5.250790781744239e-05,
+      "loss": 0.5102,
+      "step": 9842
+    },
+    {
+      "epoch": 2.9534098582039165,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.249284530802832e-05,
+      "loss": 0.4666,
+      "step": 9843
+    },
+    {
+      "epoch": 2.953709955735614,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.247778279861425e-05,
+      "loss": 0.5213,
+      "step": 9844
+    },
+    {
+      "epoch": 2.9540100532673117,
+      "grad_norm": 0.9921875,
+      "learning_rate": 5.2462720289200185e-05,
+      "loss": 0.4539,
+      "step": 9845
+    },
+    {
+      "epoch": 2.9543101507990097,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.244765777978612e-05,
+      "loss": 0.5344,
+      "step": 9846
+    },
+    {
+      "epoch": 2.9546102483307077,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.2432595270372046e-05,
+      "loss": 0.4609,
+      "step": 9847
+    },
+    {
+      "epoch": 2.9549103458624053,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.241753276095798e-05,
+      "loss": 0.5558,
+      "step": 9848
+    },
+    {
+      "epoch": 2.955210443394103,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.240247025154391e-05,
+      "loss": 0.5422,
+      "step": 9849
+    },
+    {
+      "epoch": 2.955510540925801,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.2387407742129846e-05,
+      "loss": 0.4383,
+      "step": 9850
+    },
+    {
+      "epoch": 2.955810638457499,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.2372345232715766e-05,
+      "loss": 0.5222,
+      "step": 9851
+    },
+    {
+      "epoch": 2.9561107359891965,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.235728272330171e-05,
+      "loss": 0.4877,
+      "step": 9852
+    },
+    {
+      "epoch": 2.956410833520894,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.234222021388764e-05,
+      "loss": 0.4893,
+      "step": 9853
+    },
+    {
+      "epoch": 2.956710931052592,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.2327157704473574e-05,
+      "loss": 0.5628,
+      "step": 9854
+    },
+    {
+      "epoch": 2.95701102858429,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.2312095195059494e-05,
+      "loss": 0.4346,
+      "step": 9855
+    },
+    {
+      "epoch": 2.9573111261159877,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.2297032685645434e-05,
+      "loss": 0.5022,
+      "step": 9856
+    },
+    {
+      "epoch": 2.9576112236476853,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.228197017623137e-05,
+      "loss": 0.5336,
+      "step": 9857
+    },
+    {
+      "epoch": 2.9579113211793833,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.22669076668173e-05,
+      "loss": 0.4299,
+      "step": 9858
+    },
+    {
+      "epoch": 2.9582114187110813,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.225184515740322e-05,
+      "loss": 0.6275,
+      "step": 9859
+    },
+    {
+      "epoch": 2.958511516242779,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.2236782647989155e-05,
+      "loss": 0.5057,
+      "step": 9860
+    },
+    {
+      "epoch": 2.9588116137744764,
+      "grad_norm": 1.125,
+      "learning_rate": 5.2221720138575095e-05,
+      "loss": 0.5576,
+      "step": 9861
+    },
+    {
+      "epoch": 2.9591117113061745,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.2206657629161015e-05,
+      "loss": 0.4515,
+      "step": 9862
+    },
+    {
+      "epoch": 2.9594118088378725,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.219159511974695e-05,
+      "loss": 0.4495,
+      "step": 9863
+    },
+    {
+      "epoch": 2.95971190636957,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.217653261033288e-05,
+      "loss": 0.5514,
+      "step": 9864
+    },
+    {
+      "epoch": 2.960012003901268,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.216147010091882e-05,
+      "loss": 0.4825,
+      "step": 9865
+    },
+    {
+      "epoch": 2.9603121014329656,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.214640759150474e-05,
+      "loss": 0.4732,
+      "step": 9866
+    },
+    {
+      "epoch": 2.9606121989646637,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.2131345082090676e-05,
+      "loss": 0.5491,
+      "step": 9867
+    },
+    {
+      "epoch": 2.9609122964963612,
+      "grad_norm": 1.0,
+      "learning_rate": 5.211628257267661e-05,
+      "loss": 0.5008,
+      "step": 9868
+    },
+    {
+      "epoch": 2.9612123940280592,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.210122006326254e-05,
+      "loss": 0.4814,
+      "step": 9869
+    },
+    {
+      "epoch": 2.961512491559757,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.208615755384847e-05,
+      "loss": 0.551,
+      "step": 9870
+    },
+    {
+      "epoch": 2.961812589091455,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.20710950444344e-05,
+      "loss": 0.5034,
+      "step": 9871
+    },
+    {
+      "epoch": 2.9621126866231524,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.2056032535020337e-05,
+      "loss": 0.5158,
+      "step": 9872
+    },
+    {
+      "epoch": 2.9624127841548504,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.204097002560627e-05,
+      "loss": 0.5089,
+      "step": 9873
+    },
+    {
+      "epoch": 2.962712881686548,
+      "grad_norm": 0.90625,
+      "learning_rate": 5.20259075161922e-05,
+      "loss": 0.469,
+      "step": 9874
+    },
+    {
+      "epoch": 2.963012979218246,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.201084500677813e-05,
+      "loss": 0.4769,
+      "step": 9875
+    },
+    {
+      "epoch": 2.9633130767499436,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.1995782497364064e-05,
+      "loss": 0.5374,
+      "step": 9876
+    },
+    {
+      "epoch": 2.9636131742816416,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.198071998795e-05,
+      "loss": 0.4121,
+      "step": 9877
+    },
+    {
+      "epoch": 2.963913271813339,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.1965657478535924e-05,
+      "loss": 0.603,
+      "step": 9878
+    },
+    {
+      "epoch": 2.964213369345037,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.195059496912186e-05,
+      "loss": 0.5455,
+      "step": 9879
+    },
+    {
+      "epoch": 2.964513466876735,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.193553245970779e-05,
+      "loss": 0.4833,
+      "step": 9880
+    },
+    {
+      "epoch": 2.964813564408433,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.1920469950293725e-05,
+      "loss": 0.4622,
+      "step": 9881
+    },
+    {
+      "epoch": 2.965113661940131,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.190540744087965e-05,
+      "loss": 0.6148,
+      "step": 9882
+    },
+    {
+      "epoch": 2.9654137594718284,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.1890344931465585e-05,
+      "loss": 0.4933,
+      "step": 9883
+    },
+    {
+      "epoch": 2.965713857003526,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.187528242205152e-05,
+      "loss": 0.4592,
+      "step": 9884
+    },
+    {
+      "epoch": 2.966013954535224,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.186021991263745e-05,
+      "loss": 0.5468,
+      "step": 9885
+    },
+    {
+      "epoch": 2.966314052066922,
+      "grad_norm": 1.1640625,
+      "learning_rate": 5.184515740322338e-05,
+      "loss": 0.4737,
+      "step": 9886
+    },
+    {
+      "epoch": 2.9666141495986196,
+      "grad_norm": 1.2890625,
+      "learning_rate": 5.183009489380931e-05,
+      "loss": 0.6276,
+      "step": 9887
+    },
+    {
+      "epoch": 2.966914247130317,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.1815032384395246e-05,
+      "loss": 0.5092,
+      "step": 9888
+    },
+    {
+      "epoch": 2.967214344662015,
+      "grad_norm": 1.296875,
+      "learning_rate": 5.1799969874981166e-05,
+      "loss": 0.4729,
+      "step": 9889
+    },
+    {
+      "epoch": 2.967514442193713,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.1784907365567106e-05,
+      "loss": 0.4905,
+      "step": 9890
+    },
+    {
+      "epoch": 2.9678145397254108,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.176984485615304e-05,
+      "loss": 0.5204,
+      "step": 9891
+    },
+    {
+      "epoch": 2.9681146372571083,
+      "grad_norm": 1.125,
+      "learning_rate": 5.175478234673897e-05,
+      "loss": 0.5999,
+      "step": 9892
+    },
+    {
+      "epoch": 2.9684147347888064,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.173971983732489e-05,
+      "loss": 0.5351,
+      "step": 9893
+    },
+    {
+      "epoch": 2.9687148323205044,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.1724657327910834e-05,
+      "loss": 0.5095,
+      "step": 9894
+    },
+    {
+      "epoch": 2.969014929852202,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.170959481849677e-05,
+      "loss": 0.6348,
+      "step": 9895
+    },
+    {
+      "epoch": 2.9693150273838995,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.16945323090827e-05,
+      "loss": 0.4977,
+      "step": 9896
+    },
+    {
+      "epoch": 2.9696151249155975,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.167946979966862e-05,
+      "loss": 0.4294,
+      "step": 9897
+    },
+    {
+      "epoch": 2.9699152224472956,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.1664407290254554e-05,
+      "loss": 0.5117,
+      "step": 9898
+    },
+    {
+      "epoch": 2.970215319978993,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.1649344780840494e-05,
+      "loss": 0.491,
+      "step": 9899
+    },
+    {
+      "epoch": 2.9705154175106907,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.163428227142643e-05,
+      "loss": 0.552,
+      "step": 9900
+    },
+    {
+      "epoch": 2.9708155150423887,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.161921976201235e-05,
+      "loss": 0.5566,
+      "step": 9901
+    },
+    {
+      "epoch": 2.9711156125740867,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.160415725259828e-05,
+      "loss": 0.4993,
+      "step": 9902
+    },
+    {
+      "epoch": 2.9714157101057843,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.158909474318422e-05,
+      "loss": 0.5558,
+      "step": 9903
+    },
+    {
+      "epoch": 2.9717158076374823,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.1574032233770155e-05,
+      "loss": 0.5664,
+      "step": 9904
+    },
+    {
+      "epoch": 2.97201590516918,
+      "grad_norm": 1.3046875,
+      "learning_rate": 5.1558969724356075e-05,
+      "loss": 0.5143,
+      "step": 9905
+    },
+    {
+      "epoch": 2.972316002700878,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.154390721494201e-05,
+      "loss": 0.441,
+      "step": 9906
+    },
+    {
+      "epoch": 2.9726161002325755,
+      "grad_norm": 1.21875,
+      "learning_rate": 5.152884470552794e-05,
+      "loss": 0.5338,
+      "step": 9907
+    },
+    {
+      "epoch": 2.9729161977642735,
+      "grad_norm": 0.96875,
+      "learning_rate": 5.151378219611388e-05,
+      "loss": 0.483,
+      "step": 9908
+    },
+    {
+      "epoch": 2.973216295295971,
+      "grad_norm": 0.98828125,
+      "learning_rate": 5.14987196866998e-05,
+      "loss": 0.5082,
+      "step": 9909
+    },
+    {
+      "epoch": 2.973516392827669,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.1483657177285736e-05,
+      "loss": 0.5254,
+      "step": 9910
+    },
+    {
+      "epoch": 2.9738164903593667,
+      "grad_norm": 0.99609375,
+      "learning_rate": 5.146859466787167e-05,
+      "loss": 0.4595,
+      "step": 9911
+    },
+    {
+      "epoch": 2.9741165878910647,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.145353215845761e-05,
+      "loss": 0.5212,
+      "step": 9912
+    },
+    {
+      "epoch": 2.9744166854227623,
+      "grad_norm": 0.8828125,
+      "learning_rate": 5.143846964904353e-05,
+      "loss": 0.4124,
+      "step": 9913
+    },
+    {
+      "epoch": 2.9747167829544603,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.1423407139629463e-05,
+      "loss": 0.5116,
+      "step": 9914
+    },
+    {
+      "epoch": 2.975016880486158,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.14083446302154e-05,
+      "loss": 0.4551,
+      "step": 9915
+    },
+    {
+      "epoch": 2.975316978017856,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.1393282120801324e-05,
+      "loss": 0.4498,
+      "step": 9916
+    },
+    {
+      "epoch": 2.9756170755495535,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.137821961138726e-05,
+      "loss": 0.5546,
+      "step": 9917
+    },
+    {
+      "epoch": 2.9759171730812515,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.136315710197319e-05,
+      "loss": 0.5411,
+      "step": 9918
+    },
+    {
+      "epoch": 2.976217270612949,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.1348094592559124e-05,
+      "loss": 0.4472,
+      "step": 9919
+    },
+    {
+      "epoch": 2.976517368144647,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.133303208314505e-05,
+      "loss": 0.5721,
+      "step": 9920
+    },
+    {
+      "epoch": 2.976817465676345,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.1317969573730985e-05,
+      "loss": 0.5323,
+      "step": 9921
+    },
+    {
+      "epoch": 2.9771175632080427,
+      "grad_norm": 1.03125,
+      "learning_rate": 5.130290706431692e-05,
+      "loss": 0.4792,
+      "step": 9922
+    },
+    {
+      "epoch": 2.9774176607397402,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.128784455490285e-05,
+      "loss": 0.5024,
+      "step": 9923
+    },
+    {
+      "epoch": 2.9777177582714383,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.127278204548878e-05,
+      "loss": 0.632,
+      "step": 9924
+    },
+    {
+      "epoch": 2.9780178558031363,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.125771953607471e-05,
+      "loss": 0.5299,
+      "step": 9925
+    },
+    {
+      "epoch": 2.978317953334834,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.1242657026660645e-05,
+      "loss": 0.5559,
+      "step": 9926
+    },
+    {
+      "epoch": 2.9786180508665314,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.122759451724658e-05,
+      "loss": 0.4575,
+      "step": 9927
+    },
+    {
+      "epoch": 2.9789181483982294,
+      "grad_norm": 1.3046875,
+      "learning_rate": 5.1212532007832506e-05,
+      "loss": 0.5216,
+      "step": 9928
+    },
+    {
+      "epoch": 2.9792182459299275,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.119746949841844e-05,
+      "loss": 0.5495,
+      "step": 9929
+    },
+    {
+      "epoch": 2.979518343461625,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.118240698900437e-05,
+      "loss": 0.5224,
+      "step": 9930
+    },
+    {
+      "epoch": 2.9798184409933226,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.1167344479590306e-05,
+      "loss": 0.4542,
+      "step": 9931
+    },
+    {
+      "epoch": 2.9801185385250206,
+      "grad_norm": 1.171875,
+      "learning_rate": 5.115228197017623e-05,
+      "loss": 0.4933,
+      "step": 9932
+    },
+    {
+      "epoch": 2.9804186360567186,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.113721946076217e-05,
+      "loss": 0.5775,
+      "step": 9933
+    },
+    {
+      "epoch": 2.980718733588416,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.11221569513481e-05,
+      "loss": 0.4162,
+      "step": 9934
+    },
+    {
+      "epoch": 2.981018831120114,
+      "grad_norm": 1.09375,
+      "learning_rate": 5.1107094441934034e-05,
+      "loss": 0.4495,
+      "step": 9935
+    },
+    {
+      "epoch": 2.981318928651812,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.1092031932519954e-05,
+      "loss": 0.5696,
+      "step": 9936
+    },
+    {
+      "epoch": 2.98161902618351,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.1076969423105894e-05,
+      "loss": 0.5101,
+      "step": 9937
+    },
+    {
+      "epoch": 2.9819191237152074,
+      "grad_norm": 1.140625,
+      "learning_rate": 5.106190691369183e-05,
+      "loss": 0.6217,
+      "step": 9938
+    },
+    {
+      "epoch": 2.982219221246905,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.104684440427776e-05,
+      "loss": 0.5033,
+      "step": 9939
+    },
+    {
+      "epoch": 2.982519318778603,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.103178189486368e-05,
+      "loss": 0.484,
+      "step": 9940
+    },
+    {
+      "epoch": 2.982819416310301,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.101671938544962e-05,
+      "loss": 0.6115,
+      "step": 9941
+    },
+    {
+      "epoch": 2.9831195138419986,
+      "grad_norm": 1.125,
+      "learning_rate": 5.1001656876035555e-05,
+      "loss": 0.5188,
+      "step": 9942
+    },
+    {
+      "epoch": 2.9834196113736966,
+      "grad_norm": 1.125,
+      "learning_rate": 5.0986594366621475e-05,
+      "loss": 0.5399,
+      "step": 9943
+    },
+    {
+      "epoch": 2.983719708905394,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.097153185720741e-05,
+      "loss": 0.5555,
+      "step": 9944
+    },
+    {
+      "epoch": 2.984019806437092,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.095646934779334e-05,
+      "loss": 0.5952,
+      "step": 9945
+    },
+    {
+      "epoch": 2.9843199039687898,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.094140683837928e-05,
+      "loss": 0.5455,
+      "step": 9946
+    },
+    {
+      "epoch": 2.984620001500488,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.09263443289652e-05,
+      "loss": 0.5105,
+      "step": 9947
+    },
+    {
+      "epoch": 2.9849200990321854,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.0911281819551136e-05,
+      "loss": 0.4903,
+      "step": 9948
+    },
+    {
+      "epoch": 2.9852201965638834,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.089621931013707e-05,
+      "loss": 0.5064,
+      "step": 9949
+    },
+    {
+      "epoch": 2.985520294095581,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.088115680072301e-05,
+      "loss": 0.4418,
+      "step": 9950
+    },
+    {
+      "epoch": 2.985820391627279,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.086609429130893e-05,
+      "loss": 0.4747,
+      "step": 9951
+    },
+    {
+      "epoch": 2.9861204891589765,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.085103178189486e-05,
+      "loss": 0.5052,
+      "step": 9952
+    },
+    {
+      "epoch": 2.9864205866906746,
+      "grad_norm": 1.125,
+      "learning_rate": 5.0835969272480797e-05,
+      "loss": 0.4987,
+      "step": 9953
+    },
+    {
+      "epoch": 2.986720684222372,
+      "grad_norm": 1.1015625,
+      "learning_rate": 5.082090676306674e-05,
+      "loss": 0.5064,
+      "step": 9954
+    },
+    {
+      "epoch": 2.98702078175407,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.080584425365266e-05,
+      "loss": 0.5003,
+      "step": 9955
+    },
+    {
+      "epoch": 2.9873208792857677,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.079078174423859e-05,
+      "loss": 0.44,
+      "step": 9956
+    },
+    {
+      "epoch": 2.9876209768174657,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.0775719234824524e-05,
+      "loss": 0.4437,
+      "step": 9957
+    },
+    {
+      "epoch": 2.9879210743491633,
+      "grad_norm": 1.0390625,
+      "learning_rate": 5.076065672541046e-05,
+      "loss": 0.5272,
+      "step": 9958
+    },
+    {
+      "epoch": 2.9882211718808613,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.0745594215996384e-05,
+      "loss": 0.4302,
+      "step": 9959
+    },
+    {
+      "epoch": 2.9885212694125594,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.073053170658232e-05,
+      "loss": 0.4241,
+      "step": 9960
+    },
+    {
+      "epoch": 2.988821366944257,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.071546919716825e-05,
+      "loss": 0.5634,
+      "step": 9961
+    },
+    {
+      "epoch": 2.9891214644759545,
+      "grad_norm": 1.2734375,
+      "learning_rate": 5.0700406687754185e-05,
+      "loss": 0.5872,
+      "step": 9962
+    },
+    {
+      "epoch": 2.9894215620076525,
+      "grad_norm": 1.0859375,
+      "learning_rate": 5.068534417834011e-05,
+      "loss": 0.4551,
+      "step": 9963
+    },
+    {
+      "epoch": 2.9897216595393505,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.0670281668926045e-05,
+      "loss": 0.4278,
+      "step": 9964
+    },
+    {
+      "epoch": 2.990021757071048,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.065521915951198e-05,
+      "loss": 0.4951,
+      "step": 9965
+    },
+    {
+      "epoch": 2.9903218546027457,
+      "grad_norm": 0.94140625,
+      "learning_rate": 5.064015665009791e-05,
+      "loss": 0.4728,
+      "step": 9966
+    },
+    {
+      "epoch": 2.9906219521344437,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.062509414068384e-05,
+      "loss": 0.5175,
+      "step": 9967
+    },
+    {
+      "epoch": 2.9909220496661417,
+      "grad_norm": 1.0,
+      "learning_rate": 5.061003163126977e-05,
+      "loss": 0.5266,
+      "step": 9968
+    },
+    {
+      "epoch": 2.9912221471978393,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.0594969121855706e-05,
+      "loss": 0.5395,
+      "step": 9969
+    },
+    {
+      "epoch": 2.991522244729537,
+      "grad_norm": 1.078125,
+      "learning_rate": 5.057990661244163e-05,
+      "loss": 0.4888,
+      "step": 9970
+    },
+    {
+      "epoch": 2.991822342261235,
+      "grad_norm": 1.0546875,
+      "learning_rate": 5.0564844103027566e-05,
+      "loss": 0.5825,
+      "step": 9971
+    },
+    {
+      "epoch": 2.992122439792933,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.05497815936135e-05,
+      "loss": 0.4922,
+      "step": 9972
+    },
+    {
+      "epoch": 2.9924225373246305,
+      "grad_norm": 1.2265625,
+      "learning_rate": 5.053471908419943e-05,
+      "loss": 0.4914,
+      "step": 9973
+    },
+    {
+      "epoch": 2.992722634856328,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.051965657478536e-05,
+      "loss": 0.6507,
+      "step": 9974
+    },
+    {
+      "epoch": 2.993022732388026,
+      "grad_norm": 1.1484375,
+      "learning_rate": 5.0504594065371294e-05,
+      "loss": 0.5907,
+      "step": 9975
+    },
+    {
+      "epoch": 2.993322829919724,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.048953155595723e-05,
+      "loss": 0.4842,
+      "step": 9976
+    },
+    {
+      "epoch": 2.9936229274514217,
+      "grad_norm": 1.2421875,
+      "learning_rate": 5.047446904654316e-05,
+      "loss": 0.576,
+      "step": 9977
+    },
+    {
+      "epoch": 2.9939230249831192,
+      "grad_norm": 1.0,
+      "learning_rate": 5.045940653712908e-05,
+      "loss": 0.5003,
+      "step": 9978
+    },
+    {
+      "epoch": 2.9942231225148173,
+      "grad_norm": 1.0703125,
+      "learning_rate": 5.044434402771502e-05,
+      "loss": 0.5724,
+      "step": 9979
+    },
+    {
+      "epoch": 2.9945232200465153,
+      "grad_norm": 1.015625,
+      "learning_rate": 5.0429281518300954e-05,
+      "loss": 0.5237,
+      "step": 9980
+    },
+    {
+      "epoch": 2.994823317578213,
+      "grad_norm": 0.98046875,
+      "learning_rate": 5.041421900888689e-05,
+      "loss": 0.4516,
+      "step": 9981
+    },
+    {
+      "epoch": 2.995123415109911,
+      "grad_norm": 1.125,
+      "learning_rate": 5.039915649947281e-05,
+      "loss": 0.4925,
+      "step": 9982
+    },
+    {
+      "epoch": 2.9954235126416084,
+      "grad_norm": 1.1171875,
+      "learning_rate": 5.038409399005875e-05,
+      "loss": 0.5209,
+      "step": 9983
+    },
+    {
+      "epoch": 2.9957236101733065,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.036903148064468e-05,
+      "loss": 0.4471,
+      "step": 9984
+    },
+    {
+      "epoch": 2.996023707705004,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.0353968971230615e-05,
+      "loss": 0.5811,
+      "step": 9985
+    },
+    {
+      "epoch": 2.996323805236702,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.0338906461816535e-05,
+      "loss": 0.4689,
+      "step": 9986
+    },
+    {
+      "epoch": 2.9966239027683996,
+      "grad_norm": 1.0234375,
+      "learning_rate": 5.032384395240247e-05,
+      "loss": 0.4964,
+      "step": 9987
+    },
+    {
+      "epoch": 2.9969240003000976,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.030878144298841e-05,
+      "loss": 0.483,
+      "step": 9988
+    },
+    {
+      "epoch": 2.997224097831795,
+      "grad_norm": 1.1328125,
+      "learning_rate": 5.029371893357434e-05,
+      "loss": 0.609,
+      "step": 9989
+    },
+    {
+      "epoch": 2.9975241953634932,
+      "grad_norm": 1.0078125,
+      "learning_rate": 5.027865642416026e-05,
+      "loss": 0.4969,
+      "step": 9990
+    },
+    {
+      "epoch": 2.997824292895191,
+      "grad_norm": 1.15625,
+      "learning_rate": 5.0263593914746196e-05,
+      "loss": 0.6219,
+      "step": 9991
+    },
+    {
+      "epoch": 2.998124390426889,
+      "grad_norm": 1.125,
+      "learning_rate": 5.0248531405332136e-05,
+      "loss": 0.557,
+      "step": 9992
+    },
+    {
+      "epoch": 2.9984244879585864,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.023346889591807e-05,
+      "loss": 0.5778,
+      "step": 9993
+    },
+    {
+      "epoch": 2.9987245854902844,
+      "grad_norm": 1.1796875,
+      "learning_rate": 5.021840638650399e-05,
+      "loss": 0.5866,
+      "step": 9994
+    },
+    {
+      "epoch": 2.999024683021982,
+      "grad_norm": 1.046875,
+      "learning_rate": 5.0203343877089923e-05,
+      "loss": 0.5118,
+      "step": 9995
+    },
+    {
+      "epoch": 2.99932478055368,
+      "grad_norm": 1.265625,
+      "learning_rate": 5.018828136767586e-05,
+      "loss": 0.4581,
+      "step": 9996
+    },
+    {
+      "epoch": 2.9996248780853776,
+      "grad_norm": 1.109375,
+      "learning_rate": 5.0173218858261784e-05,
+      "loss": 0.4493,
+      "step": 9997
+    },
+    {
+      "epoch": 2.9999249756170756,
+      "grad_norm": 1.1875,
+      "learning_rate": 5.015815634884772e-05,
+      "loss": 0.5449,
+      "step": 9998
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.015625,
+      "learning_rate": 5.014309383943365e-05,
+      "loss": 0.4209,
+      "step": 9999
+    },
+    {
+      "epoch": 3.000300097531698,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.0128031330019584e-05,
+      "loss": 0.1569,
+      "step": 10000
+    },
+    {
+      "epoch": 3.0006001950633956,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.011296882060551e-05,
+      "loss": 0.1551,
+      "step": 10001
+    },
+    {
+      "epoch": 3.0009002925950936,
+      "grad_norm": 0.625,
+      "learning_rate": 5.0097906311191445e-05,
+      "loss": 0.1837,
+      "step": 10002
+    },
+    {
+      "epoch": 3.001200390126791,
+      "grad_norm": 0.8984375,
+      "learning_rate": 5.008284380177738e-05,
+      "loss": 0.1633,
+      "step": 10003
+    },
+    {
+      "epoch": 3.001500487658489,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.006778129236331e-05,
+      "loss": 0.1415,
+      "step": 10004
+    },
+    {
+      "epoch": 3.0018005851901868,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.005271878294924e-05,
+      "loss": 0.1682,
+      "step": 10005
+    },
+    {
+      "epoch": 3.002100682721885,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.003765627353517e-05,
+      "loss": 0.1511,
+      "step": 10006
+    },
+    {
+      "epoch": 3.0024007802535824,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.0022593764121105e-05,
+      "loss": 0.1665,
+      "step": 10007
+    },
+    {
+      "epoch": 3.0027008777852804,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.000753125470704e-05,
+      "loss": 0.1622,
+      "step": 10008
+    },
+    {
+      "epoch": 3.003000975316978,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.999246874529297e-05,
+      "loss": 0.1404,
+      "step": 10009
+    },
+    {
+      "epoch": 3.003301072848676,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.99774062358789e-05,
+      "loss": 0.1588,
+      "step": 10010
+    },
+    {
+      "epoch": 3.0036011703803736,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.996234372646483e-05,
+      "loss": 0.1206,
+      "step": 10011
+    },
+    {
+      "epoch": 3.0039012679120716,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.994728121705076e-05,
+      "loss": 0.1042,
+      "step": 10012
+    },
+    {
+      "epoch": 3.004201365443769,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.99322187076367e-05,
+      "loss": 0.1781,
+      "step": 10013
+    },
+    {
+      "epoch": 3.004501462975467,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.991715619822263e-05,
+      "loss": 0.1513,
+      "step": 10014
+    },
+    {
+      "epoch": 3.0048015605071647,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.990209368880856e-05,
+      "loss": 0.1407,
+      "step": 10015
+    },
+    {
+      "epoch": 3.0051016580388628,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.988703117939449e-05,
+      "loss": 0.1256,
+      "step": 10016
+    },
+    {
+      "epoch": 3.0054017555705603,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.987196866998042e-05,
+      "loss": 0.1357,
+      "step": 10017
+    },
+    {
+      "epoch": 3.0057018531022583,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.9856906160566354e-05,
+      "loss": 0.1353,
+      "step": 10018
+    },
+    {
+      "epoch": 3.006001950633956,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.984184365115229e-05,
+      "loss": 0.1429,
+      "step": 10019
+    },
+    {
+      "epoch": 3.006302048165654,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.9826781141738214e-05,
+      "loss": 0.1127,
+      "step": 10020
+    },
+    {
+      "epoch": 3.0066021456973515,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.981171863232415e-05,
+      "loss": 0.1303,
+      "step": 10021
+    },
+    {
+      "epoch": 3.0069022432290495,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.979665612291008e-05,
+      "loss": 0.1215,
+      "step": 10022
+    },
+    {
+      "epoch": 3.007202340760747,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.978159361349601e-05,
+      "loss": 0.1529,
+      "step": 10023
+    },
+    {
+      "epoch": 3.007502438292445,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.976653110408194e-05,
+      "loss": 0.1503,
+      "step": 10024
+    },
+    {
+      "epoch": 3.0078025358241427,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.975146859466787e-05,
+      "loss": 0.094,
+      "step": 10025
+    },
+    {
+      "epoch": 3.0081026333558407,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.973640608525381e-05,
+      "loss": 0.1402,
+      "step": 10026
+    },
+    {
+      "epoch": 3.0084027308875383,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.9721343575839735e-05,
+      "loss": 0.1079,
+      "step": 10027
+    },
+    {
+      "epoch": 3.0087028284192363,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.970628106642567e-05,
+      "loss": 0.1251,
+      "step": 10028
+    },
+    {
+      "epoch": 3.009002925950934,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.9691218557011596e-05,
+      "loss": 0.1367,
+      "step": 10029
+    },
+    {
+      "epoch": 3.009303023482632,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.9676156047597536e-05,
+      "loss": 0.0932,
+      "step": 10030
+    },
+    {
+      "epoch": 3.0096031210143295,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.966109353818346e-05,
+      "loss": 0.1337,
+      "step": 10031
+    },
+    {
+      "epoch": 3.0099032185460275,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.9646031028769396e-05,
+      "loss": 0.1509,
+      "step": 10032
+    },
+    {
+      "epoch": 3.010203316077725,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.963096851935532e-05,
+      "loss": 0.1,
+      "step": 10033
+    },
+    {
+      "epoch": 3.010503413609423,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.9615906009941257e-05,
+      "loss": 0.1361,
+      "step": 10034
+    },
+    {
+      "epoch": 3.0108035111411207,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.960084350052719e-05,
+      "loss": 0.1347,
+      "step": 10035
+    },
+    {
+      "epoch": 3.0111036086728187,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.9585780991113124e-05,
+      "loss": 0.1226,
+      "step": 10036
+    },
+    {
+      "epoch": 3.0114037062045163,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.957071848169905e-05,
+      "loss": 0.1407,
+      "step": 10037
+    },
+    {
+      "epoch": 3.0117038037362143,
+      "grad_norm": 1.0,
+      "learning_rate": 4.9555655972284984e-05,
+      "loss": 0.1948,
+      "step": 10038
+    },
+    {
+      "epoch": 3.0120039012679123,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.954059346287092e-05,
+      "loss": 0.1236,
+      "step": 10039
+    },
+    {
+      "epoch": 3.01230399879961,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.952553095345685e-05,
+      "loss": 0.0898,
+      "step": 10040
+    },
+    {
+      "epoch": 3.012604096331308,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.951046844404278e-05,
+      "loss": 0.1227,
+      "step": 10041
+    },
+    {
+      "epoch": 3.0129041938630055,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.949540593462871e-05,
+      "loss": 0.1113,
+      "step": 10042
+    },
+    {
+      "epoch": 3.0132042913947035,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.9480343425214645e-05,
+      "loss": 0.0982,
+      "step": 10043
+    },
+    {
+      "epoch": 3.013504388926401,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.946528091580058e-05,
+      "loss": 0.1367,
+      "step": 10044
+    },
+    {
+      "epoch": 3.013804486458099,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.9450218406386505e-05,
+      "loss": 0.1047,
+      "step": 10045
+    },
+    {
+      "epoch": 3.0141045839897966,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.943515589697244e-05,
+      "loss": 0.1117,
+      "step": 10046
+    },
+    {
+      "epoch": 3.0144046815214947,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.942009338755837e-05,
+      "loss": 0.1308,
+      "step": 10047
+    },
+    {
+      "epoch": 3.0147047790531922,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.9405030878144306e-05,
+      "loss": 0.1209,
+      "step": 10048
+    },
+    {
+      "epoch": 3.0150048765848902,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.938996836873023e-05,
+      "loss": 0.1117,
+      "step": 10049
+    },
+    {
+      "epoch": 3.015304974116588,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.937490585931616e-05,
+      "loss": 0.12,
+      "step": 10050
+    },
+    {
+      "epoch": 3.015605071648286,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.93598433499021e-05,
+      "loss": 0.0991,
+      "step": 10051
+    },
+    {
+      "epoch": 3.0159051691799834,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.9344780840488026e-05,
+      "loss": 0.1487,
+      "step": 10052
+    },
+    {
+      "epoch": 3.0162052667116814,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.932971833107396e-05,
+      "loss": 0.1355,
+      "step": 10053
+    },
+    {
+      "epoch": 3.016505364243379,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.9314655821659886e-05,
+      "loss": 0.1468,
+      "step": 10054
+    },
+    {
+      "epoch": 3.016805461775077,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.929959331224582e-05,
+      "loss": 0.1785,
+      "step": 10055
+    },
+    {
+      "epoch": 3.0171055593067746,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.9284530802831754e-05,
+      "loss": 0.115,
+      "step": 10056
+    },
+    {
+      "epoch": 3.0174056568384726,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.926946829341769e-05,
+      "loss": 0.1171,
+      "step": 10057
+    },
+    {
+      "epoch": 3.01770575437017,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.9254405784003614e-05,
+      "loss": 0.1182,
+      "step": 10058
+    },
+    {
+      "epoch": 3.018005851901868,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.923934327458955e-05,
+      "loss": 0.1237,
+      "step": 10059
+    },
+    {
+      "epoch": 3.018305949433566,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.922428076517548e-05,
+      "loss": 0.1231,
+      "step": 10060
+    },
+    {
+      "epoch": 3.018606046965264,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.9209218255761414e-05,
+      "loss": 0.1245,
+      "step": 10061
+    },
+    {
+      "epoch": 3.0189061444969614,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.919415574634734e-05,
+      "loss": 0.1527,
+      "step": 10062
+    },
+    {
+      "epoch": 3.0192062420286594,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.9179093236933275e-05,
+      "loss": 0.1122,
+      "step": 10063
+    },
+    {
+      "epoch": 3.019506339560357,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.916403072751921e-05,
+      "loss": 0.1282,
+      "step": 10064
+    },
+    {
+      "epoch": 3.019806437092055,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.914896821810514e-05,
+      "loss": 0.1173,
+      "step": 10065
+    },
+    {
+      "epoch": 3.0201065346237526,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.913390570869107e-05,
+      "loss": 0.1445,
+      "step": 10066
+    },
+    {
+      "epoch": 3.0204066321554506,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.9118843199277e-05,
+      "loss": 0.1289,
+      "step": 10067
+    },
+    {
+      "epoch": 3.020706729687148,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.9103780689862936e-05,
+      "loss": 0.127,
+      "step": 10068
+    },
+    {
+      "epoch": 3.021006827218846,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.908871818044887e-05,
+      "loss": 0.1151,
+      "step": 10069
+    },
+    {
+      "epoch": 3.0213069247505437,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.9073655671034796e-05,
+      "loss": 0.1445,
+      "step": 10070
+    },
+    {
+      "epoch": 3.0216070222822418,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.905859316162073e-05,
+      "loss": 0.1215,
+      "step": 10071
+    },
+    {
+      "epoch": 3.0219071198139393,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.904353065220666e-05,
+      "loss": 0.1121,
+      "step": 10072
+    },
+    {
+      "epoch": 3.0222072173456374,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.9028468142792596e-05,
+      "loss": 0.1428,
+      "step": 10073
+    },
+    {
+      "epoch": 3.022507314877335,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.901340563337852e-05,
+      "loss": 0.1065,
+      "step": 10074
+    },
+    {
+      "epoch": 3.022807412409033,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.899834312396446e-05,
+      "loss": 0.1412,
+      "step": 10075
+    },
+    {
+      "epoch": 3.0231075099407305,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.8983280614550383e-05,
+      "loss": 0.1155,
+      "step": 10076
+    },
+    {
+      "epoch": 3.0234076074724285,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.896821810513632e-05,
+      "loss": 0.1322,
+      "step": 10077
+    },
+    {
+      "epoch": 3.0237077050041266,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.895315559572225e-05,
+      "loss": 0.1143,
+      "step": 10078
+    },
+    {
+      "epoch": 3.024007802535824,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.893809308630818e-05,
+      "loss": 0.1108,
+      "step": 10079
+    },
+    {
+      "epoch": 3.024307900067522,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.892303057689411e-05,
+      "loss": 0.1227,
+      "step": 10080
+    },
+    {
+      "epoch": 3.0246079975992197,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.8907968067480044e-05,
+      "loss": 0.124,
+      "step": 10081
+    },
+    {
+      "epoch": 3.0249080951309177,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.889290555806598e-05,
+      "loss": 0.108,
+      "step": 10082
+    },
+    {
+      "epoch": 3.0252081926626153,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.8877843048651905e-05,
+      "loss": 0.1054,
+      "step": 10083
+    },
+    {
+      "epoch": 3.0255082901943133,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.886278053923784e-05,
+      "loss": 0.1294,
+      "step": 10084
+    },
+    {
+      "epoch": 3.025808387726011,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.884771802982377e-05,
+      "loss": 0.1181,
+      "step": 10085
+    },
+    {
+      "epoch": 3.026108485257709,
+      "grad_norm": 0.625,
+      "learning_rate": 4.8832655520409705e-05,
+      "loss": 0.11,
+      "step": 10086
+    },
+    {
+      "epoch": 3.0264085827894065,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.881759301099563e-05,
+      "loss": 0.0947,
+      "step": 10087
+    },
+    {
+      "epoch": 3.0267086803211045,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.8802530501581566e-05,
+      "loss": 0.1196,
+      "step": 10088
+    },
+    {
+      "epoch": 3.027008777852802,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.87874679921675e-05,
+      "loss": 0.103,
+      "step": 10089
+    },
+    {
+      "epoch": 3.0273088753845,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.877240548275343e-05,
+      "loss": 0.107,
+      "step": 10090
+    },
+    {
+      "epoch": 3.0276089729161977,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.875734297333936e-05,
+      "loss": 0.1553,
+      "step": 10091
+    },
+    {
+      "epoch": 3.0279090704478957,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.874228046392529e-05,
+      "loss": 0.1104,
+      "step": 10092
+    },
+    {
+      "epoch": 3.0282091679795933,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.872721795451122e-05,
+      "loss": 0.0964,
+      "step": 10093
+    },
+    {
+      "epoch": 3.0285092655112913,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.871215544509716e-05,
+      "loss": 0.1593,
+      "step": 10094
+    },
+    {
+      "epoch": 3.028809363042989,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.869709293568309e-05,
+      "loss": 0.1118,
+      "step": 10095
+    },
+    {
+      "epoch": 3.029109460574687,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.868203042626902e-05,
+      "loss": 0.1259,
+      "step": 10096
+    },
+    {
+      "epoch": 3.0294095581063845,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.866696791685495e-05,
+      "loss": 0.1167,
+      "step": 10097
+    },
+    {
+      "epoch": 3.0297096556380825,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.865190540744089e-05,
+      "loss": 0.1188,
+      "step": 10098
+    },
+    {
+      "epoch": 3.03000975316978,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.8636842898026814e-05,
+      "loss": 0.1149,
+      "step": 10099
+    },
+    {
+      "epoch": 3.030309850701478,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.862178038861275e-05,
+      "loss": 0.1082,
+      "step": 10100
+    },
+    {
+      "epoch": 3.0306099482331756,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.8606717879198674e-05,
+      "loss": 0.1101,
+      "step": 10101
+    },
+    {
+      "epoch": 3.0309100457648737,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.859165536978461e-05,
+      "loss": 0.1348,
+      "step": 10102
+    },
+    {
+      "epoch": 3.0312101432965712,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.857659286037054e-05,
+      "loss": 0.1134,
+      "step": 10103
+    },
+    {
+      "epoch": 3.0315102408282693,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.856153035095647e-05,
+      "loss": 0.1227,
+      "step": 10104
+    },
+    {
+      "epoch": 3.031810338359967,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.85464678415424e-05,
+      "loss": 0.1493,
+      "step": 10105
+    },
+    {
+      "epoch": 3.032110435891665,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.8531405332128335e-05,
+      "loss": 0.106,
+      "step": 10106
+    },
+    {
+      "epoch": 3.0324105334233624,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.851634282271427e-05,
+      "loss": 0.0903,
+      "step": 10107
+    },
+    {
+      "epoch": 3.0327106309550604,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.8501280313300195e-05,
+      "loss": 0.1323,
+      "step": 10108
+    },
+    {
+      "epoch": 3.033010728486758,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.848621780388613e-05,
+      "loss": 0.1335,
+      "step": 10109
+    },
+    {
+      "epoch": 3.033310826018456,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.847115529447206e-05,
+      "loss": 0.1188,
+      "step": 10110
+    },
+    {
+      "epoch": 3.0336109235501536,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.8456092785057996e-05,
+      "loss": 0.1962,
+      "step": 10111
+    },
+    {
+      "epoch": 3.0339110210818516,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.844103027564392e-05,
+      "loss": 0.0868,
+      "step": 10112
+    },
+    {
+      "epoch": 3.034211118613549,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.8425967766229856e-05,
+      "loss": 0.097,
+      "step": 10113
+    },
+    {
+      "epoch": 3.034511216145247,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.841090525681578e-05,
+      "loss": 0.1073,
+      "step": 10114
+    },
+    {
+      "epoch": 3.0348113136769452,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.839584274740172e-05,
+      "loss": 0.1201,
+      "step": 10115
+    },
+    {
+      "epoch": 3.035111411208643,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.838078023798765e-05,
+      "loss": 0.115,
+      "step": 10116
+    },
+    {
+      "epoch": 3.035411508740341,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.8365717728573584e-05,
+      "loss": 0.1215,
+      "step": 10117
+    },
+    {
+      "epoch": 3.0357116062720384,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.835065521915951e-05,
+      "loss": 0.0983,
+      "step": 10118
+    },
+    {
+      "epoch": 3.0360117038037364,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.833559270974545e-05,
+      "loss": 0.0963,
+      "step": 10119
+    },
+    {
+      "epoch": 3.036311801335434,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.832053020033138e-05,
+      "loss": 0.116,
+      "step": 10120
+    },
+    {
+      "epoch": 3.036611898867132,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.830546769091731e-05,
+      "loss": 0.0872,
+      "step": 10121
+    },
+    {
+      "epoch": 3.0369119963988296,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.829040518150324e-05,
+      "loss": 0.1058,
+      "step": 10122
+    },
+    {
+      "epoch": 3.0372120939305276,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.827534267208917e-05,
+      "loss": 0.1001,
+      "step": 10123
+    },
+    {
+      "epoch": 3.037512191462225,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.8260280162675105e-05,
+      "loss": 0.1159,
+      "step": 10124
+    },
+    {
+      "epoch": 3.037812288993923,
+      "grad_norm": 0.75,
+      "learning_rate": 4.824521765326104e-05,
+      "loss": 0.1062,
+      "step": 10125
+    },
+    {
+      "epoch": 3.0381123865256208,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.8230155143846965e-05,
+      "loss": 0.1247,
+      "step": 10126
+    },
+    {
+      "epoch": 3.038412484057319,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.82150926344329e-05,
+      "loss": 0.1338,
+      "step": 10127
+    },
+    {
+      "epoch": 3.0387125815890164,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.820003012501883e-05,
+      "loss": 0.119,
+      "step": 10128
+    },
+    {
+      "epoch": 3.0390126791207144,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.8184967615604766e-05,
+      "loss": 0.1518,
+      "step": 10129
+    },
+    {
+      "epoch": 3.039312776652412,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.816990510619069e-05,
+      "loss": 0.1227,
+      "step": 10130
+    },
+    {
+      "epoch": 3.03961287418411,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.8154842596776626e-05,
+      "loss": 0.1028,
+      "step": 10131
+    },
+    {
+      "epoch": 3.0399129717158075,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.813978008736256e-05,
+      "loss": 0.1015,
+      "step": 10132
+    },
+    {
+      "epoch": 3.0402130692475056,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.8124717577948486e-05,
+      "loss": 0.1445,
+      "step": 10133
+    },
+    {
+      "epoch": 3.040513166779203,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.810965506853442e-05,
+      "loss": 0.1248,
+      "step": 10134
+    },
+    {
+      "epoch": 3.040813264310901,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.8094592559120347e-05,
+      "loss": 0.1268,
+      "step": 10135
+    },
+    {
+      "epoch": 3.0411133618425987,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.807953004970629e-05,
+      "loss": 0.1069,
+      "step": 10136
+    },
+    {
+      "epoch": 3.0414134593742967,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.8064467540292214e-05,
+      "loss": 0.1162,
+      "step": 10137
+    },
+    {
+      "epoch": 3.0417135569059943,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.804940503087815e-05,
+      "loss": 0.1014,
+      "step": 10138
+    },
+    {
+      "epoch": 3.0420136544376923,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.8034342521464074e-05,
+      "loss": 0.1159,
+      "step": 10139
+    },
+    {
+      "epoch": 3.04231375196939,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.8019280012050014e-05,
+      "loss": 0.089,
+      "step": 10140
+    },
+    {
+      "epoch": 3.042613849501088,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.800421750263594e-05,
+      "loss": 0.1095,
+      "step": 10141
+    },
+    {
+      "epoch": 3.0429139470327855,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.7989154993221874e-05,
+      "loss": 0.0975,
+      "step": 10142
+    },
+    {
+      "epoch": 3.0432140445644835,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.79740924838078e-05,
+      "loss": 0.1234,
+      "step": 10143
+    },
+    {
+      "epoch": 3.043514142096181,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.7959029974393735e-05,
+      "loss": 0.1161,
+      "step": 10144
+    },
+    {
+      "epoch": 3.043814239627879,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.794396746497967e-05,
+      "loss": 0.1057,
+      "step": 10145
+    },
+    {
+      "epoch": 3.0441143371595767,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.79289049555656e-05,
+      "loss": 0.129,
+      "step": 10146
+    },
+    {
+      "epoch": 3.0444144346912747,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.791384244615153e-05,
+      "loss": 0.1315,
+      "step": 10147
+    },
+    {
+      "epoch": 3.0447145322229723,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.789877993673746e-05,
+      "loss": 0.1108,
+      "step": 10148
+    },
+    {
+      "epoch": 3.0450146297546703,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.7883717427323396e-05,
+      "loss": 0.1093,
+      "step": 10149
+    },
+    {
+      "epoch": 3.045314727286368,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.786865491790933e-05,
+      "loss": 0.1294,
+      "step": 10150
+    },
+    {
+      "epoch": 3.045614824818066,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.7853592408495256e-05,
+      "loss": 0.1212,
+      "step": 10151
+    },
+    {
+      "epoch": 3.0459149223497635,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.783852989908119e-05,
+      "loss": 0.0907,
+      "step": 10152
+    },
+    {
+      "epoch": 3.0462150198814615,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.782346738966712e-05,
+      "loss": 0.128,
+      "step": 10153
+    },
+    {
+      "epoch": 3.0465151174131595,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.7808404880253056e-05,
+      "loss": 0.1201,
+      "step": 10154
+    },
+    {
+      "epoch": 3.046815214944857,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.779334237083898e-05,
+      "loss": 0.0943,
+      "step": 10155
+    },
+    {
+      "epoch": 3.047115312476555,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.777827986142492e-05,
+      "loss": 0.1159,
+      "step": 10156
+    },
+    {
+      "epoch": 3.0474154100082527,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.776321735201085e-05,
+      "loss": 0.1324,
+      "step": 10157
+    },
+    {
+      "epoch": 3.0477155075399507,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.774815484259678e-05,
+      "loss": 0.1253,
+      "step": 10158
+    },
+    {
+      "epoch": 3.0480156050716483,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.773309233318271e-05,
+      "loss": 0.1016,
+      "step": 10159
+    },
+    {
+      "epoch": 3.0483157026033463,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.771802982376864e-05,
+      "loss": 0.1263,
+      "step": 10160
+    },
+    {
+      "epoch": 3.048615800135044,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.770296731435457e-05,
+      "loss": 0.1064,
+      "step": 10161
+    },
+    {
+      "epoch": 3.048915897666742,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.7687904804940504e-05,
+      "loss": 0.0971,
+      "step": 10162
+    },
+    {
+      "epoch": 3.0492159951984394,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.767284229552644e-05,
+      "loss": 0.1399,
+      "step": 10163
+    },
+    {
+      "epoch": 3.0495160927301375,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.7657779786112365e-05,
+      "loss": 0.0995,
+      "step": 10164
+    },
+    {
+      "epoch": 3.049816190261835,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.76427172766983e-05,
+      "loss": 0.1052,
+      "step": 10165
+    },
+    {
+      "epoch": 3.050116287793533,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.762765476728423e-05,
+      "loss": 0.1198,
+      "step": 10166
+    },
+    {
+      "epoch": 3.0504163853252306,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.7612592257870165e-05,
+      "loss": 0.0839,
+      "step": 10167
+    },
+    {
+      "epoch": 3.0507164828569286,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.759752974845609e-05,
+      "loss": 0.1118,
+      "step": 10168
+    },
+    {
+      "epoch": 3.051016580388626,
+      "grad_norm": 0.75,
+      "learning_rate": 4.7582467239042026e-05,
+      "loss": 0.1174,
+      "step": 10169
+    },
+    {
+      "epoch": 3.0513166779203242,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.756740472962796e-05,
+      "loss": 0.1461,
+      "step": 10170
+    },
+    {
+      "epoch": 3.051616775452022,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.755234222021389e-05,
+      "loss": 0.1344,
+      "step": 10171
+    },
+    {
+      "epoch": 3.05191687298372,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.753727971079982e-05,
+      "loss": 0.1117,
+      "step": 10172
+    },
+    {
+      "epoch": 3.0522169705154174,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.752221720138575e-05,
+      "loss": 0.1257,
+      "step": 10173
+    },
+    {
+      "epoch": 3.0525170680471154,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.7507154691971686e-05,
+      "loss": 0.1639,
+      "step": 10174
+    },
+    {
+      "epoch": 3.052817165578813,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.749209218255762e-05,
+      "loss": 0.1056,
+      "step": 10175
+    },
+    {
+      "epoch": 3.053117263110511,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.747702967314355e-05,
+      "loss": 0.1142,
+      "step": 10176
+    },
+    {
+      "epoch": 3.0534173606422086,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.746196716372948e-05,
+      "loss": 0.1248,
+      "step": 10177
+    },
+    {
+      "epoch": 3.0537174581739066,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.7446904654315414e-05,
+      "loss": 0.1504,
+      "step": 10178
+    },
+    {
+      "epoch": 3.054017555705604,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.743184214490135e-05,
+      "loss": 0.1261,
+      "step": 10179
+    },
+    {
+      "epoch": 3.054317653237302,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.7416779635487274e-05,
+      "loss": 0.1413,
+      "step": 10180
+    },
+    {
+      "epoch": 3.0546177507689998,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.740171712607321e-05,
+      "loss": 0.1407,
+      "step": 10181
+    },
+    {
+      "epoch": 3.054917848300698,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.7386654616659134e-05,
+      "loss": 0.1599,
+      "step": 10182
+    },
+    {
+      "epoch": 3.0552179458323954,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.7371592107245075e-05,
+      "loss": 0.1058,
+      "step": 10183
+    },
+    {
+      "epoch": 3.0555180433640934,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.7356529597831e-05,
+      "loss": 0.125,
+      "step": 10184
+    },
+    {
+      "epoch": 3.055818140895791,
+      "grad_norm": 0.75,
+      "learning_rate": 4.734146708841693e-05,
+      "loss": 0.1201,
+      "step": 10185
+    },
+    {
+      "epoch": 3.056118238427489,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.732640457900286e-05,
+      "loss": 0.1215,
+      "step": 10186
+    },
+    {
+      "epoch": 3.0564183359591865,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.7311342069588795e-05,
+      "loss": 0.1138,
+      "step": 10187
+    },
+    {
+      "epoch": 3.0567184334908846,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.729627956017473e-05,
+      "loss": 0.1139,
+      "step": 10188
+    },
+    {
+      "epoch": 3.057018531022582,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.7281217050760655e-05,
+      "loss": 0.1305,
+      "step": 10189
+    },
+    {
+      "epoch": 3.05731862855428,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.726615454134659e-05,
+      "loss": 0.1172,
+      "step": 10190
+    },
+    {
+      "epoch": 3.0576187260859777,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.725109203193252e-05,
+      "loss": 0.1168,
+      "step": 10191
+    },
+    {
+      "epoch": 3.0579188236176758,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.7236029522518456e-05,
+      "loss": 0.1081,
+      "step": 10192
+    },
+    {
+      "epoch": 3.0582189211493738,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.722096701310438e-05,
+      "loss": 0.1028,
+      "step": 10193
+    },
+    {
+      "epoch": 3.0585190186810713,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.7205904503690316e-05,
+      "loss": 0.1126,
+      "step": 10194
+    },
+    {
+      "epoch": 3.0588191162127694,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.719084199427625e-05,
+      "loss": 0.139,
+      "step": 10195
+    },
+    {
+      "epoch": 3.059119213744467,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.717577948486218e-05,
+      "loss": 0.1188,
+      "step": 10196
+    },
+    {
+      "epoch": 3.059419311276165,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.716071697544811e-05,
+      "loss": 0.1072,
+      "step": 10197
+    },
+    {
+      "epoch": 3.0597194088078625,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.7145654466034044e-05,
+      "loss": 0.1426,
+      "step": 10198
+    },
+    {
+      "epoch": 3.0600195063395605,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.713059195661998e-05,
+      "loss": 0.1172,
+      "step": 10199
+    },
+    {
+      "epoch": 3.060319603871258,
+      "grad_norm": 1.125,
+      "learning_rate": 4.711552944720591e-05,
+      "loss": 0.1819,
+      "step": 10200
+    },
+    {
+      "epoch": 3.060619701402956,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.710046693779184e-05,
+      "loss": 0.0929,
+      "step": 10201
+    },
+    {
+      "epoch": 3.0609197989346537,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.708540442837777e-05,
+      "loss": 0.1317,
+      "step": 10202
+    },
+    {
+      "epoch": 3.0612198964663517,
+      "grad_norm": 0.75,
+      "learning_rate": 4.70703419189637e-05,
+      "loss": 0.1031,
+      "step": 10203
+    },
+    {
+      "epoch": 3.0615199939980493,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.705527940954964e-05,
+      "loss": 0.107,
+      "step": 10204
+    },
+    {
+      "epoch": 3.0618200915297473,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.7040216900135565e-05,
+      "loss": 0.1116,
+      "step": 10205
+    },
+    {
+      "epoch": 3.062120189061445,
+      "grad_norm": 0.75,
+      "learning_rate": 4.70251543907215e-05,
+      "loss": 0.1054,
+      "step": 10206
+    },
+    {
+      "epoch": 3.062420286593143,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.7010091881307425e-05,
+      "loss": 0.117,
+      "step": 10207
+    },
+    {
+      "epoch": 3.0627203841248405,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.6995029371893365e-05,
+      "loss": 0.1127,
+      "step": 10208
+    },
+    {
+      "epoch": 3.0630204816565385,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.697996686247929e-05,
+      "loss": 0.0984,
+      "step": 10209
+    },
+    {
+      "epoch": 3.063320579188236,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.6964904353065226e-05,
+      "loss": 0.1185,
+      "step": 10210
+    },
+    {
+      "epoch": 3.063620676719934,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.694984184365115e-05,
+      "loss": 0.0887,
+      "step": 10211
+    },
+    {
+      "epoch": 3.0639207742516317,
+      "grad_norm": 0.75,
+      "learning_rate": 4.6934779334237086e-05,
+      "loss": 0.1255,
+      "step": 10212
+    },
+    {
+      "epoch": 3.0642208717833297,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.691971682482302e-05,
+      "loss": 0.1243,
+      "step": 10213
+    },
+    {
+      "epoch": 3.0645209693150273,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.6904654315408946e-05,
+      "loss": 0.1418,
+      "step": 10214
+    },
+    {
+      "epoch": 3.0648210668467253,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.688959180599488e-05,
+      "loss": 0.1095,
+      "step": 10215
+    },
+    {
+      "epoch": 3.065121164378423,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.687452929658081e-05,
+      "loss": 0.1261,
+      "step": 10216
+    },
+    {
+      "epoch": 3.065421261910121,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.685946678716675e-05,
+      "loss": 0.1226,
+      "step": 10217
+    },
+    {
+      "epoch": 3.0657213594418184,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.6844404277752674e-05,
+      "loss": 0.1449,
+      "step": 10218
+    },
+    {
+      "epoch": 3.0660214569735165,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.682934176833861e-05,
+      "loss": 0.1056,
+      "step": 10219
+    },
+    {
+      "epoch": 3.066321554505214,
+      "grad_norm": 0.75,
+      "learning_rate": 4.6814279258924534e-05,
+      "loss": 0.1195,
+      "step": 10220
+    },
+    {
+      "epoch": 3.066621652036912,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.6799216749510474e-05,
+      "loss": 0.1071,
+      "step": 10221
+    },
+    {
+      "epoch": 3.0669217495686096,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.67841542400964e-05,
+      "loss": 0.1033,
+      "step": 10222
+    },
+    {
+      "epoch": 3.0672218471003077,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.6769091730682334e-05,
+      "loss": 0.1106,
+      "step": 10223
+    },
+    {
+      "epoch": 3.0675219446320052,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.675402922126826e-05,
+      "loss": 0.0932,
+      "step": 10224
+    },
+    {
+      "epoch": 3.0678220421637032,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.67389667118542e-05,
+      "loss": 0.1359,
+      "step": 10225
+    },
+    {
+      "epoch": 3.068122139695401,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.672390420244013e-05,
+      "loss": 0.1122,
+      "step": 10226
+    },
+    {
+      "epoch": 3.068422237227099,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.670884169302606e-05,
+      "loss": 0.1194,
+      "step": 10227
+    },
+    {
+      "epoch": 3.0687223347587964,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.669377918361199e-05,
+      "loss": 0.1067,
+      "step": 10228
+    },
+    {
+      "epoch": 3.0690224322904944,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.667871667419792e-05,
+      "loss": 0.098,
+      "step": 10229
+    },
+    {
+      "epoch": 3.069322529822192,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.6663654164783856e-05,
+      "loss": 0.1214,
+      "step": 10230
+    },
+    {
+      "epoch": 3.06962262735389,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.664859165536979e-05,
+      "loss": 0.1142,
+      "step": 10231
+    },
+    {
+      "epoch": 3.069922724885588,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.6633529145955716e-05,
+      "loss": 0.0976,
+      "step": 10232
+    },
+    {
+      "epoch": 3.0702228224172856,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.661846663654165e-05,
+      "loss": 0.1155,
+      "step": 10233
+    },
+    {
+      "epoch": 3.0705229199489836,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.660340412712758e-05,
+      "loss": 0.1121,
+      "step": 10234
+    },
+    {
+      "epoch": 3.070823017480681,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.6588341617713516e-05,
+      "loss": 0.099,
+      "step": 10235
+    },
+    {
+      "epoch": 3.0711231150123792,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.657327910829944e-05,
+      "loss": 0.123,
+      "step": 10236
+    },
+    {
+      "epoch": 3.071423212544077,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.655821659888538e-05,
+      "loss": 0.1436,
+      "step": 10237
+    },
+    {
+      "epoch": 3.071723310075775,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.654315408947131e-05,
+      "loss": 0.1235,
+      "step": 10238
+    },
+    {
+      "epoch": 3.0720234076074724,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.652809158005724e-05,
+      "loss": 0.1163,
+      "step": 10239
+    },
+    {
+      "epoch": 3.0723235051391704,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.651302907064317e-05,
+      "loss": 0.1292,
+      "step": 10240
+    },
+    {
+      "epoch": 3.072623602670868,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.64979665612291e-05,
+      "loss": 0.1158,
+      "step": 10241
+    },
+    {
+      "epoch": 3.072923700202566,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.648290405181504e-05,
+      "loss": 0.1254,
+      "step": 10242
+    },
+    {
+      "epoch": 3.0732237977342636,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.6467841542400964e-05,
+      "loss": 0.0945,
+      "step": 10243
+    },
+    {
+      "epoch": 3.0735238952659616,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.64527790329869e-05,
+      "loss": 0.1097,
+      "step": 10244
+    },
+    {
+      "epoch": 3.073823992797659,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.6437716523572825e-05,
+      "loss": 0.1177,
+      "step": 10245
+    },
+    {
+      "epoch": 3.074124090329357,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.6422654014158765e-05,
+      "loss": 0.1338,
+      "step": 10246
+    },
+    {
+      "epoch": 3.0744241878610548,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.640759150474469e-05,
+      "loss": 0.1021,
+      "step": 10247
+    },
+    {
+      "epoch": 3.0747242853927528,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.6392528995330625e-05,
+      "loss": 0.1391,
+      "step": 10248
+    },
+    {
+      "epoch": 3.0750243829244504,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.637746648591655e-05,
+      "loss": 0.1285,
+      "step": 10249
+    },
+    {
+      "epoch": 3.0753244804561484,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.6362403976502486e-05,
+      "loss": 0.0897,
+      "step": 10250
+    },
+    {
+      "epoch": 3.075624577987846,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.634734146708842e-05,
+      "loss": 0.1188,
+      "step": 10251
+    },
+    {
+      "epoch": 3.075924675519544,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.633227895767435e-05,
+      "loss": 0.1149,
+      "step": 10252
+    },
+    {
+      "epoch": 3.0762247730512415,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.631721644826028e-05,
+      "loss": 0.1105,
+      "step": 10253
+    },
+    {
+      "epoch": 3.0765248705829396,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.630215393884621e-05,
+      "loss": 0.1203,
+      "step": 10254
+    },
+    {
+      "epoch": 3.076824968114637,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.6287091429432146e-05,
+      "loss": 0.132,
+      "step": 10255
+    },
+    {
+      "epoch": 3.077125065646335,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.627202892001808e-05,
+      "loss": 0.1387,
+      "step": 10256
+    },
+    {
+      "epoch": 3.0774251631780327,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.625696641060401e-05,
+      "loss": 0.1093,
+      "step": 10257
+    },
+    {
+      "epoch": 3.0777252607097307,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.624190390118994e-05,
+      "loss": 0.1027,
+      "step": 10258
+    },
+    {
+      "epoch": 3.0780253582414283,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.6226841391775874e-05,
+      "loss": 0.0877,
+      "step": 10259
+    },
+    {
+      "epoch": 3.0783254557731263,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.621177888236181e-05,
+      "loss": 0.2376,
+      "step": 10260
+    },
+    {
+      "epoch": 3.078625553304824,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.6196716372947734e-05,
+      "loss": 0.1002,
+      "step": 10261
+    },
+    {
+      "epoch": 3.078925650836522,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.618165386353367e-05,
+      "loss": 0.1745,
+      "step": 10262
+    },
+    {
+      "epoch": 3.0792257483682195,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.61665913541196e-05,
+      "loss": 0.1072,
+      "step": 10263
+    },
+    {
+      "epoch": 3.0795258458999175,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.6151528844705535e-05,
+      "loss": 0.1011,
+      "step": 10264
+    },
+    {
+      "epoch": 3.079825943431615,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.613646633529146e-05,
+      "loss": 0.1023,
+      "step": 10265
+    },
+    {
+      "epoch": 3.080126040963313,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.612140382587739e-05,
+      "loss": 0.1141,
+      "step": 10266
+    },
+    {
+      "epoch": 3.0804261384950107,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.610634131646333e-05,
+      "loss": 0.1225,
+      "step": 10267
+    },
+    {
+      "epoch": 3.0807262360267087,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.6091278807049255e-05,
+      "loss": 0.1,
+      "step": 10268
+    },
+    {
+      "epoch": 3.0810263335584063,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.607621629763519e-05,
+      "loss": 0.1201,
+      "step": 10269
+    },
+    {
+      "epoch": 3.0813264310901043,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.6061153788221115e-05,
+      "loss": 0.1158,
+      "step": 10270
+    },
+    {
+      "epoch": 3.0816265286218023,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.604609127880705e-05,
+      "loss": 0.1586,
+      "step": 10271
+    },
+    {
+      "epoch": 3.0819266261535,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.603102876939298e-05,
+      "loss": 0.1114,
+      "step": 10272
+    },
+    {
+      "epoch": 3.082226723685198,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.6015966259978916e-05,
+      "loss": 0.1256,
+      "step": 10273
+    },
+    {
+      "epoch": 3.0825268212168955,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.600090375056484e-05,
+      "loss": 0.1261,
+      "step": 10274
+    },
+    {
+      "epoch": 3.0828269187485935,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.5985841241150776e-05,
+      "loss": 0.1017,
+      "step": 10275
+    },
+    {
+      "epoch": 3.083127016280291,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.597077873173671e-05,
+      "loss": 0.0841,
+      "step": 10276
+    },
+    {
+      "epoch": 3.083427113811989,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.5955716222322643e-05,
+      "loss": 0.0939,
+      "step": 10277
+    },
+    {
+      "epoch": 3.0837272113436867,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.594065371290857e-05,
+      "loss": 0.1473,
+      "step": 10278
+    },
+    {
+      "epoch": 3.0840273088753847,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.5925591203494504e-05,
+      "loss": 0.1068,
+      "step": 10279
+    },
+    {
+      "epoch": 3.0843274064070823,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.591052869408044e-05,
+      "loss": 0.128,
+      "step": 10280
+    },
+    {
+      "epoch": 3.0846275039387803,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.589546618466637e-05,
+      "loss": 0.1082,
+      "step": 10281
+    },
+    {
+      "epoch": 3.084927601470478,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.58804036752523e-05,
+      "loss": 0.0888,
+      "step": 10282
+    },
+    {
+      "epoch": 3.085227699002176,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.586534116583823e-05,
+      "loss": 0.1162,
+      "step": 10283
+    },
+    {
+      "epoch": 3.0855277965338734,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.5850278656424165e-05,
+      "loss": 0.1128,
+      "step": 10284
+    },
+    {
+      "epoch": 3.0858278940655715,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.58352161470101e-05,
+      "loss": 0.0967,
+      "step": 10285
+    },
+    {
+      "epoch": 3.086127991597269,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.5820153637596025e-05,
+      "loss": 0.1076,
+      "step": 10286
+    },
+    {
+      "epoch": 3.086428089128967,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.580509112818196e-05,
+      "loss": 0.1161,
+      "step": 10287
+    },
+    {
+      "epoch": 3.0867281866606646,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.5790028618767885e-05,
+      "loss": 0.096,
+      "step": 10288
+    },
+    {
+      "epoch": 3.0870282841923626,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.5774966109353825e-05,
+      "loss": 0.111,
+      "step": 10289
+    },
+    {
+      "epoch": 3.08732838172406,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.575990359993975e-05,
+      "loss": 0.123,
+      "step": 10290
+    },
+    {
+      "epoch": 3.0876284792557582,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.5744841090525686e-05,
+      "loss": 0.135,
+      "step": 10291
+    },
+    {
+      "epoch": 3.087928576787456,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.572977858111161e-05,
+      "loss": 0.1021,
+      "step": 10292
+    },
+    {
+      "epoch": 3.088228674319154,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.571471607169755e-05,
+      "loss": 0.0851,
+      "step": 10293
+    },
+    {
+      "epoch": 3.0885287718508514,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.569965356228348e-05,
+      "loss": 0.1076,
+      "step": 10294
+    },
+    {
+      "epoch": 3.0888288693825494,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.5684591052869406e-05,
+      "loss": 0.1142,
+      "step": 10295
+    },
+    {
+      "epoch": 3.089128966914247,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.566952854345534e-05,
+      "loss": 0.095,
+      "step": 10296
+    },
+    {
+      "epoch": 3.089429064445945,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.565446603404127e-05,
+      "loss": 0.0878,
+      "step": 10297
+    },
+    {
+      "epoch": 3.0897291619776426,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.563940352462721e-05,
+      "loss": 0.0936,
+      "step": 10298
+    },
+    {
+      "epoch": 3.0900292595093406,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.5624341015213134e-05,
+      "loss": 0.1228,
+      "step": 10299
+    },
+    {
+      "epoch": 3.090329357041038,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.560927850579907e-05,
+      "loss": 0.1019,
+      "step": 10300
+    },
+    {
+      "epoch": 3.090629454572736,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.5594215996385e-05,
+      "loss": 0.1138,
+      "step": 10301
+    },
+    {
+      "epoch": 3.0909295521044338,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.5579153486970934e-05,
+      "loss": 0.1202,
+      "step": 10302
+    },
+    {
+      "epoch": 3.091229649636132,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.556409097755686e-05,
+      "loss": 0.1387,
+      "step": 10303
+    },
+    {
+      "epoch": 3.0915297471678294,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.5549028468142794e-05,
+      "loss": 0.095,
+      "step": 10304
+    },
+    {
+      "epoch": 3.0918298446995274,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.553396595872873e-05,
+      "loss": 0.1519,
+      "step": 10305
+    },
+    {
+      "epoch": 3.092129942231225,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.551890344931466e-05,
+      "loss": 0.1048,
+      "step": 10306
+    },
+    {
+      "epoch": 3.092430039762923,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.550384093990059e-05,
+      "loss": 0.1325,
+      "step": 10307
+    },
+    {
+      "epoch": 3.0927301372946205,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.548877843048652e-05,
+      "loss": 0.1258,
+      "step": 10308
+    },
+    {
+      "epoch": 3.0930302348263186,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.547371592107245e-05,
+      "loss": 0.1004,
+      "step": 10309
+    },
+    {
+      "epoch": 3.0933303323580166,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.545865341165839e-05,
+      "loss": 0.147,
+      "step": 10310
+    },
+    {
+      "epoch": 3.093630429889714,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.5443590902244316e-05,
+      "loss": 0.0888,
+      "step": 10311
+    },
+    {
+      "epoch": 3.093930527421412,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.542852839283025e-05,
+      "loss": 0.1148,
+      "step": 10312
+    },
+    {
+      "epoch": 3.0942306249531097,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.5413465883416176e-05,
+      "loss": 0.1216,
+      "step": 10313
+    },
+    {
+      "epoch": 3.0945307224848078,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.5398403374002116e-05,
+      "loss": 0.0952,
+      "step": 10314
+    },
+    {
+      "epoch": 3.0948308200165053,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.538334086458804e-05,
+      "loss": 0.0994,
+      "step": 10315
+    },
+    {
+      "epoch": 3.0951309175482034,
+      "grad_norm": 0.75,
+      "learning_rate": 4.5368278355173976e-05,
+      "loss": 0.1201,
+      "step": 10316
+    },
+    {
+      "epoch": 3.095431015079901,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.53532158457599e-05,
+      "loss": 0.1005,
+      "step": 10317
+    },
+    {
+      "epoch": 3.095731112611599,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.533815333634584e-05,
+      "loss": 0.168,
+      "step": 10318
+    },
+    {
+      "epoch": 3.0960312101432965,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.532309082693177e-05,
+      "loss": 0.1438,
+      "step": 10319
+    },
+    {
+      "epoch": 3.0963313076749945,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.5308028317517704e-05,
+      "loss": 0.1106,
+      "step": 10320
+    },
+    {
+      "epoch": 3.096631405206692,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.529296580810363e-05,
+      "loss": 0.1168,
+      "step": 10321
+    },
+    {
+      "epoch": 3.09693150273839,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.5277903298689564e-05,
+      "loss": 0.1258,
+      "step": 10322
+    },
+    {
+      "epoch": 3.0972316002700877,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.52628407892755e-05,
+      "loss": 0.1595,
+      "step": 10323
+    },
+    {
+      "epoch": 3.0975316978017857,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.5247778279861424e-05,
+      "loss": 0.1361,
+      "step": 10324
+    },
+    {
+      "epoch": 3.0978317953334833,
+      "grad_norm": 1.25,
+      "learning_rate": 4.523271577044736e-05,
+      "loss": 0.1513,
+      "step": 10325
+    },
+    {
+      "epoch": 3.0981318928651813,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.521765326103329e-05,
+      "loss": 0.1097,
+      "step": 10326
+    },
+    {
+      "epoch": 3.098431990396879,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.5202590751619225e-05,
+      "loss": 0.1105,
+      "step": 10327
+    },
+    {
+      "epoch": 3.098732087928577,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.518752824220515e-05,
+      "loss": 0.1151,
+      "step": 10328
+    },
+    {
+      "epoch": 3.0990321854602745,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.5172465732791085e-05,
+      "loss": 0.119,
+      "step": 10329
+    },
+    {
+      "epoch": 3.0993322829919725,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.515740322337701e-05,
+      "loss": 0.0991,
+      "step": 10330
+    },
+    {
+      "epoch": 3.09963238052367,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.514234071396295e-05,
+      "loss": 0.1057,
+      "step": 10331
+    },
+    {
+      "epoch": 3.099932478055368,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.512727820454888e-05,
+      "loss": 0.1162,
+      "step": 10332
+    },
+    {
+      "epoch": 3.1002325755870657,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.511221569513481e-05,
+      "loss": 0.112,
+      "step": 10333
+    },
+    {
+      "epoch": 3.1005326731187637,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.509715318572074e-05,
+      "loss": 0.1125,
+      "step": 10334
+    },
+    {
+      "epoch": 3.1008327706504613,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.508209067630668e-05,
+      "loss": 0.105,
+      "step": 10335
+    },
+    {
+      "epoch": 3.1011328681821593,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.5067028166892606e-05,
+      "loss": 0.1304,
+      "step": 10336
+    },
+    {
+      "epoch": 3.101432965713857,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.505196565747854e-05,
+      "loss": 0.1628,
+      "step": 10337
+    },
+    {
+      "epoch": 3.101733063245555,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.503690314806447e-05,
+      "loss": 0.1193,
+      "step": 10338
+    },
+    {
+      "epoch": 3.1020331607772524,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.50218406386504e-05,
+      "loss": 0.1327,
+      "step": 10339
+    },
+    {
+      "epoch": 3.1023332583089505,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.5006778129236334e-05,
+      "loss": 0.1092,
+      "step": 10340
+    },
+    {
+      "epoch": 3.102633355840648,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.499171561982227e-05,
+      "loss": 0.1113,
+      "step": 10341
+    },
+    {
+      "epoch": 3.102933453372346,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.4976653110408194e-05,
+      "loss": 0.103,
+      "step": 10342
+    },
+    {
+      "epoch": 3.1032335509040436,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.496159060099413e-05,
+      "loss": 0.1385,
+      "step": 10343
+    },
+    {
+      "epoch": 3.1035336484357416,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.494652809158006e-05,
+      "loss": 0.128,
+      "step": 10344
+    },
+    {
+      "epoch": 3.103833745967439,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.4931465582165995e-05,
+      "loss": 0.1034,
+      "step": 10345
+    },
+    {
+      "epoch": 3.1041338434991372,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.491640307275192e-05,
+      "loss": 0.1066,
+      "step": 10346
+    },
+    {
+      "epoch": 3.104433941030835,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.4901340563337855e-05,
+      "loss": 0.0854,
+      "step": 10347
+    },
+    {
+      "epoch": 3.104734038562533,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.488627805392379e-05,
+      "loss": 0.1231,
+      "step": 10348
+    },
+    {
+      "epoch": 3.105034136094231,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.4871215544509715e-05,
+      "loss": 0.1096,
+      "step": 10349
+    },
+    {
+      "epoch": 3.1053342336259284,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.485615303509565e-05,
+      "loss": 0.1005,
+      "step": 10350
+    },
+    {
+      "epoch": 3.1056343311576264,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.4841090525681575e-05,
+      "loss": 0.1098,
+      "step": 10351
+    },
+    {
+      "epoch": 3.105934428689324,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.4826028016267516e-05,
+      "loss": 0.1173,
+      "step": 10352
+    },
+    {
+      "epoch": 3.106234526221022,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.481096550685344e-05,
+      "loss": 0.1567,
+      "step": 10353
+    },
+    {
+      "epoch": 3.1065346237527196,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.4795902997439376e-05,
+      "loss": 0.1132,
+      "step": 10354
+    },
+    {
+      "epoch": 3.1068347212844176,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.47808404880253e-05,
+      "loss": 0.1054,
+      "step": 10355
+    },
+    {
+      "epoch": 3.107134818816115,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.4765777978611236e-05,
+      "loss": 0.1009,
+      "step": 10356
+    },
+    {
+      "epoch": 3.107434916347813,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.475071546919717e-05,
+      "loss": 0.1006,
+      "step": 10357
+    },
+    {
+      "epoch": 3.107735013879511,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.4735652959783103e-05,
+      "loss": 0.114,
+      "step": 10358
+    },
+    {
+      "epoch": 3.108035111411209,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.472059045036903e-05,
+      "loss": 0.1385,
+      "step": 10359
+    },
+    {
+      "epoch": 3.1083352089429064,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.4705527940954964e-05,
+      "loss": 0.1025,
+      "step": 10360
+    },
+    {
+      "epoch": 3.1086353064746044,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.46904654315409e-05,
+      "loss": 0.106,
+      "step": 10361
+    },
+    {
+      "epoch": 3.108935404006302,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.467540292212683e-05,
+      "loss": 0.1226,
+      "step": 10362
+    },
+    {
+      "epoch": 3.109235501538,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.466034041271276e-05,
+      "loss": 0.127,
+      "step": 10363
+    },
+    {
+      "epoch": 3.1095355990696976,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.464527790329869e-05,
+      "loss": 0.171,
+      "step": 10364
+    },
+    {
+      "epoch": 3.1098356966013956,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.4630215393884625e-05,
+      "loss": 0.1084,
+      "step": 10365
+    },
+    {
+      "epoch": 3.110135794133093,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.461515288447056e-05,
+      "loss": 0.1069,
+      "step": 10366
+    },
+    {
+      "epoch": 3.110435891664791,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.4600090375056485e-05,
+      "loss": 0.1141,
+      "step": 10367
+    },
+    {
+      "epoch": 3.1107359891964887,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.458502786564242e-05,
+      "loss": 0.1802,
+      "step": 10368
+    },
+    {
+      "epoch": 3.1110360867281868,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.456996535622835e-05,
+      "loss": 0.1058,
+      "step": 10369
+    },
+    {
+      "epoch": 3.1113361842598843,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.4554902846814285e-05,
+      "loss": 0.1002,
+      "step": 10370
+    },
+    {
+      "epoch": 3.1116362817915824,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.453984033740021e-05,
+      "loss": 0.0999,
+      "step": 10371
+    },
+    {
+      "epoch": 3.11193637932328,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.4524777827986146e-05,
+      "loss": 0.1246,
+      "step": 10372
+    },
+    {
+      "epoch": 3.112236476854978,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.450971531857208e-05,
+      "loss": 0.121,
+      "step": 10373
+    },
+    {
+      "epoch": 3.1125365743866755,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.449465280915801e-05,
+      "loss": 0.1106,
+      "step": 10374
+    },
+    {
+      "epoch": 3.1128366719183735,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.447959029974394e-05,
+      "loss": 0.1137,
+      "step": 10375
+    },
+    {
+      "epoch": 3.113136769450071,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.4464527790329866e-05,
+      "loss": 0.1201,
+      "step": 10376
+    },
+    {
+      "epoch": 3.113436866981769,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.44494652809158e-05,
+      "loss": 0.1278,
+      "step": 10377
+    },
+    {
+      "epoch": 3.1137369645134667,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.443440277150173e-05,
+      "loss": 0.1551,
+      "step": 10378
+    },
+    {
+      "epoch": 3.1140370620451647,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.441934026208767e-05,
+      "loss": 0.1185,
+      "step": 10379
+    },
+    {
+      "epoch": 3.1143371595768623,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.4404277752673594e-05,
+      "loss": 0.0939,
+      "step": 10380
+    },
+    {
+      "epoch": 3.1146372571085603,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.438921524325953e-05,
+      "loss": 0.1163,
+      "step": 10381
+    },
+    {
+      "epoch": 3.114937354640258,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.437415273384546e-05,
+      "loss": 0.12,
+      "step": 10382
+    },
+    {
+      "epoch": 3.115237452171956,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.4359090224431394e-05,
+      "loss": 0.1018,
+      "step": 10383
+    },
+    {
+      "epoch": 3.1155375497036535,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.434402771501732e-05,
+      "loss": 0.1006,
+      "step": 10384
+    },
+    {
+      "epoch": 3.1158376472353515,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.4328965205603254e-05,
+      "loss": 0.1166,
+      "step": 10385
+    },
+    {
+      "epoch": 3.116137744767049,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.431390269618919e-05,
+      "loss": 0.1071,
+      "step": 10386
+    },
+    {
+      "epoch": 3.116437842298747,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.429884018677512e-05,
+      "loss": 0.1198,
+      "step": 10387
+    },
+    {
+      "epoch": 3.116737939830445,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.428377767736105e-05,
+      "loss": 0.1194,
+      "step": 10388
+    },
+    {
+      "epoch": 3.1170380373621427,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.426871516794698e-05,
+      "loss": 0.1099,
+      "step": 10389
+    },
+    {
+      "epoch": 3.1173381348938407,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.4253652658532915e-05,
+      "loss": 0.0939,
+      "step": 10390
+    },
+    {
+      "epoch": 3.1176382324255383,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.423859014911885e-05,
+      "loss": 0.1237,
+      "step": 10391
+    },
+    {
+      "epoch": 3.1179383299572363,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.4223527639704776e-05,
+      "loss": 0.1057,
+      "step": 10392
+    },
+    {
+      "epoch": 3.118238427488934,
+      "grad_norm": 0.75,
+      "learning_rate": 4.420846513029071e-05,
+      "loss": 0.1001,
+      "step": 10393
+    },
+    {
+      "epoch": 3.118538525020632,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.419340262087664e-05,
+      "loss": 0.0851,
+      "step": 10394
+    },
+    {
+      "epoch": 3.1188386225523295,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.4178340111462576e-05,
+      "loss": 0.1179,
+      "step": 10395
+    },
+    {
+      "epoch": 3.1191387200840275,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.41632776020485e-05,
+      "loss": 0.091,
+      "step": 10396
+    },
+    {
+      "epoch": 3.119438817615725,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.4148215092634437e-05,
+      "loss": 0.1023,
+      "step": 10397
+    },
+    {
+      "epoch": 3.119738915147423,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.413315258322036e-05,
+      "loss": 0.1426,
+      "step": 10398
+    },
+    {
+      "epoch": 3.1200390126791206,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.4118090073806304e-05,
+      "loss": 0.114,
+      "step": 10399
+    },
+    {
+      "epoch": 3.1203391102108187,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.410302756439223e-05,
+      "loss": 0.1565,
+      "step": 10400
+    },
+    {
+      "epoch": 3.1206392077425162,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.4087965054978164e-05,
+      "loss": 0.0988,
+      "step": 10401
+    },
+    {
+      "epoch": 3.1209393052742143,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.407290254556409e-05,
+      "loss": 0.1015,
+      "step": 10402
+    },
+    {
+      "epoch": 3.121239402805912,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.4057840036150024e-05,
+      "loss": 0.1179,
+      "step": 10403
+    },
+    {
+      "epoch": 3.12153950033761,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.404277752673596e-05,
+      "loss": 0.1033,
+      "step": 10404
+    },
+    {
+      "epoch": 3.1218395978693074,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.4027715017321884e-05,
+      "loss": 0.0968,
+      "step": 10405
+    },
+    {
+      "epoch": 3.1221396954010054,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.401265250790782e-05,
+      "loss": 0.1309,
+      "step": 10406
+    },
+    {
+      "epoch": 3.122439792932703,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.399758999849375e-05,
+      "loss": 0.0991,
+      "step": 10407
+    },
+    {
+      "epoch": 3.122739890464401,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.3982527489079685e-05,
+      "loss": 0.1133,
+      "step": 10408
+    },
+    {
+      "epoch": 3.1230399879960986,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.396746497966561e-05,
+      "loss": 0.1076,
+      "step": 10409
+    },
+    {
+      "epoch": 3.1233400855277966,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.3952402470251545e-05,
+      "loss": 0.1038,
+      "step": 10410
+    },
+    {
+      "epoch": 3.123640183059494,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.393733996083748e-05,
+      "loss": 0.1227,
+      "step": 10411
+    },
+    {
+      "epoch": 3.123940280591192,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.392227745142341e-05,
+      "loss": 0.1169,
+      "step": 10412
+    },
+    {
+      "epoch": 3.12424037812289,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.390721494200934e-05,
+      "loss": 0.1045,
+      "step": 10413
+    },
+    {
+      "epoch": 3.124540475654588,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.389215243259527e-05,
+      "loss": 0.5133,
+      "step": 10414
+    },
+    {
+      "epoch": 3.1248405731862854,
+      "grad_norm": 0.75,
+      "learning_rate": 4.38770899231812e-05,
+      "loss": 0.1101,
+      "step": 10415
+    },
+    {
+      "epoch": 3.1251406707179834,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.386202741376714e-05,
+      "loss": 0.0968,
+      "step": 10416
+    },
+    {
+      "epoch": 3.125440768249681,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.3846964904353066e-05,
+      "loss": 0.1019,
+      "step": 10417
+    },
+    {
+      "epoch": 3.125740865781379,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.3831902394939e-05,
+      "loss": 0.1082,
+      "step": 10418
+    },
+    {
+      "epoch": 3.1260409633130766,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.381683988552493e-05,
+      "loss": 0.1273,
+      "step": 10419
+    },
+    {
+      "epoch": 3.1263410608447746,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.380177737611087e-05,
+      "loss": 0.1395,
+      "step": 10420
+    },
+    {
+      "epoch": 3.126641158376472,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.3786714866696794e-05,
+      "loss": 0.1408,
+      "step": 10421
+    },
+    {
+      "epoch": 3.12694125590817,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.377165235728273e-05,
+      "loss": 0.0987,
+      "step": 10422
+    },
+    {
+      "epoch": 3.1272413534398678,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.3756589847868654e-05,
+      "loss": 0.1209,
+      "step": 10423
+    },
+    {
+      "epoch": 3.1275414509715658,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.374152733845459e-05,
+      "loss": 0.1016,
+      "step": 10424
+    },
+    {
+      "epoch": 3.1278415485032633,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.372646482904052e-05,
+      "loss": 0.1075,
+      "step": 10425
+    },
+    {
+      "epoch": 3.1281416460349614,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.3711402319626455e-05,
+      "loss": 0.1406,
+      "step": 10426
+    },
+    {
+      "epoch": 3.1284417435666594,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.369633981021238e-05,
+      "loss": 0.1296,
+      "step": 10427
+    },
+    {
+      "epoch": 3.128741841098357,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.3681277300798315e-05,
+      "loss": 0.0907,
+      "step": 10428
+    },
+    {
+      "epoch": 3.129041938630055,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.366621479138425e-05,
+      "loss": 0.1151,
+      "step": 10429
+    },
+    {
+      "epoch": 3.1293420361617525,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.3651152281970175e-05,
+      "loss": 0.1216,
+      "step": 10430
+    },
+    {
+      "epoch": 3.1296421336934506,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.363608977255611e-05,
+      "loss": 0.1132,
+      "step": 10431
+    },
+    {
+      "epoch": 3.129942231225148,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.362102726314204e-05,
+      "loss": 0.1236,
+      "step": 10432
+    },
+    {
+      "epoch": 3.130242328756846,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.3605964753727976e-05,
+      "loss": 0.1045,
+      "step": 10433
+    },
+    {
+      "epoch": 3.1305424262885437,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.35909022443139e-05,
+      "loss": 0.106,
+      "step": 10434
+    },
+    {
+      "epoch": 3.1308425238202418,
+      "grad_norm": 1.0,
+      "learning_rate": 4.3575839734899836e-05,
+      "loss": 0.1855,
+      "step": 10435
+    },
+    {
+      "epoch": 3.1311426213519393,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.356077722548576e-05,
+      "loss": 0.1007,
+      "step": 10436
+    },
+    {
+      "epoch": 3.1314427188836373,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.35457147160717e-05,
+      "loss": 0.1013,
+      "step": 10437
+    },
+    {
+      "epoch": 3.131742816415335,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.353065220665763e-05,
+      "loss": 0.1121,
+      "step": 10438
+    },
+    {
+      "epoch": 3.132042913947033,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.3515589697243563e-05,
+      "loss": 0.1156,
+      "step": 10439
+    },
+    {
+      "epoch": 3.1323430114787305,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.350052718782949e-05,
+      "loss": 0.122,
+      "step": 10440
+    },
+    {
+      "epoch": 3.1326431090104285,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.348546467841543e-05,
+      "loss": 0.1344,
+      "step": 10441
+    },
+    {
+      "epoch": 3.132943206542126,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.347040216900136e-05,
+      "loss": 0.1181,
+      "step": 10442
+    },
+    {
+      "epoch": 3.133243304073824,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.345533965958729e-05,
+      "loss": 0.1182,
+      "step": 10443
+    },
+    {
+      "epoch": 3.1335434016055217,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.344027715017322e-05,
+      "loss": 0.1038,
+      "step": 10444
+    },
+    {
+      "epoch": 3.1338434991372197,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.342521464075915e-05,
+      "loss": 0.1237,
+      "step": 10445
+    },
+    {
+      "epoch": 3.1341435966689173,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.3410152131345085e-05,
+      "loss": 0.1025,
+      "step": 10446
+    },
+    {
+      "epoch": 3.1344436942006153,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.339508962193102e-05,
+      "loss": 0.1309,
+      "step": 10447
+    },
+    {
+      "epoch": 3.134743791732313,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.3380027112516945e-05,
+      "loss": 0.0965,
+      "step": 10448
+    },
+    {
+      "epoch": 3.135043889264011,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.336496460310288e-05,
+      "loss": 0.0998,
+      "step": 10449
+    },
+    {
+      "epoch": 3.1353439867957085,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.334990209368881e-05,
+      "loss": 0.1029,
+      "step": 10450
+    },
+    {
+      "epoch": 3.1356440843274065,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.3334839584274745e-05,
+      "loss": 0.1116,
+      "step": 10451
+    },
+    {
+      "epoch": 3.135944181859104,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.331977707486067e-05,
+      "loss": 0.1195,
+      "step": 10452
+    },
+    {
+      "epoch": 3.136244279390802,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.3304714565446606e-05,
+      "loss": 0.12,
+      "step": 10453
+    },
+    {
+      "epoch": 3.1365443769224997,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.328965205603254e-05,
+      "loss": 0.0977,
+      "step": 10454
+    },
+    {
+      "epoch": 3.1368444744541977,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.327458954661847e-05,
+      "loss": 0.0894,
+      "step": 10455
+    },
+    {
+      "epoch": 3.1371445719858952,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.32595270372044e-05,
+      "loss": 0.0949,
+      "step": 10456
+    },
+    {
+      "epoch": 3.1374446695175933,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.3244464527790326e-05,
+      "loss": 0.1105,
+      "step": 10457
+    },
+    {
+      "epoch": 3.137744767049291,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.3229402018376267e-05,
+      "loss": 0.0871,
+      "step": 10458
+    },
+    {
+      "epoch": 3.138044864580989,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.321433950896219e-05,
+      "loss": 0.1136,
+      "step": 10459
+    },
+    {
+      "epoch": 3.1383449621126864,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.319927699954813e-05,
+      "loss": 0.0944,
+      "step": 10460
+    },
+    {
+      "epoch": 3.1386450596443844,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.3184214490134054e-05,
+      "loss": 0.1051,
+      "step": 10461
+    },
+    {
+      "epoch": 3.138945157176082,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.3169151980719994e-05,
+      "loss": 0.1378,
+      "step": 10462
+    },
+    {
+      "epoch": 3.13924525470778,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.315408947130592e-05,
+      "loss": 0.1163,
+      "step": 10463
+    },
+    {
+      "epoch": 3.1395453522394776,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.3139026961891854e-05,
+      "loss": 0.1265,
+      "step": 10464
+    },
+    {
+      "epoch": 3.1398454497711756,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.312396445247778e-05,
+      "loss": 0.0897,
+      "step": 10465
+    },
+    {
+      "epoch": 3.1401455473028737,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.3108901943063715e-05,
+      "loss": 0.1177,
+      "step": 10466
+    },
+    {
+      "epoch": 3.1404456448345712,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.309383943364965e-05,
+      "loss": 0.1287,
+      "step": 10467
+    },
+    {
+      "epoch": 3.1407457423662692,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.307877692423558e-05,
+      "loss": 0.0842,
+      "step": 10468
+    },
+    {
+      "epoch": 3.141045839897967,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.306371441482151e-05,
+      "loss": 0.1091,
+      "step": 10469
+    },
+    {
+      "epoch": 3.141345937429665,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.304865190540744e-05,
+      "loss": 0.1007,
+      "step": 10470
+    },
+    {
+      "epoch": 3.1416460349613624,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.3033589395993375e-05,
+      "loss": 0.1097,
+      "step": 10471
+    },
+    {
+      "epoch": 3.1419461324930604,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.301852688657931e-05,
+      "loss": 0.0947,
+      "step": 10472
+    },
+    {
+      "epoch": 3.142246230024758,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.3003464377165236e-05,
+      "loss": 0.1067,
+      "step": 10473
+    },
+    {
+      "epoch": 3.142546327556456,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.298840186775117e-05,
+      "loss": 0.0992,
+      "step": 10474
+    },
+    {
+      "epoch": 3.1428464250881536,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.29733393583371e-05,
+      "loss": 0.0993,
+      "step": 10475
+    },
+    {
+      "epoch": 3.1431465226198516,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.2958276848923036e-05,
+      "loss": 0.1395,
+      "step": 10476
+    },
+    {
+      "epoch": 3.143446620151549,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.294321433950896e-05,
+      "loss": 0.1102,
+      "step": 10477
+    },
+    {
+      "epoch": 3.143746717683247,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.2928151830094897e-05,
+      "loss": 0.1472,
+      "step": 10478
+    },
+    {
+      "epoch": 3.144046815214945,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.291308932068083e-05,
+      "loss": 0.1072,
+      "step": 10479
+    },
+    {
+      "epoch": 3.144346912746643,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.2898026811266764e-05,
+      "loss": 0.0901,
+      "step": 10480
+    },
+    {
+      "epoch": 3.1446470102783404,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.288296430185269e-05,
+      "loss": 0.1243,
+      "step": 10481
+    },
+    {
+      "epoch": 3.1449471078100384,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.2867901792438624e-05,
+      "loss": 0.1366,
+      "step": 10482
+    },
+    {
+      "epoch": 3.145247205341736,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.285283928302455e-05,
+      "loss": 0.1165,
+      "step": 10483
+    },
+    {
+      "epoch": 3.145547302873434,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.2837776773610484e-05,
+      "loss": 0.0921,
+      "step": 10484
+    },
+    {
+      "epoch": 3.1458474004051316,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.282271426419642e-05,
+      "loss": 0.1075,
+      "step": 10485
+    },
+    {
+      "epoch": 3.1461474979368296,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.2807651754782344e-05,
+      "loss": 0.1471,
+      "step": 10486
+    },
+    {
+      "epoch": 3.146447595468527,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.279258924536828e-05,
+      "loss": 0.1024,
+      "step": 10487
+    },
+    {
+      "epoch": 3.146747693000225,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.277752673595421e-05,
+      "loss": 0.1456,
+      "step": 10488
+    },
+    {
+      "epoch": 3.1470477905319227,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.2762464226540145e-05,
+      "loss": 0.1063,
+      "step": 10489
+    },
+    {
+      "epoch": 3.1473478880636208,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.274740171712607e-05,
+      "loss": 0.0922,
+      "step": 10490
+    },
+    {
+      "epoch": 3.1476479855953183,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.2732339207712005e-05,
+      "loss": 0.1073,
+      "step": 10491
+    },
+    {
+      "epoch": 3.1479480831270163,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.271727669829794e-05,
+      "loss": 0.1175,
+      "step": 10492
+    },
+    {
+      "epoch": 3.148248180658714,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.270221418888387e-05,
+      "loss": 0.1149,
+      "step": 10493
+    },
+    {
+      "epoch": 3.148548278190412,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.26871516794698e-05,
+      "loss": 0.1122,
+      "step": 10494
+    },
+    {
+      "epoch": 3.1488483757221095,
+      "grad_norm": 0.75,
+      "learning_rate": 4.267208917005573e-05,
+      "loss": 0.101,
+      "step": 10495
+    },
+    {
+      "epoch": 3.1491484732538075,
+      "grad_norm": 0.75,
+      "learning_rate": 4.2657026660641666e-05,
+      "loss": 0.1066,
+      "step": 10496
+    },
+    {
+      "epoch": 3.149448570785505,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.26419641512276e-05,
+      "loss": 0.0922,
+      "step": 10497
+    },
+    {
+      "epoch": 3.149748668317203,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.2626901641813526e-05,
+      "loss": 0.1094,
+      "step": 10498
+    },
+    {
+      "epoch": 3.1500487658489007,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.261183913239946e-05,
+      "loss": 0.1294,
+      "step": 10499
+    },
+    {
+      "epoch": 3.1503488633805987,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.2596776622985394e-05,
+      "loss": 0.107,
+      "step": 10500
+    },
+    {
+      "epoch": 3.1506489609122963,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.258171411357133e-05,
+      "loss": 0.1376,
+      "step": 10501
+    },
+    {
+      "epoch": 3.1509490584439943,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.2566651604157254e-05,
+      "loss": 0.1221,
+      "step": 10502
+    },
+    {
+      "epoch": 3.151249155975692,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.255158909474319e-05,
+      "loss": 0.0959,
+      "step": 10503
+    },
+    {
+      "epoch": 3.15154925350739,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.2536526585329114e-05,
+      "loss": 0.1036,
+      "step": 10504
+    },
+    {
+      "epoch": 3.151849351039088,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.2521464075915054e-05,
+      "loss": 0.094,
+      "step": 10505
+    },
+    {
+      "epoch": 3.1521494485707855,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.250640156650098e-05,
+      "loss": 0.1168,
+      "step": 10506
+    },
+    {
+      "epoch": 3.1524495461024835,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.2491339057086915e-05,
+      "loss": 0.1044,
+      "step": 10507
+    },
+    {
+      "epoch": 3.152749643634181,
+      "grad_norm": 0.875,
+      "learning_rate": 4.247627654767284e-05,
+      "loss": 0.1151,
+      "step": 10508
+    },
+    {
+      "epoch": 3.153049741165879,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.246121403825878e-05,
+      "loss": 0.1332,
+      "step": 10509
+    },
+    {
+      "epoch": 3.1533498386975767,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.244615152884471e-05,
+      "loss": 0.1045,
+      "step": 10510
+    },
+    {
+      "epoch": 3.1536499362292747,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.2431089019430635e-05,
+      "loss": 0.0937,
+      "step": 10511
+    },
+    {
+      "epoch": 3.1539500337609723,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.241602651001657e-05,
+      "loss": 0.1102,
+      "step": 10512
+    },
+    {
+      "epoch": 3.1542501312926703,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.24009640006025e-05,
+      "loss": 0.1027,
+      "step": 10513
+    },
+    {
+      "epoch": 3.154550228824368,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.2385901491188436e-05,
+      "loss": 0.1527,
+      "step": 10514
+    },
+    {
+      "epoch": 3.154850326356066,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.237083898177436e-05,
+      "loss": 0.106,
+      "step": 10515
+    },
+    {
+      "epoch": 3.1551504238877635,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.2355776472360296e-05,
+      "loss": 0.1143,
+      "step": 10516
+    },
+    {
+      "epoch": 3.1554505214194615,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.234071396294623e-05,
+      "loss": 0.1183,
+      "step": 10517
+    },
+    {
+      "epoch": 3.155750618951159,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.232565145353216e-05,
+      "loss": 0.1037,
+      "step": 10518
+    },
+    {
+      "epoch": 3.156050716482857,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.231058894411809e-05,
+      "loss": 0.0962,
+      "step": 10519
+    },
+    {
+      "epoch": 3.1563508140145546,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.2295526434704023e-05,
+      "loss": 0.1004,
+      "step": 10520
+    },
+    {
+      "epoch": 3.1566509115462527,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.228046392528996e-05,
+      "loss": 0.1336,
+      "step": 10521
+    },
+    {
+      "epoch": 3.1569510090779502,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.226540141587589e-05,
+      "loss": 0.1091,
+      "step": 10522
+    },
+    {
+      "epoch": 3.1572511066096483,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.225033890646182e-05,
+      "loss": 0.1258,
+      "step": 10523
+    },
+    {
+      "epoch": 3.157551204141346,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.223527639704775e-05,
+      "loss": 0.0963,
+      "step": 10524
+    },
+    {
+      "epoch": 3.157851301673044,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.222021388763368e-05,
+      "loss": 0.0883,
+      "step": 10525
+    },
+    {
+      "epoch": 3.1581513992047414,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.220515137821962e-05,
+      "loss": 0.1305,
+      "step": 10526
+    },
+    {
+      "epoch": 3.1584514967364394,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.2190088868805545e-05,
+      "loss": 0.1577,
+      "step": 10527
+    },
+    {
+      "epoch": 3.158751594268137,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.217502635939148e-05,
+      "loss": 0.1143,
+      "step": 10528
+    },
+    {
+      "epoch": 3.159051691799835,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.2159963849977405e-05,
+      "loss": 0.0869,
+      "step": 10529
+    },
+    {
+      "epoch": 3.1593517893315326,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.2144901340563345e-05,
+      "loss": 0.1338,
+      "step": 10530
+    },
+    {
+      "epoch": 3.1596518868632306,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.212983883114927e-05,
+      "loss": 0.1672,
+      "step": 10531
+    },
+    {
+      "epoch": 3.159951984394928,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.2114776321735205e-05,
+      "loss": 0.0935,
+      "step": 10532
+    },
+    {
+      "epoch": 3.160252081926626,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.209971381232113e-05,
+      "loss": 0.1152,
+      "step": 10533
+    },
+    {
+      "epoch": 3.160552179458324,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.2084651302907066e-05,
+      "loss": 0.0968,
+      "step": 10534
+    },
+    {
+      "epoch": 3.160852276990022,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.2069588793493e-05,
+      "loss": 0.1126,
+      "step": 10535
+    },
+    {
+      "epoch": 3.1611523745217194,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.205452628407893e-05,
+      "loss": 0.1205,
+      "step": 10536
+    },
+    {
+      "epoch": 3.1614524720534174,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.203946377466486e-05,
+      "loss": 0.1428,
+      "step": 10537
+    },
+    {
+      "epoch": 3.161752569585115,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.202440126525079e-05,
+      "loss": 0.1267,
+      "step": 10538
+    },
+    {
+      "epoch": 3.162052667116813,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.200933875583673e-05,
+      "loss": 0.1078,
+      "step": 10539
+    },
+    {
+      "epoch": 3.1623527646485106,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.199427624642265e-05,
+      "loss": 0.2752,
+      "step": 10540
+    },
+    {
+      "epoch": 3.1626528621802086,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.197921373700859e-05,
+      "loss": 0.1239,
+      "step": 10541
+    },
+    {
+      "epoch": 3.162952959711906,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.1964151227594514e-05,
+      "loss": 0.1526,
+      "step": 10542
+    },
+    {
+      "epoch": 3.163253057243604,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.1949088718180454e-05,
+      "loss": 0.1147,
+      "step": 10543
+    },
+    {
+      "epoch": 3.163553154775302,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.193402620876638e-05,
+      "loss": 0.1101,
+      "step": 10544
+    },
+    {
+      "epoch": 3.1638532523069998,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.1918963699352314e-05,
+      "loss": 0.125,
+      "step": 10545
+    },
+    {
+      "epoch": 3.164153349838698,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.190390118993824e-05,
+      "loss": 0.1063,
+      "step": 10546
+    },
+    {
+      "epoch": 3.1644534473703954,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.188883868052418e-05,
+      "loss": 0.1167,
+      "step": 10547
+    },
+    {
+      "epoch": 3.1647535449020934,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.187377617111011e-05,
+      "loss": 0.1014,
+      "step": 10548
+    },
+    {
+      "epoch": 3.165053642433791,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.185871366169604e-05,
+      "loss": 0.1293,
+      "step": 10549
+    },
+    {
+      "epoch": 3.165353739965489,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.184365115228197e-05,
+      "loss": 0.1336,
+      "step": 10550
+    },
+    {
+      "epoch": 3.1656538374971865,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.18285886428679e-05,
+      "loss": 0.1178,
+      "step": 10551
+    },
+    {
+      "epoch": 3.1659539350288846,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.1813526133453835e-05,
+      "loss": 0.1228,
+      "step": 10552
+    },
+    {
+      "epoch": 3.166254032560582,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.179846362403977e-05,
+      "loss": 0.085,
+      "step": 10553
+    },
+    {
+      "epoch": 3.16655413009228,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.1783401114625696e-05,
+      "loss": 0.1137,
+      "step": 10554
+    },
+    {
+      "epoch": 3.1668542276239777,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.176833860521163e-05,
+      "loss": 0.1122,
+      "step": 10555
+    },
+    {
+      "epoch": 3.1671543251556757,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.175327609579756e-05,
+      "loss": 0.0972,
+      "step": 10556
+    },
+    {
+      "epoch": 3.1674544226873733,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.1738213586383496e-05,
+      "loss": 0.1147,
+      "step": 10557
+    },
+    {
+      "epoch": 3.1677545202190713,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.172315107696942e-05,
+      "loss": 0.1043,
+      "step": 10558
+    },
+    {
+      "epoch": 3.168054617750769,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.1708088567555357e-05,
+      "loss": 0.1207,
+      "step": 10559
+    },
+    {
+      "epoch": 3.168354715282467,
+      "grad_norm": 0.75,
+      "learning_rate": 4.169302605814129e-05,
+      "loss": 0.128,
+      "step": 10560
+    },
+    {
+      "epoch": 3.1686548128141645,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.1677963548727224e-05,
+      "loss": 0.0995,
+      "step": 10561
+    },
+    {
+      "epoch": 3.1689549103458625,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.166290103931315e-05,
+      "loss": 0.1259,
+      "step": 10562
+    },
+    {
+      "epoch": 3.16925500787756,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.1647838529899084e-05,
+      "loss": 0.0989,
+      "step": 10563
+    },
+    {
+      "epoch": 3.169555105409258,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.163277602048502e-05,
+      "loss": 0.1081,
+      "step": 10564
+    },
+    {
+      "epoch": 3.1698552029409557,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.1617713511070944e-05,
+      "loss": 0.094,
+      "step": 10565
+    },
+    {
+      "epoch": 3.1701553004726537,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.160265100165688e-05,
+      "loss": 0.1063,
+      "step": 10566
+    },
+    {
+      "epoch": 3.1704553980043513,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.1587588492242804e-05,
+      "loss": 0.1017,
+      "step": 10567
+    },
+    {
+      "epoch": 3.1707554955360493,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.1572525982828745e-05,
+      "loss": 0.1488,
+      "step": 10568
+    },
+    {
+      "epoch": 3.171055593067747,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.155746347341467e-05,
+      "loss": 0.1233,
+      "step": 10569
+    },
+    {
+      "epoch": 3.171355690599445,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.1542400964000605e-05,
+      "loss": 0.1427,
+      "step": 10570
+    },
+    {
+      "epoch": 3.1716557881311425,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.152733845458653e-05,
+      "loss": 0.0957,
+      "step": 10571
+    },
+    {
+      "epoch": 3.1719558856628405,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.1512275945172465e-05,
+      "loss": 0.1177,
+      "step": 10572
+    },
+    {
+      "epoch": 3.172255983194538,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.14972134357584e-05,
+      "loss": 0.1262,
+      "step": 10573
+    },
+    {
+      "epoch": 3.172556080726236,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.148215092634433e-05,
+      "loss": 0.0995,
+      "step": 10574
+    },
+    {
+      "epoch": 3.1728561782579336,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.146708841693026e-05,
+      "loss": 0.1303,
+      "step": 10575
+    },
+    {
+      "epoch": 3.1731562757896317,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.145202590751619e-05,
+      "loss": 0.121,
+      "step": 10576
+    },
+    {
+      "epoch": 3.1734563733213292,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.1436963398102126e-05,
+      "loss": 0.1298,
+      "step": 10577
+    },
+    {
+      "epoch": 3.1737564708530273,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.142190088868806e-05,
+      "loss": 0.0971,
+      "step": 10578
+    },
+    {
+      "epoch": 3.174056568384725,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.1406838379273986e-05,
+      "loss": 0.1139,
+      "step": 10579
+    },
+    {
+      "epoch": 3.174356665916423,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.139177586985992e-05,
+      "loss": 0.1091,
+      "step": 10580
+    },
+    {
+      "epoch": 3.1746567634481204,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.1376713360445854e-05,
+      "loss": 0.0852,
+      "step": 10581
+    },
+    {
+      "epoch": 3.1749568609798184,
+      "grad_norm": 1.0,
+      "learning_rate": 4.136165085103179e-05,
+      "loss": 0.1617,
+      "step": 10582
+    },
+    {
+      "epoch": 3.1752569585115165,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.1346588341617714e-05,
+      "loss": 0.0937,
+      "step": 10583
+    },
+    {
+      "epoch": 3.175557056043214,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.133152583220365e-05,
+      "loss": 0.131,
+      "step": 10584
+    },
+    {
+      "epoch": 3.175857153574912,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.131646332278958e-05,
+      "loss": 0.1037,
+      "step": 10585
+    },
+    {
+      "epoch": 3.1761572511066096,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.1301400813375514e-05,
+      "loss": 0.1204,
+      "step": 10586
+    },
+    {
+      "epoch": 3.1764573486383076,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.128633830396144e-05,
+      "loss": 0.1094,
+      "step": 10587
+    },
+    {
+      "epoch": 3.176757446170005,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.1271275794547375e-05,
+      "loss": 0.1077,
+      "step": 10588
+    },
+    {
+      "epoch": 3.1770575437017032,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.125621328513331e-05,
+      "loss": 0.1055,
+      "step": 10589
+    },
+    {
+      "epoch": 3.177357641233401,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.124115077571924e-05,
+      "loss": 0.1145,
+      "step": 10590
+    },
+    {
+      "epoch": 3.177657738765099,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.122608826630517e-05,
+      "loss": 0.1035,
+      "step": 10591
+    },
+    {
+      "epoch": 3.1779578362967964,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.1211025756891095e-05,
+      "loss": 0.1146,
+      "step": 10592
+    },
+    {
+      "epoch": 3.1782579338284944,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.119596324747703e-05,
+      "loss": 0.1332,
+      "step": 10593
+    },
+    {
+      "epoch": 3.178558031360192,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.118090073806296e-05,
+      "loss": 0.1099,
+      "step": 10594
+    },
+    {
+      "epoch": 3.17885812889189,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.1165838228648896e-05,
+      "loss": 0.1324,
+      "step": 10595
+    },
+    {
+      "epoch": 3.1791582264235876,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.115077571923482e-05,
+      "loss": 0.0992,
+      "step": 10596
+    },
+    {
+      "epoch": 3.1794583239552856,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.1135713209820756e-05,
+      "loss": 0.1203,
+      "step": 10597
+    },
+    {
+      "epoch": 3.179758421486983,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.112065070040669e-05,
+      "loss": 0.1106,
+      "step": 10598
+    },
+    {
+      "epoch": 3.180058519018681,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.110558819099262e-05,
+      "loss": 0.1284,
+      "step": 10599
+    },
+    {
+      "epoch": 3.1803586165503788,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.109052568157855e-05,
+      "loss": 0.1116,
+      "step": 10600
+    },
+    {
+      "epoch": 3.180658714082077,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.1075463172164483e-05,
+      "loss": 0.1192,
+      "step": 10601
+    },
+    {
+      "epoch": 3.1809588116137744,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.106040066275042e-05,
+      "loss": 0.1336,
+      "step": 10602
+    },
+    {
+      "epoch": 3.1812589091454724,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.104533815333635e-05,
+      "loss": 0.1139,
+      "step": 10603
+    },
+    {
+      "epoch": 3.18155900667717,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.103027564392228e-05,
+      "loss": 0.1227,
+      "step": 10604
+    },
+    {
+      "epoch": 3.181859104208868,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.101521313450821e-05,
+      "loss": 0.1239,
+      "step": 10605
+    },
+    {
+      "epoch": 3.1821592017405655,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.1000150625094144e-05,
+      "loss": 0.0957,
+      "step": 10606
+    },
+    {
+      "epoch": 3.1824592992722636,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.098508811568008e-05,
+      "loss": 0.1045,
+      "step": 10607
+    },
+    {
+      "epoch": 3.182759396803961,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.0970025606266005e-05,
+      "loss": 0.1075,
+      "step": 10608
+    },
+    {
+      "epoch": 3.183059494335659,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.095496309685194e-05,
+      "loss": 0.1249,
+      "step": 10609
+    },
+    {
+      "epoch": 3.1833595918673567,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.0939900587437865e-05,
+      "loss": 0.1014,
+      "step": 10610
+    },
+    {
+      "epoch": 3.1836596893990547,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.0924838078023805e-05,
+      "loss": 0.1277,
+      "step": 10611
+    },
+    {
+      "epoch": 3.1839597869307523,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.090977556860973e-05,
+      "loss": 0.1192,
+      "step": 10612
+    },
+    {
+      "epoch": 3.1842598844624503,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.0894713059195665e-05,
+      "loss": 0.1288,
+      "step": 10613
+    },
+    {
+      "epoch": 3.184559981994148,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.087965054978159e-05,
+      "loss": 0.0818,
+      "step": 10614
+    },
+    {
+      "epoch": 3.184860079525846,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.086458804036753e-05,
+      "loss": 0.128,
+      "step": 10615
+    },
+    {
+      "epoch": 3.1851601770575435,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.084952553095346e-05,
+      "loss": 0.112,
+      "step": 10616
+    },
+    {
+      "epoch": 3.1854602745892415,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.083446302153939e-05,
+      "loss": 0.1065,
+      "step": 10617
+    },
+    {
+      "epoch": 3.185760372120939,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.081940051212532e-05,
+      "loss": 0.0796,
+      "step": 10618
+    },
+    {
+      "epoch": 3.186060469652637,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.080433800271125e-05,
+      "loss": 0.1451,
+      "step": 10619
+    },
+    {
+      "epoch": 3.1863605671843347,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.078927549329719e-05,
+      "loss": 0.1031,
+      "step": 10620
+    },
+    {
+      "epoch": 3.1866606647160327,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.077421298388311e-05,
+      "loss": 0.1082,
+      "step": 10621
+    },
+    {
+      "epoch": 3.1869607622477307,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.075915047446905e-05,
+      "loss": 0.1107,
+      "step": 10622
+    },
+    {
+      "epoch": 3.1872608597794283,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.074408796505498e-05,
+      "loss": 0.1059,
+      "step": 10623
+    },
+    {
+      "epoch": 3.1875609573111263,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.0729025455640914e-05,
+      "loss": 0.0991,
+      "step": 10624
+    },
+    {
+      "epoch": 3.187861054842824,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.071396294622684e-05,
+      "loss": 0.0861,
+      "step": 10625
+    },
+    {
+      "epoch": 3.188161152374522,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.0698900436812774e-05,
+      "loss": 0.0972,
+      "step": 10626
+    },
+    {
+      "epoch": 3.1884612499062195,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.068383792739871e-05,
+      "loss": 0.1246,
+      "step": 10627
+    },
+    {
+      "epoch": 3.1887613474379175,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.066877541798464e-05,
+      "loss": 0.1091,
+      "step": 10628
+    },
+    {
+      "epoch": 3.189061444969615,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.065371290857057e-05,
+      "loss": 0.0911,
+      "step": 10629
+    },
+    {
+      "epoch": 3.189361542501313,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.06386503991565e-05,
+      "loss": 0.1139,
+      "step": 10630
+    },
+    {
+      "epoch": 3.1896616400330107,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.062358788974243e-05,
+      "loss": 0.0961,
+      "step": 10631
+    },
+    {
+      "epoch": 3.1899617375647087,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.060852538032837e-05,
+      "loss": 0.1076,
+      "step": 10632
+    },
+    {
+      "epoch": 3.1902618350964063,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.0593462870914295e-05,
+      "loss": 0.1116,
+      "step": 10633
+    },
+    {
+      "epoch": 3.1905619326281043,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.057840036150023e-05,
+      "loss": 0.1526,
+      "step": 10634
+    },
+    {
+      "epoch": 3.190862030159802,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.0563337852086156e-05,
+      "loss": 0.1315,
+      "step": 10635
+    },
+    {
+      "epoch": 3.1911621276915,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.0548275342672096e-05,
+      "loss": 0.1185,
+      "step": 10636
+    },
+    {
+      "epoch": 3.1914622252231974,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.053321283325802e-05,
+      "loss": 0.1149,
+      "step": 10637
+    },
+    {
+      "epoch": 3.1917623227548955,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.0518150323843956e-05,
+      "loss": 0.1007,
+      "step": 10638
+    },
+    {
+      "epoch": 3.192062420286593,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.050308781442988e-05,
+      "loss": 0.1275,
+      "step": 10639
+    },
+    {
+      "epoch": 3.192362517818291,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.0488025305015817e-05,
+      "loss": 0.126,
+      "step": 10640
+    },
+    {
+      "epoch": 3.1926626153499886,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.047296279560175e-05,
+      "loss": 0.0926,
+      "step": 10641
+    },
+    {
+      "epoch": 3.1929627128816866,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.0457900286187684e-05,
+      "loss": 0.0894,
+      "step": 10642
+    },
+    {
+      "epoch": 3.1932628104133842,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.044283777677361e-05,
+      "loss": 0.1137,
+      "step": 10643
+    },
+    {
+      "epoch": 3.1935629079450822,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.0427775267359544e-05,
+      "loss": 0.1182,
+      "step": 10644
+    },
+    {
+      "epoch": 3.19386300547678,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.041271275794548e-05,
+      "loss": 0.0832,
+      "step": 10645
+    },
+    {
+      "epoch": 3.194163103008478,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.0397650248531404e-05,
+      "loss": 0.0995,
+      "step": 10646
+    },
+    {
+      "epoch": 3.1944632005401754,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.038258773911734e-05,
+      "loss": 0.0917,
+      "step": 10647
+    },
+    {
+      "epoch": 3.1947632980718734,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.036752522970327e-05,
+      "loss": 0.1324,
+      "step": 10648
+    },
+    {
+      "epoch": 3.195063395603571,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.0352462720289205e-05,
+      "loss": 0.1155,
+      "step": 10649
+    },
+    {
+      "epoch": 3.195363493135269,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.033740021087513e-05,
+      "loss": 0.0939,
+      "step": 10650
+    },
+    {
+      "epoch": 3.1956635906669666,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.0322337701461065e-05,
+      "loss": 0.1298,
+      "step": 10651
+    },
+    {
+      "epoch": 3.1959636881986646,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.030727519204699e-05,
+      "loss": 0.1205,
+      "step": 10652
+    },
+    {
+      "epoch": 3.196263785730362,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.029221268263293e-05,
+      "loss": 0.1036,
+      "step": 10653
+    },
+    {
+      "epoch": 3.19656388326206,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.027715017321886e-05,
+      "loss": 0.1176,
+      "step": 10654
+    },
+    {
+      "epoch": 3.1968639807937578,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.026208766380479e-05,
+      "loss": 0.1353,
+      "step": 10655
+    },
+    {
+      "epoch": 3.197164078325456,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.024702515439072e-05,
+      "loss": 0.1139,
+      "step": 10656
+    },
+    {
+      "epoch": 3.1974641758571534,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.023196264497666e-05,
+      "loss": 0.1069,
+      "step": 10657
+    },
+    {
+      "epoch": 3.1977642733888514,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.0216900135562586e-05,
+      "loss": 0.105,
+      "step": 10658
+    },
+    {
+      "epoch": 3.198064370920549,
+      "grad_norm": 0.75,
+      "learning_rate": 4.020183762614852e-05,
+      "loss": 0.1014,
+      "step": 10659
+    },
+    {
+      "epoch": 3.198364468452247,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.0186775116734446e-05,
+      "loss": 0.0949,
+      "step": 10660
+    },
+    {
+      "epoch": 3.198664565983945,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.017171260732038e-05,
+      "loss": 0.1192,
+      "step": 10661
+    },
+    {
+      "epoch": 3.1989646635156426,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.0156650097906314e-05,
+      "loss": 0.1192,
+      "step": 10662
+    },
+    {
+      "epoch": 3.1992647610473406,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.014158758849225e-05,
+      "loss": 0.1139,
+      "step": 10663
+    },
+    {
+      "epoch": 3.199564858579038,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.0126525079078174e-05,
+      "loss": 0.1173,
+      "step": 10664
+    },
+    {
+      "epoch": 3.199864956110736,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.011146256966411e-05,
+      "loss": 0.089,
+      "step": 10665
+    },
+    {
+      "epoch": 3.2001650536424338,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.009640006025004e-05,
+      "loss": 0.111,
+      "step": 10666
+    },
+    {
+      "epoch": 3.2004651511741318,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.0081337550835974e-05,
+      "loss": 0.1268,
+      "step": 10667
+    },
+    {
+      "epoch": 3.2007652487058293,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.00662750414219e-05,
+      "loss": 0.0941,
+      "step": 10668
+    },
+    {
+      "epoch": 3.2010653462375274,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.0051212532007835e-05,
+      "loss": 0.0935,
+      "step": 10669
+    },
+    {
+      "epoch": 3.201365443769225,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.003615002259377e-05,
+      "loss": 0.1008,
+      "step": 10670
+    },
+    {
+      "epoch": 3.201665541300923,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.00210875131797e-05,
+      "loss": 0.1338,
+      "step": 10671
+    },
+    {
+      "epoch": 3.2019656388326205,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.000602500376563e-05,
+      "loss": 0.095,
+      "step": 10672
+    },
+    {
+      "epoch": 3.2022657363643185,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.9990962494351555e-05,
+      "loss": 0.084,
+      "step": 10673
+    },
+    {
+      "epoch": 3.202565833896016,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.9975899984937496e-05,
+      "loss": 0.1307,
+      "step": 10674
+    },
+    {
+      "epoch": 3.202865931427714,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.996083747552342e-05,
+      "loss": 0.1043,
+      "step": 10675
+    },
+    {
+      "epoch": 3.2031660289594117,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.9945774966109356e-05,
+      "loss": 0.1403,
+      "step": 10676
+    },
+    {
+      "epoch": 3.2034661264911097,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.993071245669528e-05,
+      "loss": 0.1217,
+      "step": 10677
+    },
+    {
+      "epoch": 3.2037662240228073,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.991564994728122e-05,
+      "loss": 0.107,
+      "step": 10678
+    },
+    {
+      "epoch": 3.2040663215545053,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.990058743786715e-05,
+      "loss": 0.1147,
+      "step": 10679
+    },
+    {
+      "epoch": 3.204366419086203,
+      "grad_norm": 0.875,
+      "learning_rate": 3.988552492845308e-05,
+      "loss": 0.1125,
+      "step": 10680
+    },
+    {
+      "epoch": 3.204666516617901,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.987046241903901e-05,
+      "loss": 0.0954,
+      "step": 10681
+    },
+    {
+      "epoch": 3.2049666141495985,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.9855399909624943e-05,
+      "loss": 0.1103,
+      "step": 10682
+    },
+    {
+      "epoch": 3.2052667116812965,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.984033740021088e-05,
+      "loss": 0.1146,
+      "step": 10683
+    },
+    {
+      "epoch": 3.205566809212994,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.982527489079681e-05,
+      "loss": 0.1074,
+      "step": 10684
+    },
+    {
+      "epoch": 3.205866906744692,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.981021238138274e-05,
+      "loss": 0.1059,
+      "step": 10685
+    },
+    {
+      "epoch": 3.2061670042763897,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.979514987196867e-05,
+      "loss": 0.1274,
+      "step": 10686
+    },
+    {
+      "epoch": 3.2064671018080877,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.9780087362554604e-05,
+      "loss": 0.1135,
+      "step": 10687
+    },
+    {
+      "epoch": 3.2067671993397853,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.976502485314054e-05,
+      "loss": 0.0974,
+      "step": 10688
+    },
+    {
+      "epoch": 3.2070672968714833,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.9749962343726465e-05,
+      "loss": 0.1258,
+      "step": 10689
+    },
+    {
+      "epoch": 3.207367394403181,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.97348998343124e-05,
+      "loss": 0.1294,
+      "step": 10690
+    },
+    {
+      "epoch": 3.207667491934879,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.971983732489833e-05,
+      "loss": 0.1268,
+      "step": 10691
+    },
+    {
+      "epoch": 3.2079675894665765,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.9704774815484265e-05,
+      "loss": 0.0822,
+      "step": 10692
+    },
+    {
+      "epoch": 3.2082676869982745,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.968971230607019e-05,
+      "loss": 0.0956,
+      "step": 10693
+    },
+    {
+      "epoch": 3.208567784529972,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.9674649796656125e-05,
+      "loss": 0.1165,
+      "step": 10694
+    },
+    {
+      "epoch": 3.20886788206167,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.965958728724206e-05,
+      "loss": 0.1107,
+      "step": 10695
+    },
+    {
+      "epoch": 3.2091679795933676,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.964452477782799e-05,
+      "loss": 0.0925,
+      "step": 10696
+    },
+    {
+      "epoch": 3.2094680771250657,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.962946226841392e-05,
+      "loss": 0.1065,
+      "step": 10697
+    },
+    {
+      "epoch": 3.2097681746567632,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.961439975899985e-05,
+      "loss": 0.1208,
+      "step": 10698
+    },
+    {
+      "epoch": 3.2100682721884612,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.959933724958578e-05,
+      "loss": 0.1133,
+      "step": 10699
+    },
+    {
+      "epoch": 3.2103683697201593,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.958427474017171e-05,
+      "loss": 0.1277,
+      "step": 10700
+    },
+    {
+      "epoch": 3.210668467251857,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.956921223075765e-05,
+      "loss": 0.0971,
+      "step": 10701
+    },
+    {
+      "epoch": 3.210968564783555,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.9554149721343573e-05,
+      "loss": 0.1002,
+      "step": 10702
+    },
+    {
+      "epoch": 3.2112686623152524,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.953908721192951e-05,
+      "loss": 0.0942,
+      "step": 10703
+    },
+    {
+      "epoch": 3.2115687598469504,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.952402470251544e-05,
+      "loss": 0.1066,
+      "step": 10704
+    },
+    {
+      "epoch": 3.211868857378648,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.9508962193101374e-05,
+      "loss": 0.0979,
+      "step": 10705
+    },
+    {
+      "epoch": 3.212168954910346,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.94938996836873e-05,
+      "loss": 0.1198,
+      "step": 10706
+    },
+    {
+      "epoch": 3.2124690524420436,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.9478837174273234e-05,
+      "loss": 0.1237,
+      "step": 10707
+    },
+    {
+      "epoch": 3.2127691499737416,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.946377466485917e-05,
+      "loss": 0.0922,
+      "step": 10708
+    },
+    {
+      "epoch": 3.213069247505439,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.94487121554451e-05,
+      "loss": 0.1042,
+      "step": 10709
+    },
+    {
+      "epoch": 3.2133693450371372,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.943364964603103e-05,
+      "loss": 0.1162,
+      "step": 10710
+    },
+    {
+      "epoch": 3.213669442568835,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.941858713661696e-05,
+      "loss": 0.127,
+      "step": 10711
+    },
+    {
+      "epoch": 3.213969540100533,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.9403524627202895e-05,
+      "loss": 0.1067,
+      "step": 10712
+    },
+    {
+      "epoch": 3.2142696376322304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.938846211778883e-05,
+      "loss": 0.0828,
+      "step": 10713
+    },
+    {
+      "epoch": 3.2145697351639284,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.9373399608374755e-05,
+      "loss": 0.1032,
+      "step": 10714
+    },
+    {
+      "epoch": 3.214869832695626,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.935833709896069e-05,
+      "loss": 0.1022,
+      "step": 10715
+    },
+    {
+      "epoch": 3.215169930227324,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.934327458954662e-05,
+      "loss": 0.109,
+      "step": 10716
+    },
+    {
+      "epoch": 3.2154700277590216,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.9328212080132556e-05,
+      "loss": 0.1177,
+      "step": 10717
+    },
+    {
+      "epoch": 3.2157701252907196,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.931314957071848e-05,
+      "loss": 0.1422,
+      "step": 10718
+    },
+    {
+      "epoch": 3.216070222822417,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.9298087061304416e-05,
+      "loss": 0.1143,
+      "step": 10719
+    },
+    {
+      "epoch": 3.216370320354115,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.928302455189034e-05,
+      "loss": 0.0985,
+      "step": 10720
+    },
+    {
+      "epoch": 3.2166704178858128,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.926796204247628e-05,
+      "loss": 0.1345,
+      "step": 10721
+    },
+    {
+      "epoch": 3.216970515417511,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.925289953306221e-05,
+      "loss": 0.1183,
+      "step": 10722
+    },
+    {
+      "epoch": 3.2172706129492084,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.9237837023648144e-05,
+      "loss": 0.1192,
+      "step": 10723
+    },
+    {
+      "epoch": 3.2175707104809064,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.922277451423407e-05,
+      "loss": 0.1064,
+      "step": 10724
+    },
+    {
+      "epoch": 3.217870808012604,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.920771200482001e-05,
+      "loss": 0.1807,
+      "step": 10725
+    },
+    {
+      "epoch": 3.218170905544302,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.919264949540594e-05,
+      "loss": 0.1024,
+      "step": 10726
+    },
+    {
+      "epoch": 3.2184710030759995,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.9177586985991864e-05,
+      "loss": 0.0967,
+      "step": 10727
+    },
+    {
+      "epoch": 3.2187711006076976,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.91625244765778e-05,
+      "loss": 0.1103,
+      "step": 10728
+    },
+    {
+      "epoch": 3.219071198139395,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.914746196716373e-05,
+      "loss": 0.1221,
+      "step": 10729
+    },
+    {
+      "epoch": 3.219371295671093,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.9132399457749665e-05,
+      "loss": 0.1337,
+      "step": 10730
+    },
+    {
+      "epoch": 3.2196713932027907,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.911733694833559e-05,
+      "loss": 0.0999,
+      "step": 10731
+    },
+    {
+      "epoch": 3.2199714907344887,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.9102274438921525e-05,
+      "loss": 0.1179,
+      "step": 10732
+    },
+    {
+      "epoch": 3.2202715882661863,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.908721192950746e-05,
+      "loss": 0.1177,
+      "step": 10733
+    },
+    {
+      "epoch": 3.2205716857978843,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.907214942009339e-05,
+      "loss": 0.1044,
+      "step": 10734
+    },
+    {
+      "epoch": 3.2208717833295823,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.905708691067932e-05,
+      "loss": 0.0875,
+      "step": 10735
+    },
+    {
+      "epoch": 3.22117188086128,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.904202440126525e-05,
+      "loss": 0.0909,
+      "step": 10736
+    },
+    {
+      "epoch": 3.2214719783929775,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.9026961891851186e-05,
+      "loss": 0.1132,
+      "step": 10737
+    },
+    {
+      "epoch": 3.2217720759246755,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.901189938243712e-05,
+      "loss": 0.1259,
+      "step": 10738
+    },
+    {
+      "epoch": 3.2220721734563735,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.8996836873023046e-05,
+      "loss": 0.128,
+      "step": 10739
+    },
+    {
+      "epoch": 3.222372270988071,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.898177436360898e-05,
+      "loss": 0.1303,
+      "step": 10740
+    },
+    {
+      "epoch": 3.222672368519769,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.8966711854194906e-05,
+      "loss": 0.1007,
+      "step": 10741
+    },
+    {
+      "epoch": 3.2229724660514667,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.895164934478085e-05,
+      "loss": 0.0912,
+      "step": 10742
+    },
+    {
+      "epoch": 3.2232725635831647,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.8936586835366774e-05,
+      "loss": 0.1014,
+      "step": 10743
+    },
+    {
+      "epoch": 3.2235726611148623,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.892152432595271e-05,
+      "loss": 0.1132,
+      "step": 10744
+    },
+    {
+      "epoch": 3.2238727586465603,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.8906461816538634e-05,
+      "loss": 0.1221,
+      "step": 10745
+    },
+    {
+      "epoch": 3.224172856178258,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.8891399307124574e-05,
+      "loss": 0.0798,
+      "step": 10746
+    },
+    {
+      "epoch": 3.224472953709956,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.88763367977105e-05,
+      "loss": 0.1138,
+      "step": 10747
+    },
+    {
+      "epoch": 3.2247730512416535,
+      "grad_norm": 1.0,
+      "learning_rate": 3.8861274288296434e-05,
+      "loss": 0.1499,
+      "step": 10748
+    },
+    {
+      "epoch": 3.2250731487733515,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.884621177888236e-05,
+      "loss": 0.1096,
+      "step": 10749
+    },
+    {
+      "epoch": 3.225373246305049,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.8831149269468295e-05,
+      "loss": 0.1034,
+      "step": 10750
+    },
+    {
+      "epoch": 3.225673343836747,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.881608676005423e-05,
+      "loss": 0.1087,
+      "step": 10751
+    },
+    {
+      "epoch": 3.2259734413684447,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.880102425064016e-05,
+      "loss": 0.0948,
+      "step": 10752
+    },
+    {
+      "epoch": 3.2262735389001427,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.878596174122609e-05,
+      "loss": 0.1056,
+      "step": 10753
+    },
+    {
+      "epoch": 3.2265736364318403,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.877089923181202e-05,
+      "loss": 0.1006,
+      "step": 10754
+    },
+    {
+      "epoch": 3.2268737339635383,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.8755836722397956e-05,
+      "loss": 0.0974,
+      "step": 10755
+    },
+    {
+      "epoch": 3.227173831495236,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.874077421298388e-05,
+      "loss": 0.0881,
+      "step": 10756
+    },
+    {
+      "epoch": 3.227473929026934,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.8725711703569816e-05,
+      "loss": 0.1225,
+      "step": 10757
+    },
+    {
+      "epoch": 3.2277740265586314,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.871064919415574e-05,
+      "loss": 0.1216,
+      "step": 10758
+    },
+    {
+      "epoch": 3.2280741240903295,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.869558668474168e-05,
+      "loss": 0.1234,
+      "step": 10759
+    },
+    {
+      "epoch": 3.228374221622027,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.868052417532761e-05,
+      "loss": 0.107,
+      "step": 10760
+    },
+    {
+      "epoch": 3.228674319153725,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.866546166591354e-05,
+      "loss": 0.1448,
+      "step": 10761
+    },
+    {
+      "epoch": 3.2289744166854226,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.865039915649947e-05,
+      "loss": 0.1149,
+      "step": 10762
+    },
+    {
+      "epoch": 3.2292745142171206,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.863533664708541e-05,
+      "loss": 0.1329,
+      "step": 10763
+    },
+    {
+      "epoch": 3.229574611748818,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.862027413767134e-05,
+      "loss": 0.1031,
+      "step": 10764
+    },
+    {
+      "epoch": 3.2298747092805162,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.860521162825727e-05,
+      "loss": 0.1052,
+      "step": 10765
+    },
+    {
+      "epoch": 3.230174806812214,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.85901491188432e-05,
+      "loss": 0.0886,
+      "step": 10766
+    },
+    {
+      "epoch": 3.230474904343912,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.857508660942913e-05,
+      "loss": 0.1223,
+      "step": 10767
+    },
+    {
+      "epoch": 3.2307750018756094,
+      "grad_norm": 0.75,
+      "learning_rate": 3.8560024100015064e-05,
+      "loss": 0.1083,
+      "step": 10768
+    },
+    {
+      "epoch": 3.2310750994073074,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.8544961590601e-05,
+      "loss": 0.1065,
+      "step": 10769
+    },
+    {
+      "epoch": 3.231375196939005,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.8529899081186925e-05,
+      "loss": 0.0943,
+      "step": 10770
+    },
+    {
+      "epoch": 3.231675294470703,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.851483657177286e-05,
+      "loss": 0.1311,
+      "step": 10771
+    },
+    {
+      "epoch": 3.2319753920024006,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.849977406235879e-05,
+      "loss": 0.0964,
+      "step": 10772
+    },
+    {
+      "epoch": 3.2322754895340986,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.8484711552944725e-05,
+      "loss": 0.0962,
+      "step": 10773
+    },
+    {
+      "epoch": 3.2325755870657966,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.846964904353065e-05,
+      "loss": 0.0851,
+      "step": 10774
+    },
+    {
+      "epoch": 3.232875684597494,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.8454586534116586e-05,
+      "loss": 0.1179,
+      "step": 10775
+    },
+    {
+      "epoch": 3.2331757821291918,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.843952402470252e-05,
+      "loss": 0.0936,
+      "step": 10776
+    },
+    {
+      "epoch": 3.23347587966089,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.842446151528845e-05,
+      "loss": 0.127,
+      "step": 10777
+    },
+    {
+      "epoch": 3.233775977192588,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.840939900587438e-05,
+      "loss": 0.111,
+      "step": 10778
+    },
+    {
+      "epoch": 3.2340760747242854,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.839433649646031e-05,
+      "loss": 0.1264,
+      "step": 10779
+    },
+    {
+      "epoch": 3.2343761722559834,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.8379273987046246e-05,
+      "loss": 0.0925,
+      "step": 10780
+    },
+    {
+      "epoch": 3.234676269787681,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.836421147763217e-05,
+      "loss": 0.1072,
+      "step": 10781
+    },
+    {
+      "epoch": 3.234976367319379,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.834914896821811e-05,
+      "loss": 0.1207,
+      "step": 10782
+    },
+    {
+      "epoch": 3.2352764648510766,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.8334086458804033e-05,
+      "loss": 0.1132,
+      "step": 10783
+    },
+    {
+      "epoch": 3.2355765623827746,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.8319023949389974e-05,
+      "loss": 0.114,
+      "step": 10784
+    },
+    {
+      "epoch": 3.235876659914472,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.83039614399759e-05,
+      "loss": 0.103,
+      "step": 10785
+    },
+    {
+      "epoch": 3.23617675744617,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.8288898930561834e-05,
+      "loss": 0.1009,
+      "step": 10786
+    },
+    {
+      "epoch": 3.2364768549778677,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.827383642114776e-05,
+      "loss": 0.0934,
+      "step": 10787
+    },
+    {
+      "epoch": 3.2367769525095658,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.8258773911733694e-05,
+      "loss": 0.1341,
+      "step": 10788
+    },
+    {
+      "epoch": 3.2370770500412633,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.824371140231963e-05,
+      "loss": 0.1055,
+      "step": 10789
+    },
+    {
+      "epoch": 3.2373771475729614,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.822864889290556e-05,
+      "loss": 0.1014,
+      "step": 10790
+    },
+    {
+      "epoch": 3.237677245104659,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.821358638349149e-05,
+      "loss": 0.1344,
+      "step": 10791
+    },
+    {
+      "epoch": 3.237977342636357,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.819852387407742e-05,
+      "loss": 0.1213,
+      "step": 10792
+    },
+    {
+      "epoch": 3.2382774401680545,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.8183461364663355e-05,
+      "loss": 0.132,
+      "step": 10793
+    },
+    {
+      "epoch": 3.2385775376997525,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.816839885524929e-05,
+      "loss": 0.1012,
+      "step": 10794
+    },
+    {
+      "epoch": 3.23887763523145,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.8153336345835215e-05,
+      "loss": 0.1286,
+      "step": 10795
+    },
+    {
+      "epoch": 3.239177732763148,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.813827383642115e-05,
+      "loss": 0.0935,
+      "step": 10796
+    },
+    {
+      "epoch": 3.2394778302948457,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.812321132700708e-05,
+      "loss": 0.1232,
+      "step": 10797
+    },
+    {
+      "epoch": 3.2397779278265437,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.8108148817593016e-05,
+      "loss": 0.1064,
+      "step": 10798
+    },
+    {
+      "epoch": 3.2400780253582413,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.809308630817894e-05,
+      "loss": 0.1164,
+      "step": 10799
+    },
+    {
+      "epoch": 3.2403781228899393,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.8078023798764876e-05,
+      "loss": 0.095,
+      "step": 10800
+    },
+    {
+      "epoch": 3.240678220421637,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.806296128935081e-05,
+      "loss": 0.0908,
+      "step": 10801
+    },
+    {
+      "epoch": 3.240978317953335,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.804789877993674e-05,
+      "loss": 0.138,
+      "step": 10802
+    },
+    {
+      "epoch": 3.2412784154850325,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.803283627052267e-05,
+      "loss": 0.1162,
+      "step": 10803
+    },
+    {
+      "epoch": 3.2415785130167305,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.8017773761108604e-05,
+      "loss": 0.0997,
+      "step": 10804
+    },
+    {
+      "epoch": 3.241878610548428,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.800271125169454e-05,
+      "loss": 0.1181,
+      "step": 10805
+    },
+    {
+      "epoch": 3.242178708080126,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.798764874228047e-05,
+      "loss": 0.1198,
+      "step": 10806
+    },
+    {
+      "epoch": 3.2424788056118237,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.79725862328664e-05,
+      "loss": 0.1242,
+      "step": 10807
+    },
+    {
+      "epoch": 3.2427789031435217,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.7957523723452324e-05,
+      "loss": 0.103,
+      "step": 10808
+    },
+    {
+      "epoch": 3.2430790006752193,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.794246121403826e-05,
+      "loss": 0.118,
+      "step": 10809
+    },
+    {
+      "epoch": 3.2433790982069173,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.792739870462419e-05,
+      "loss": 0.1319,
+      "step": 10810
+    },
+    {
+      "epoch": 3.243679195738615,
+      "grad_norm": 0.75,
+      "learning_rate": 3.7912336195210125e-05,
+      "loss": 0.0981,
+      "step": 10811
+    },
+    {
+      "epoch": 3.243979293270313,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.789727368579605e-05,
+      "loss": 0.0991,
+      "step": 10812
+    },
+    {
+      "epoch": 3.244279390802011,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.7882211176381985e-05,
+      "loss": 0.1271,
+      "step": 10813
+    },
+    {
+      "epoch": 3.2445794883337085,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.786714866696792e-05,
+      "loss": 0.0856,
+      "step": 10814
+    },
+    {
+      "epoch": 3.244879585865406,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.785208615755385e-05,
+      "loss": 0.0876,
+      "step": 10815
+    },
+    {
+      "epoch": 3.245179683397104,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.783702364813978e-05,
+      "loss": 0.111,
+      "step": 10816
+    },
+    {
+      "epoch": 3.245479780928802,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.782196113872571e-05,
+      "loss": 0.0892,
+      "step": 10817
+    },
+    {
+      "epoch": 3.2457798784604996,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.7806898629311646e-05,
+      "loss": 0.1187,
+      "step": 10818
+    },
+    {
+      "epoch": 3.2460799759921977,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.779183611989758e-05,
+      "loss": 0.1707,
+      "step": 10819
+    },
+    {
+      "epoch": 3.2463800735238952,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.7776773610483506e-05,
+      "loss": 0.1224,
+      "step": 10820
+    },
+    {
+      "epoch": 3.2466801710555933,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.776171110106944e-05,
+      "loss": 0.1203,
+      "step": 10821
+    },
+    {
+      "epoch": 3.246980268587291,
+      "grad_norm": 0.875,
+      "learning_rate": 3.774664859165537e-05,
+      "loss": 0.127,
+      "step": 10822
+    },
+    {
+      "epoch": 3.247280366118989,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.773158608224131e-05,
+      "loss": 0.1154,
+      "step": 10823
+    },
+    {
+      "epoch": 3.2475804636506864,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.7716523572827234e-05,
+      "loss": 0.1031,
+      "step": 10824
+    },
+    {
+      "epoch": 3.2478805611823844,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.770146106341317e-05,
+      "loss": 0.1303,
+      "step": 10825
+    },
+    {
+      "epoch": 3.248180658714082,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.7686398553999094e-05,
+      "loss": 0.0996,
+      "step": 10826
+    },
+    {
+      "epoch": 3.24848075624578,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.7671336044585034e-05,
+      "loss": 0.0982,
+      "step": 10827
+    },
+    {
+      "epoch": 3.2487808537774776,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.765627353517096e-05,
+      "loss": 0.1031,
+      "step": 10828
+    },
+    {
+      "epoch": 3.2490809513091756,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.7641211025756894e-05,
+      "loss": 0.1089,
+      "step": 10829
+    },
+    {
+      "epoch": 3.249381048840873,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.762614851634282e-05,
+      "loss": 0.0816,
+      "step": 10830
+    },
+    {
+      "epoch": 3.249681146372571,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.761108600692876e-05,
+      "loss": 0.1175,
+      "step": 10831
+    },
+    {
+      "epoch": 3.249981243904269,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.759602349751469e-05,
+      "loss": 0.1073,
+      "step": 10832
+    },
+    {
+      "epoch": 3.250281341435967,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.758096098810062e-05,
+      "loss": 0.1098,
+      "step": 10833
+    },
+    {
+      "epoch": 3.2505814389676644,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.756589847868655e-05,
+      "loss": 0.1049,
+      "step": 10834
+    },
+    {
+      "epoch": 3.2508815364993624,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.755083596927248e-05,
+      "loss": 0.0918,
+      "step": 10835
+    },
+    {
+      "epoch": 3.25118163403106,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.7535773459858416e-05,
+      "loss": 0.0873,
+      "step": 10836
+    },
+    {
+      "epoch": 3.251481731562758,
+      "grad_norm": 0.75,
+      "learning_rate": 3.752071095044434e-05,
+      "loss": 0.0962,
+      "step": 10837
+    },
+    {
+      "epoch": 3.2517818290944556,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.7505648441030276e-05,
+      "loss": 0.1109,
+      "step": 10838
+    },
+    {
+      "epoch": 3.2520819266261536,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.749058593161621e-05,
+      "loss": 0.0999,
+      "step": 10839
+    },
+    {
+      "epoch": 3.252382024157851,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.747552342220214e-05,
+      "loss": 0.114,
+      "step": 10840
+    },
+    {
+      "epoch": 3.252682121689549,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.746046091278807e-05,
+      "loss": 0.0869,
+      "step": 10841
+    },
+    {
+      "epoch": 3.2529822192212468,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.7445398403374e-05,
+      "loss": 0.114,
+      "step": 10842
+    },
+    {
+      "epoch": 3.2532823167529448,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.743033589395994e-05,
+      "loss": 0.0902,
+      "step": 10843
+    },
+    {
+      "epoch": 3.2535824142846423,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.741527338454587e-05,
+      "loss": 0.115,
+      "step": 10844
+    },
+    {
+      "epoch": 3.2538825118163404,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.74002108751318e-05,
+      "loss": 0.0921,
+      "step": 10845
+    },
+    {
+      "epoch": 3.254182609348038,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.738514836571773e-05,
+      "loss": 0.0815,
+      "step": 10846
+    },
+    {
+      "epoch": 3.254482706879736,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.737008585630366e-05,
+      "loss": 0.1335,
+      "step": 10847
+    },
+    {
+      "epoch": 3.2547828044114335,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.73550233468896e-05,
+      "loss": 0.1107,
+      "step": 10848
+    },
+    {
+      "epoch": 3.2550829019431315,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.7339960837475524e-05,
+      "loss": 0.1025,
+      "step": 10849
+    },
+    {
+      "epoch": 3.255382999474829,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.732489832806146e-05,
+      "loss": 0.0854,
+      "step": 10850
+    },
+    {
+      "epoch": 3.255683097006527,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.7309835818647385e-05,
+      "loss": 0.1203,
+      "step": 10851
+    },
+    {
+      "epoch": 3.255983194538225,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.7294773309233325e-05,
+      "loss": 0.1069,
+      "step": 10852
+    },
+    {
+      "epoch": 3.2562832920699227,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.727971079981925e-05,
+      "loss": 0.0968,
+      "step": 10853
+    },
+    {
+      "epoch": 3.2565833896016203,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.7264648290405185e-05,
+      "loss": 0.1074,
+      "step": 10854
+    },
+    {
+      "epoch": 3.2568834871333183,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.724958578099111e-05,
+      "loss": 0.1079,
+      "step": 10855
+    },
+    {
+      "epoch": 3.2571835846650163,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.7234523271577046e-05,
+      "loss": 0.101,
+      "step": 10856
+    },
+    {
+      "epoch": 3.257483682196714,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.721946076216298e-05,
+      "loss": 0.077,
+      "step": 10857
+    },
+    {
+      "epoch": 3.2577837797284115,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.720439825274891e-05,
+      "loss": 0.0952,
+      "step": 10858
+    },
+    {
+      "epoch": 3.2580838772601095,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.718933574333484e-05,
+      "loss": 0.1068,
+      "step": 10859
+    },
+    {
+      "epoch": 3.2583839747918075,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.717427323392077e-05,
+      "loss": 0.1216,
+      "step": 10860
+    },
+    {
+      "epoch": 3.258684072323505,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.7159210724506706e-05,
+      "loss": 0.0839,
+      "step": 10861
+    },
+    {
+      "epoch": 3.258984169855203,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.714414821509264e-05,
+      "loss": 0.1067,
+      "step": 10862
+    },
+    {
+      "epoch": 3.2592842673869007,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.712908570567857e-05,
+      "loss": 0.1028,
+      "step": 10863
+    },
+    {
+      "epoch": 3.2595843649185987,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.71140231962645e-05,
+      "loss": 0.1064,
+      "step": 10864
+    },
+    {
+      "epoch": 3.2598844624502963,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.7098960686850434e-05,
+      "loss": 0.1095,
+      "step": 10865
+    },
+    {
+      "epoch": 3.2601845599819943,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.708389817743636e-05,
+      "loss": 0.1111,
+      "step": 10866
+    },
+    {
+      "epoch": 3.260484657513692,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.7068835668022294e-05,
+      "loss": 0.0887,
+      "step": 10867
+    },
+    {
+      "epoch": 3.26078475504539,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.705377315860822e-05,
+      "loss": 0.1272,
+      "step": 10868
+    },
+    {
+      "epoch": 3.2610848525770875,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.703871064919416e-05,
+      "loss": 0.0951,
+      "step": 10869
+    },
+    {
+      "epoch": 3.2613849501087855,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.702364813978009e-05,
+      "loss": 0.121,
+      "step": 10870
+    },
+    {
+      "epoch": 3.261685047640483,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.700858563036602e-05,
+      "loss": 0.1266,
+      "step": 10871
+    },
+    {
+      "epoch": 3.261985145172181,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.699352312095195e-05,
+      "loss": 0.1102,
+      "step": 10872
+    },
+    {
+      "epoch": 3.2622852427038787,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.697846061153789e-05,
+      "loss": 0.1011,
+      "step": 10873
+    },
+    {
+      "epoch": 3.2625853402355767,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.6963398102123815e-05,
+      "loss": 0.1107,
+      "step": 10874
+    },
+    {
+      "epoch": 3.2628854377672742,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.694833559270975e-05,
+      "loss": 0.1135,
+      "step": 10875
+    },
+    {
+      "epoch": 3.2631855352989723,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.6933273083295675e-05,
+      "loss": 0.0915,
+      "step": 10876
+    },
+    {
+      "epoch": 3.26348563283067,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.691821057388161e-05,
+      "loss": 0.1436,
+      "step": 10877
+    },
+    {
+      "epoch": 3.263785730362368,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.690314806446754e-05,
+      "loss": 0.1219,
+      "step": 10878
+    },
+    {
+      "epoch": 3.2640858278940654,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.6888085555053476e-05,
+      "loss": 0.108,
+      "step": 10879
+    },
+    {
+      "epoch": 3.2643859254257634,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.68730230456394e-05,
+      "loss": 0.1117,
+      "step": 10880
+    },
+    {
+      "epoch": 3.264686022957461,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.6857960536225336e-05,
+      "loss": 0.119,
+      "step": 10881
+    },
+    {
+      "epoch": 3.264986120489159,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.684289802681127e-05,
+      "loss": 0.1216,
+      "step": 10882
+    },
+    {
+      "epoch": 3.2652862180208566,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.68278355173972e-05,
+      "loss": 0.1153,
+      "step": 10883
+    },
+    {
+      "epoch": 3.2655863155525546,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.681277300798313e-05,
+      "loss": 0.1032,
+      "step": 10884
+    },
+    {
+      "epoch": 3.265886413084252,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.6797710498569064e-05,
+      "loss": 0.1142,
+      "step": 10885
+    },
+    {
+      "epoch": 3.2661865106159502,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.6782647989155e-05,
+      "loss": 0.144,
+      "step": 10886
+    },
+    {
+      "epoch": 3.266486608147648,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.676758547974093e-05,
+      "loss": 0.1091,
+      "step": 10887
+    },
+    {
+      "epoch": 3.266786705679346,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.675252297032686e-05,
+      "loss": 0.1208,
+      "step": 10888
+    },
+    {
+      "epoch": 3.2670868032110434,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.673746046091279e-05,
+      "loss": 0.1059,
+      "step": 10889
+    },
+    {
+      "epoch": 3.2673869007427414,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.6722397951498725e-05,
+      "loss": 0.1116,
+      "step": 10890
+    },
+    {
+      "epoch": 3.2676869982744394,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.670733544208465e-05,
+      "loss": 0.0889,
+      "step": 10891
+    },
+    {
+      "epoch": 3.267987095806137,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.6692272932670585e-05,
+      "loss": 0.1109,
+      "step": 10892
+    },
+    {
+      "epoch": 3.2682871933378346,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.667721042325651e-05,
+      "loss": 0.1421,
+      "step": 10893
+    },
+    {
+      "epoch": 3.2685872908695326,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.6662147913842445e-05,
+      "loss": 0.105,
+      "step": 10894
+    },
+    {
+      "epoch": 3.2688873884012306,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.664708540442838e-05,
+      "loss": 0.1266,
+      "step": 10895
+    },
+    {
+      "epoch": 3.269187485932928,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.663202289501431e-05,
+      "loss": 0.1132,
+      "step": 10896
+    },
+    {
+      "epoch": 3.2694875834646258,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.661696038560024e-05,
+      "loss": 0.0984,
+      "step": 10897
+    },
+    {
+      "epoch": 3.2697876809963238,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.660189787618617e-05,
+      "loss": 0.1477,
+      "step": 10898
+    },
+    {
+      "epoch": 3.270087778528022,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.6586835366772106e-05,
+      "loss": 0.1418,
+      "step": 10899
+    },
+    {
+      "epoch": 3.2703878760597194,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.657177285735804e-05,
+      "loss": 0.117,
+      "step": 10900
+    },
+    {
+      "epoch": 3.2706879735914174,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.6556710347943966e-05,
+      "loss": 0.1225,
+      "step": 10901
+    },
+    {
+      "epoch": 3.270988071123115,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.65416478385299e-05,
+      "loss": 0.0928,
+      "step": 10902
+    },
+    {
+      "epoch": 3.271288168654813,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.652658532911583e-05,
+      "loss": 0.1102,
+      "step": 10903
+    },
+    {
+      "epoch": 3.2715882661865106,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.651152281970177e-05,
+      "loss": 0.1257,
+      "step": 10904
+    },
+    {
+      "epoch": 3.2718883637182086,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.6496460310287694e-05,
+      "loss": 0.104,
+      "step": 10905
+    },
+    {
+      "epoch": 3.272188461249906,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.648139780087363e-05,
+      "loss": 0.0957,
+      "step": 10906
+    },
+    {
+      "epoch": 3.272488558781604,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.646633529145956e-05,
+      "loss": 0.1471,
+      "step": 10907
+    },
+    {
+      "epoch": 3.2727886563133017,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.6451272782045494e-05,
+      "loss": 0.0854,
+      "step": 10908
+    },
+    {
+      "epoch": 3.2730887538449998,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.643621027263142e-05,
+      "loss": 0.1207,
+      "step": 10909
+    },
+    {
+      "epoch": 3.2733888513766973,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.6421147763217354e-05,
+      "loss": 0.1303,
+      "step": 10910
+    },
+    {
+      "epoch": 3.2736889489083953,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.640608525380329e-05,
+      "loss": 0.1074,
+      "step": 10911
+    },
+    {
+      "epoch": 3.273989046440093,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.639102274438922e-05,
+      "loss": 0.1014,
+      "step": 10912
+    },
+    {
+      "epoch": 3.274289143971791,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.637596023497515e-05,
+      "loss": 0.1091,
+      "step": 10913
+    },
+    {
+      "epoch": 3.2745892415034885,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.636089772556108e-05,
+      "loss": 0.1062,
+      "step": 10914
+    },
+    {
+      "epoch": 3.2748893390351865,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.634583521614701e-05,
+      "loss": 0.1177,
+      "step": 10915
+    },
+    {
+      "epoch": 3.275189436566884,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.633077270673295e-05,
+      "loss": 0.1144,
+      "step": 10916
+    },
+    {
+      "epoch": 3.275489534098582,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.6315710197318876e-05,
+      "loss": 0.1058,
+      "step": 10917
+    },
+    {
+      "epoch": 3.2757896316302797,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.63006476879048e-05,
+      "loss": 0.1104,
+      "step": 10918
+    },
+    {
+      "epoch": 3.2760897291619777,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.6285585178490736e-05,
+      "loss": 0.1272,
+      "step": 10919
+    },
+    {
+      "epoch": 3.2763898266936753,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.627052266907667e-05,
+      "loss": 0.1195,
+      "step": 10920
+    },
+    {
+      "epoch": 3.2766899242253733,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.62554601596626e-05,
+      "loss": 0.1359,
+      "step": 10921
+    },
+    {
+      "epoch": 3.276990021757071,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.624039765024853e-05,
+      "loss": 0.0875,
+      "step": 10922
+    },
+    {
+      "epoch": 3.277290119288769,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.622533514083446e-05,
+      "loss": 0.1105,
+      "step": 10923
+    },
+    {
+      "epoch": 3.2775902168204665,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.62102726314204e-05,
+      "loss": 0.0946,
+      "step": 10924
+    },
+    {
+      "epoch": 3.2778903143521645,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.619521012200633e-05,
+      "loss": 0.1004,
+      "step": 10925
+    },
+    {
+      "epoch": 3.278190411883862,
+      "grad_norm": 0.75,
+      "learning_rate": 3.618014761259226e-05,
+      "loss": 0.1065,
+      "step": 10926
+    },
+    {
+      "epoch": 3.27849050941556,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.616508510317819e-05,
+      "loss": 0.0914,
+      "step": 10927
+    },
+    {
+      "epoch": 3.2787906069472577,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.6150022593764124e-05,
+      "loss": 0.0998,
+      "step": 10928
+    },
+    {
+      "epoch": 3.2790907044789557,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.613496008435006e-05,
+      "loss": 0.1032,
+      "step": 10929
+    },
+    {
+      "epoch": 3.2793908020106537,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.6119897574935984e-05,
+      "loss": 0.111,
+      "step": 10930
+    },
+    {
+      "epoch": 3.2796908995423513,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.610483506552192e-05,
+      "loss": 0.0939,
+      "step": 10931
+    },
+    {
+      "epoch": 3.279990997074049,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.608977255610785e-05,
+      "loss": 0.1241,
+      "step": 10932
+    },
+    {
+      "epoch": 3.280291094605747,
+      "grad_norm": 0.75,
+      "learning_rate": 3.6074710046693785e-05,
+      "loss": 0.116,
+      "step": 10933
+    },
+    {
+      "epoch": 3.280591192137445,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.605964753727971e-05,
+      "loss": 0.1121,
+      "step": 10934
+    },
+    {
+      "epoch": 3.2808912896691425,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.6044585027865645e-05,
+      "loss": 0.0918,
+      "step": 10935
+    },
+    {
+      "epoch": 3.28119138720084,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.602952251845157e-05,
+      "loss": 0.105,
+      "step": 10936
+    },
+    {
+      "epoch": 3.281491484732538,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.601446000903751e-05,
+      "loss": 0.1068,
+      "step": 10937
+    },
+    {
+      "epoch": 3.281791582264236,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.599939749962344e-05,
+      "loss": 0.1012,
+      "step": 10938
+    },
+    {
+      "epoch": 3.2820916797959336,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.598433499020937e-05,
+      "loss": 0.1129,
+      "step": 10939
+    },
+    {
+      "epoch": 3.2823917773276317,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.59692724807953e-05,
+      "loss": 0.0926,
+      "step": 10940
+    },
+    {
+      "epoch": 3.2826918748593292,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.595420997138124e-05,
+      "loss": 0.0906,
+      "step": 10941
+    },
+    {
+      "epoch": 3.2829919723910272,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.5939147461967166e-05,
+      "loss": 0.0926,
+      "step": 10942
+    },
+    {
+      "epoch": 3.283292069922725,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.59240849525531e-05,
+      "loss": 0.1082,
+      "step": 10943
+    },
+    {
+      "epoch": 3.283592167454423,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.590902244313903e-05,
+      "loss": 0.126,
+      "step": 10944
+    },
+    {
+      "epoch": 3.2838922649861204,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.589395993372496e-05,
+      "loss": 0.1192,
+      "step": 10945
+    },
+    {
+      "epoch": 3.2841923625178184,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.5878897424310894e-05,
+      "loss": 0.0858,
+      "step": 10946
+    },
+    {
+      "epoch": 3.284492460049516,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.586383491489682e-05,
+      "loss": 0.0993,
+      "step": 10947
+    },
+    {
+      "epoch": 3.284792557581214,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.5848772405482754e-05,
+      "loss": 0.0984,
+      "step": 10948
+    },
+    {
+      "epoch": 3.2850926551129116,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.583370989606869e-05,
+      "loss": 0.1013,
+      "step": 10949
+    },
+    {
+      "epoch": 3.2853927526446096,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.581864738665462e-05,
+      "loss": 0.0814,
+      "step": 10950
+    },
+    {
+      "epoch": 3.285692850176307,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.580358487724055e-05,
+      "loss": 0.1024,
+      "step": 10951
+    },
+    {
+      "epoch": 3.285992947708005,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.578852236782648e-05,
+      "loss": 0.1168,
+      "step": 10952
+    },
+    {
+      "epoch": 3.286293045239703,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.577345985841241e-05,
+      "loss": 0.1216,
+      "step": 10953
+    },
+    {
+      "epoch": 3.286593142771401,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.575839734899835e-05,
+      "loss": 0.1173,
+      "step": 10954
+    },
+    {
+      "epoch": 3.2868932403030984,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.5743334839584275e-05,
+      "loss": 0.1295,
+      "step": 10955
+    },
+    {
+      "epoch": 3.2871933378347964,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.572827233017021e-05,
+      "loss": 0.0933,
+      "step": 10956
+    },
+    {
+      "epoch": 3.287493435366494,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.5713209820756135e-05,
+      "loss": 0.1207,
+      "step": 10957
+    },
+    {
+      "epoch": 3.287793532898192,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.5698147311342076e-05,
+      "loss": 0.1084,
+      "step": 10958
+    },
+    {
+      "epoch": 3.2880936304298896,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.5683084801928e-05,
+      "loss": 0.0998,
+      "step": 10959
+    },
+    {
+      "epoch": 3.2883937279615876,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.5668022292513936e-05,
+      "loss": 0.1079,
+      "step": 10960
+    },
+    {
+      "epoch": 3.288693825493285,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.565295978309986e-05,
+      "loss": 0.1198,
+      "step": 10961
+    },
+    {
+      "epoch": 3.288993923024983,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.5637897273685796e-05,
+      "loss": 0.1128,
+      "step": 10962
+    },
+    {
+      "epoch": 3.2892940205566807,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.562283476427173e-05,
+      "loss": 0.1017,
+      "step": 10963
+    },
+    {
+      "epoch": 3.2895941180883788,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.5607772254857663e-05,
+      "loss": 0.1155,
+      "step": 10964
+    },
+    {
+      "epoch": 3.2898942156200763,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.559270974544359e-05,
+      "loss": 0.1081,
+      "step": 10965
+    },
+    {
+      "epoch": 3.2901943131517744,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.5577647236029524e-05,
+      "loss": 0.1193,
+      "step": 10966
+    },
+    {
+      "epoch": 3.290494410683472,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.556258472661546e-05,
+      "loss": 0.0939,
+      "step": 10967
+    },
+    {
+      "epoch": 3.29079450821517,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.554752221720139e-05,
+      "loss": 0.0928,
+      "step": 10968
+    },
+    {
+      "epoch": 3.291094605746868,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.553245970778732e-05,
+      "loss": 0.1015,
+      "step": 10969
+    },
+    {
+      "epoch": 3.2913947032785655,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.551739719837325e-05,
+      "loss": 0.0988,
+      "step": 10970
+    },
+    {
+      "epoch": 3.291694800810263,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.5502334688959185e-05,
+      "loss": 0.1011,
+      "step": 10971
+    },
+    {
+      "epoch": 3.291994898341961,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.548727217954511e-05,
+      "loss": 0.0978,
+      "step": 10972
+    },
+    {
+      "epoch": 3.292294995873659,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.5472209670131045e-05,
+      "loss": 0.1094,
+      "step": 10973
+    },
+    {
+      "epoch": 3.2925950934053567,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.545714716071697e-05,
+      "loss": 0.1041,
+      "step": 10974
+    },
+    {
+      "epoch": 3.2928951909370543,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.544208465130291e-05,
+      "loss": 0.1117,
+      "step": 10975
+    },
+    {
+      "epoch": 3.2931952884687523,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.542702214188884e-05,
+      "loss": 0.0813,
+      "step": 10976
+    },
+    {
+      "epoch": 3.2934953860004503,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.541195963247477e-05,
+      "loss": 0.1084,
+      "step": 10977
+    },
+    {
+      "epoch": 3.293795483532148,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.53968971230607e-05,
+      "loss": 0.0989,
+      "step": 10978
+    },
+    {
+      "epoch": 3.294095581063846,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.538183461364664e-05,
+      "loss": 0.0926,
+      "step": 10979
+    },
+    {
+      "epoch": 3.2943956785955435,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.5366772104232566e-05,
+      "loss": 0.1094,
+      "step": 10980
+    },
+    {
+      "epoch": 3.2946957761272415,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.53517095948185e-05,
+      "loss": 0.1163,
+      "step": 10981
+    },
+    {
+      "epoch": 3.294995873658939,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.5336647085404426e-05,
+      "loss": 0.1299,
+      "step": 10982
+    },
+    {
+      "epoch": 3.295295971190637,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.532158457599036e-05,
+      "loss": 0.0958,
+      "step": 10983
+    },
+    {
+      "epoch": 3.2955960687223347,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.530652206657629e-05,
+      "loss": 0.134,
+      "step": 10984
+    },
+    {
+      "epoch": 3.2958961662540327,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.529145955716223e-05,
+      "loss": 0.0871,
+      "step": 10985
+    },
+    {
+      "epoch": 3.2961962637857303,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.5276397047748154e-05,
+      "loss": 0.1104,
+      "step": 10986
+    },
+    {
+      "epoch": 3.2964963613174283,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.526133453833409e-05,
+      "loss": 0.1163,
+      "step": 10987
+    },
+    {
+      "epoch": 3.296796458849126,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.524627202892002e-05,
+      "loss": 0.1291,
+      "step": 10988
+    },
+    {
+      "epoch": 3.297096556380824,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.5231209519505954e-05,
+      "loss": 0.1235,
+      "step": 10989
+    },
+    {
+      "epoch": 3.2973966539125215,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.521614701009188e-05,
+      "loss": 0.0981,
+      "step": 10990
+    },
+    {
+      "epoch": 3.2976967514442195,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.5201084500677814e-05,
+      "loss": 0.1004,
+      "step": 10991
+    },
+    {
+      "epoch": 3.297996848975917,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.518602199126375e-05,
+      "loss": 0.1123,
+      "step": 10992
+    },
+    {
+      "epoch": 3.298296946507615,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.517095948184968e-05,
+      "loss": 0.1025,
+      "step": 10993
+    },
+    {
+      "epoch": 3.2985970440393126,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.515589697243561e-05,
+      "loss": 0.0858,
+      "step": 10994
+    },
+    {
+      "epoch": 3.2988971415710107,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.514083446302154e-05,
+      "loss": 0.1189,
+      "step": 10995
+    },
+    {
+      "epoch": 3.2991972391027082,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.5125771953607475e-05,
+      "loss": 0.125,
+      "step": 10996
+    },
+    {
+      "epoch": 3.2994973366344063,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.511070944419341e-05,
+      "loss": 0.1027,
+      "step": 10997
+    },
+    {
+      "epoch": 3.299797434166104,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.5095646934779336e-05,
+      "loss": 0.1229,
+      "step": 10998
+    },
+    {
+      "epoch": 3.300097531697802,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.508058442536526e-05,
+      "loss": 0.1195,
+      "step": 10999
+    },
+    {
+      "epoch": 3.3003976292294994,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.50655219159512e-05,
+      "loss": 0.1086,
+      "step": 11000
+    },
+    {
+      "epoch": 3.3006977267611974,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.505045940653713e-05,
+      "loss": 0.1098,
+      "step": 11001
+    },
+    {
+      "epoch": 3.300997824292895,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.503539689712306e-05,
+      "loss": 0.1143,
+      "step": 11002
+    },
+    {
+      "epoch": 3.301297921824593,
+      "grad_norm": 0.75,
+      "learning_rate": 3.502033438770899e-05,
+      "loss": 0.0895,
+      "step": 11003
+    },
+    {
+      "epoch": 3.3015980193562906,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.500527187829492e-05,
+      "loss": 0.0906,
+      "step": 11004
+    },
+    {
+      "epoch": 3.3018981168879886,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.499020936888086e-05,
+      "loss": 0.0963,
+      "step": 11005
+    },
+    {
+      "epoch": 3.302198214419686,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.497514685946679e-05,
+      "loss": 0.0998,
+      "step": 11006
+    },
+    {
+      "epoch": 3.302498311951384,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.496008435005272e-05,
+      "loss": 0.0917,
+      "step": 11007
+    },
+    {
+      "epoch": 3.3027984094830822,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.494502184063865e-05,
+      "loss": 0.1078,
+      "step": 11008
+    },
+    {
+      "epoch": 3.30309850701478,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.4929959331224584e-05,
+      "loss": 0.101,
+      "step": 11009
+    },
+    {
+      "epoch": 3.3033986045464774,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.491489682181052e-05,
+      "loss": 0.1177,
+      "step": 11010
+    },
+    {
+      "epoch": 3.3036987020781754,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.4899834312396444e-05,
+      "loss": 0.0834,
+      "step": 11011
+    },
+    {
+      "epoch": 3.3039987996098734,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.488477180298238e-05,
+      "loss": 0.1278,
+      "step": 11012
+    },
+    {
+      "epoch": 3.304298897141571,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.486970929356831e-05,
+      "loss": 0.1183,
+      "step": 11013
+    },
+    {
+      "epoch": 3.3045989946732686,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.4854646784154245e-05,
+      "loss": 0.0965,
+      "step": 11014
+    },
+    {
+      "epoch": 3.3048990922049666,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.483958427474017e-05,
+      "loss": 0.1191,
+      "step": 11015
+    },
+    {
+      "epoch": 3.3051991897366646,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.4824521765326105e-05,
+      "loss": 0.1178,
+      "step": 11016
+    },
+    {
+      "epoch": 3.305499287268362,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.480945925591204e-05,
+      "loss": 0.0916,
+      "step": 11017
+    },
+    {
+      "epoch": 3.30579938480006,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.479439674649797e-05,
+      "loss": 0.1711,
+      "step": 11018
+    },
+    {
+      "epoch": 3.3060994823317578,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.47793342370839e-05,
+      "loss": 0.1139,
+      "step": 11019
+    },
+    {
+      "epoch": 3.306399579863456,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.476427172766983e-05,
+      "loss": 0.098,
+      "step": 11020
+    },
+    {
+      "epoch": 3.3066996773951534,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.474920921825576e-05,
+      "loss": 0.104,
+      "step": 11021
+    },
+    {
+      "epoch": 3.3069997749268514,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.47341467088417e-05,
+      "loss": 0.1446,
+      "step": 11022
+    },
+    {
+      "epoch": 3.307299872458549,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.4719084199427626e-05,
+      "loss": 0.1102,
+      "step": 11023
+    },
+    {
+      "epoch": 3.307599969990247,
+      "grad_norm": 0.75,
+      "learning_rate": 3.470402169001356e-05,
+      "loss": 0.0999,
+      "step": 11024
+    },
+    {
+      "epoch": 3.3079000675219445,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.468895918059949e-05,
+      "loss": 0.1222,
+      "step": 11025
+    },
+    {
+      "epoch": 3.3082001650536426,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.467389667118542e-05,
+      "loss": 0.1072,
+      "step": 11026
+    },
+    {
+      "epoch": 3.30850026258534,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.4658834161771354e-05,
+      "loss": 0.1012,
+      "step": 11027
+    },
+    {
+      "epoch": 3.308800360117038,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.464377165235728e-05,
+      "loss": 0.117,
+      "step": 11028
+    },
+    {
+      "epoch": 3.3091004576487357,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.4628709142943214e-05,
+      "loss": 0.1153,
+      "step": 11029
+    },
+    {
+      "epoch": 3.3094005551804337,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.461364663352915e-05,
+      "loss": 0.1184,
+      "step": 11030
+    },
+    {
+      "epoch": 3.3097006527121313,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.459858412411508e-05,
+      "loss": 0.1174,
+      "step": 11031
+    },
+    {
+      "epoch": 3.3100007502438293,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.458352161470101e-05,
+      "loss": 0.0891,
+      "step": 11032
+    },
+    {
+      "epoch": 3.310300847775527,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.456845910528694e-05,
+      "loss": 0.139,
+      "step": 11033
+    },
+    {
+      "epoch": 3.310600945307225,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.4553396595872875e-05,
+      "loss": 0.127,
+      "step": 11034
+    },
+    {
+      "epoch": 3.3109010428389225,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.453833408645881e-05,
+      "loss": 0.1259,
+      "step": 11035
+    },
+    {
+      "epoch": 3.3112011403706205,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.4523271577044735e-05,
+      "loss": 0.1056,
+      "step": 11036
+    },
+    {
+      "epoch": 3.311501237902318,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.450820906763067e-05,
+      "loss": 0.0921,
+      "step": 11037
+    },
+    {
+      "epoch": 3.311801335434016,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.44931465582166e-05,
+      "loss": 0.0931,
+      "step": 11038
+    },
+    {
+      "epoch": 3.3121014329657137,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.4478084048802536e-05,
+      "loss": 0.0927,
+      "step": 11039
+    },
+    {
+      "epoch": 3.3124015304974117,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.446302153938846e-05,
+      "loss": 0.1794,
+      "step": 11040
+    },
+    {
+      "epoch": 3.3127016280291093,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.4447959029974396e-05,
+      "loss": 0.1283,
+      "step": 11041
+    },
+    {
+      "epoch": 3.3130017255608073,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.443289652056032e-05,
+      "loss": 0.1091,
+      "step": 11042
+    },
+    {
+      "epoch": 3.313301823092505,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.441783401114626e-05,
+      "loss": 0.1306,
+      "step": 11043
+    },
+    {
+      "epoch": 3.313601920624203,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.440277150173219e-05,
+      "loss": 0.1131,
+      "step": 11044
+    },
+    {
+      "epoch": 3.3139020181559005,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.4387708992318123e-05,
+      "loss": 0.1146,
+      "step": 11045
+    },
+    {
+      "epoch": 3.3142021156875985,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.437264648290405e-05,
+      "loss": 0.0996,
+      "step": 11046
+    },
+    {
+      "epoch": 3.3145022132192965,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.435758397348999e-05,
+      "loss": 0.1281,
+      "step": 11047
+    },
+    {
+      "epoch": 3.314802310750994,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.434252146407592e-05,
+      "loss": 0.1071,
+      "step": 11048
+    },
+    {
+      "epoch": 3.3151024082826916,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.432745895466185e-05,
+      "loss": 0.1191,
+      "step": 11049
+    },
+    {
+      "epoch": 3.3154025058143897,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.431239644524778e-05,
+      "loss": 0.0916,
+      "step": 11050
+    },
+    {
+      "epoch": 3.3157026033460877,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.429733393583371e-05,
+      "loss": 0.0947,
+      "step": 11051
+    },
+    {
+      "epoch": 3.3160027008777853,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.4282271426419645e-05,
+      "loss": 0.0956,
+      "step": 11052
+    },
+    {
+      "epoch": 3.316302798409483,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.426720891700557e-05,
+      "loss": 0.1152,
+      "step": 11053
+    },
+    {
+      "epoch": 3.316602895941181,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.4252146407591505e-05,
+      "loss": 0.0958,
+      "step": 11054
+    },
+    {
+      "epoch": 3.316902993472879,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.423708389817744e-05,
+      "loss": 0.122,
+      "step": 11055
+    },
+    {
+      "epoch": 3.3172030910045764,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.422202138876337e-05,
+      "loss": 0.0833,
+      "step": 11056
+    },
+    {
+      "epoch": 3.3175031885362745,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.42069588793493e-05,
+      "loss": 0.1185,
+      "step": 11057
+    },
+    {
+      "epoch": 3.317803286067972,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.419189636993523e-05,
+      "loss": 0.1411,
+      "step": 11058
+    },
+    {
+      "epoch": 3.31810338359967,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.4176833860521166e-05,
+      "loss": 0.0734,
+      "step": 11059
+    },
+    {
+      "epoch": 3.3184034811313676,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.41617713511071e-05,
+      "loss": 0.0912,
+      "step": 11060
+    },
+    {
+      "epoch": 3.3187035786630656,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.4146708841693026e-05,
+      "loss": 0.1406,
+      "step": 11061
+    },
+    {
+      "epoch": 3.319003676194763,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.413164633227896e-05,
+      "loss": 0.101,
+      "step": 11062
+    },
+    {
+      "epoch": 3.3193037737264612,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.4116583822864886e-05,
+      "loss": 0.1049,
+      "step": 11063
+    },
+    {
+      "epoch": 3.319603871258159,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.4101521313450827e-05,
+      "loss": 0.1078,
+      "step": 11064
+    },
+    {
+      "epoch": 3.319903968789857,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.408645880403675e-05,
+      "loss": 0.1251,
+      "step": 11065
+    },
+    {
+      "epoch": 3.3202040663215544,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.407139629462269e-05,
+      "loss": 0.1043,
+      "step": 11066
+    },
+    {
+      "epoch": 3.3205041638532524,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.4056333785208614e-05,
+      "loss": 0.1176,
+      "step": 11067
+    },
+    {
+      "epoch": 3.32080426138495,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.4041271275794554e-05,
+      "loss": 0.1015,
+      "step": 11068
+    },
+    {
+      "epoch": 3.321104358916648,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.402620876638048e-05,
+      "loss": 0.1259,
+      "step": 11069
+    },
+    {
+      "epoch": 3.3214044564483456,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.4011146256966414e-05,
+      "loss": 0.0795,
+      "step": 11070
+    },
+    {
+      "epoch": 3.3217045539800436,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.399608374755234e-05,
+      "loss": 0.1275,
+      "step": 11071
+    },
+    {
+      "epoch": 3.322004651511741,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.3981021238138274e-05,
+      "loss": 0.0952,
+      "step": 11072
+    },
+    {
+      "epoch": 3.322304749043439,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.396595872872421e-05,
+      "loss": 0.1137,
+      "step": 11073
+    },
+    {
+      "epoch": 3.3226048465751368,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.395089621931014e-05,
+      "loss": 0.1121,
+      "step": 11074
+    },
+    {
+      "epoch": 3.322904944106835,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.393583370989607e-05,
+      "loss": 0.0971,
+      "step": 11075
+    },
+    {
+      "epoch": 3.3232050416385324,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.3920771200482e-05,
+      "loss": 0.1413,
+      "step": 11076
+    },
+    {
+      "epoch": 3.3235051391702304,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.3905708691067935e-05,
+      "loss": 0.1275,
+      "step": 11077
+    },
+    {
+      "epoch": 3.323805236701928,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.389064618165387e-05,
+      "loss": 0.1133,
+      "step": 11078
+    },
+    {
+      "epoch": 3.324105334233626,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.3875583672239796e-05,
+      "loss": 0.1075,
+      "step": 11079
+    },
+    {
+      "epoch": 3.3244054317653235,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.386052116282572e-05,
+      "loss": 0.0857,
+      "step": 11080
+    },
+    {
+      "epoch": 3.3247055292970216,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.384545865341166e-05,
+      "loss": 0.0998,
+      "step": 11081
+    },
+    {
+      "epoch": 3.325005626828719,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.383039614399759e-05,
+      "loss": 0.1132,
+      "step": 11082
+    },
+    {
+      "epoch": 3.325305724360417,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.381533363458352e-05,
+      "loss": 0.091,
+      "step": 11083
+    },
+    {
+      "epoch": 3.3256058218921147,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.380027112516945e-05,
+      "loss": 0.0922,
+      "step": 11084
+    },
+    {
+      "epoch": 3.3259059194238128,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.378520861575539e-05,
+      "loss": 0.0953,
+      "step": 11085
+    },
+    {
+      "epoch": 3.3262060169555108,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.377014610634132e-05,
+      "loss": 0.0959,
+      "step": 11086
+    },
+    {
+      "epoch": 3.3265061144872083,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.375508359692725e-05,
+      "loss": 0.096,
+      "step": 11087
+    },
+    {
+      "epoch": 3.326806212018906,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.374002108751318e-05,
+      "loss": 0.0938,
+      "step": 11088
+    },
+    {
+      "epoch": 3.327106309550604,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.372495857809911e-05,
+      "loss": 0.0966,
+      "step": 11089
+    },
+    {
+      "epoch": 3.327406407082302,
+      "grad_norm": 0.625,
+      "learning_rate": 3.3709896068685044e-05,
+      "loss": 0.0777,
+      "step": 11090
+    },
+    {
+      "epoch": 3.3277065046139995,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.369483355927098e-05,
+      "loss": 0.1007,
+      "step": 11091
+    },
+    {
+      "epoch": 3.3280066021456975,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.3679771049856904e-05,
+      "loss": 0.1131,
+      "step": 11092
+    },
+    {
+      "epoch": 3.328306699677395,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.366470854044284e-05,
+      "loss": 0.0914,
+      "step": 11093
+    },
+    {
+      "epoch": 3.328606797209093,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.364964603102877e-05,
+      "loss": 0.1152,
+      "step": 11094
+    },
+    {
+      "epoch": 3.3289068947407907,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.3634583521614705e-05,
+      "loss": 0.0947,
+      "step": 11095
+    },
+    {
+      "epoch": 3.3292069922724887,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.361952101220063e-05,
+      "loss": 0.0992,
+      "step": 11096
+    },
+    {
+      "epoch": 3.3295070898041863,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.3604458502786565e-05,
+      "loss": 0.0985,
+      "step": 11097
+    },
+    {
+      "epoch": 3.3298071873358843,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.35893959933725e-05,
+      "loss": 0.1045,
+      "step": 11098
+    },
+    {
+      "epoch": 3.330107284867582,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.357433348395843e-05,
+      "loss": 0.111,
+      "step": 11099
+    },
+    {
+      "epoch": 3.33040738239928,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.355927097454436e-05,
+      "loss": 0.1114,
+      "step": 11100
+    },
+    {
+      "epoch": 3.3307074799309775,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.354420846513029e-05,
+      "loss": 0.0917,
+      "step": 11101
+    },
+    {
+      "epoch": 3.3310075774626755,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.3529145955716226e-05,
+      "loss": 0.1124,
+      "step": 11102
+    },
+    {
+      "epoch": 3.331307674994373,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.351408344630216e-05,
+      "loss": 0.0756,
+      "step": 11103
+    },
+    {
+      "epoch": 3.331607772526071,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.3499020936888086e-05,
+      "loss": 0.0949,
+      "step": 11104
+    },
+    {
+      "epoch": 3.3319078700577687,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.348395842747402e-05,
+      "loss": 0.0994,
+      "step": 11105
+    },
+    {
+      "epoch": 3.3322079675894667,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.3468895918059953e-05,
+      "loss": 0.1177,
+      "step": 11106
+    },
+    {
+      "epoch": 3.3325080651211643,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.345383340864588e-05,
+      "loss": 0.1394,
+      "step": 11107
+    },
+    {
+      "epoch": 3.3328081626528623,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.3438770899231814e-05,
+      "loss": 0.1214,
+      "step": 11108
+    },
+    {
+      "epoch": 3.33310826018456,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.342370838981774e-05,
+      "loss": 0.1105,
+      "step": 11109
+    },
+    {
+      "epoch": 3.333408357716258,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.3408645880403674e-05,
+      "loss": 0.0957,
+      "step": 11110
+    },
+    {
+      "epoch": 3.3337084552479554,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.339358337098961e-05,
+      "loss": 0.1107,
+      "step": 11111
+    },
+    {
+      "epoch": 3.3340085527796535,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.337852086157554e-05,
+      "loss": 0.0949,
+      "step": 11112
+    },
+    {
+      "epoch": 3.334308650311351,
+      "grad_norm": 0.875,
+      "learning_rate": 3.336345835216147e-05,
+      "loss": 0.1163,
+      "step": 11113
+    },
+    {
+      "epoch": 3.334608747843049,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.33483958427474e-05,
+      "loss": 0.1261,
+      "step": 11114
+    },
+    {
+      "epoch": 3.3349088453747466,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.1185,
+      "step": 11115
+    },
+    {
+      "epoch": 3.3352089429064447,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.331827082391927e-05,
+      "loss": 0.1088,
+      "step": 11116
+    },
+    {
+      "epoch": 3.3355090404381422,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.3303208314505195e-05,
+      "loss": 0.0922,
+      "step": 11117
+    },
+    {
+      "epoch": 3.3358091379698402,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.328814580509113e-05,
+      "loss": 0.0941,
+      "step": 11118
+    },
+    {
+      "epoch": 3.336109235501538,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.327308329567706e-05,
+      "loss": 0.1228,
+      "step": 11119
+    },
+    {
+      "epoch": 3.336409333033236,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.3258020786262996e-05,
+      "loss": 0.1014,
+      "step": 11120
+    },
+    {
+      "epoch": 3.3367094305649334,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.324295827684892e-05,
+      "loss": 0.1245,
+      "step": 11121
+    },
+    {
+      "epoch": 3.3370095280966314,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.3227895767434856e-05,
+      "loss": 0.1388,
+      "step": 11122
+    },
+    {
+      "epoch": 3.337309625628329,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.321283325802079e-05,
+      "loss": 0.122,
+      "step": 11123
+    },
+    {
+      "epoch": 3.337609723160027,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.319777074860672e-05,
+      "loss": 0.0872,
+      "step": 11124
+    },
+    {
+      "epoch": 3.337909820691725,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.318270823919265e-05,
+      "loss": 0.1241,
+      "step": 11125
+    },
+    {
+      "epoch": 3.3382099182234226,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.3167645729778583e-05,
+      "loss": 0.1084,
+      "step": 11126
+    },
+    {
+      "epoch": 3.33851001575512,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.315258322036452e-05,
+      "loss": 0.123,
+      "step": 11127
+    },
+    {
+      "epoch": 3.338810113286818,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.313752071095045e-05,
+      "loss": 0.0856,
+      "step": 11128
+    },
+    {
+      "epoch": 3.3391102108185162,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.312245820153638e-05,
+      "loss": 0.1127,
+      "step": 11129
+    },
+    {
+      "epoch": 3.339410308350214,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.310739569212231e-05,
+      "loss": 0.1264,
+      "step": 11130
+    },
+    {
+      "epoch": 3.339710405881912,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.309233318270824e-05,
+      "loss": 0.1278,
+      "step": 11131
+    },
+    {
+      "epoch": 3.3400105034136094,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.307727067329418e-05,
+      "loss": 0.1185,
+      "step": 11132
+    },
+    {
+      "epoch": 3.3403106009453074,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.3062208163880105e-05,
+      "loss": 0.0856,
+      "step": 11133
+    },
+    {
+      "epoch": 3.340610698477005,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.304714565446603e-05,
+      "loss": 0.1015,
+      "step": 11134
+    },
+    {
+      "epoch": 3.340910796008703,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.3032083145051965e-05,
+      "loss": 0.1006,
+      "step": 11135
+    },
+    {
+      "epoch": 3.3412108935404006,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.30170206356379e-05,
+      "loss": 0.1099,
+      "step": 11136
+    },
+    {
+      "epoch": 3.3415109910720986,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.300195812622383e-05,
+      "loss": 0.0921,
+      "step": 11137
+    },
+    {
+      "epoch": 3.341811088603796,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.298689561680976e-05,
+      "loss": 0.1491,
+      "step": 11138
+    },
+    {
+      "epoch": 3.342111186135494,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.297183310739569e-05,
+      "loss": 0.1448,
+      "step": 11139
+    },
+    {
+      "epoch": 3.3424112836671918,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.2956770597981626e-05,
+      "loss": 0.1127,
+      "step": 11140
+    },
+    {
+      "epoch": 3.3427113811988898,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.294170808856756e-05,
+      "loss": 0.1234,
+      "step": 11141
+    },
+    {
+      "epoch": 3.3430114787305873,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.2926645579153486e-05,
+      "loss": 0.0904,
+      "step": 11142
+    },
+    {
+      "epoch": 3.3433115762622854,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.291158306973942e-05,
+      "loss": 0.0936,
+      "step": 11143
+    },
+    {
+      "epoch": 3.343611673793983,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.289652056032535e-05,
+      "loss": 0.1045,
+      "step": 11144
+    },
+    {
+      "epoch": 3.343911771325681,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.2881458050911287e-05,
+      "loss": 0.1048,
+      "step": 11145
+    },
+    {
+      "epoch": 3.3442118688573785,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.286639554149721e-05,
+      "loss": 0.1026,
+      "step": 11146
+    },
+    {
+      "epoch": 3.3445119663890766,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.285133303208315e-05,
+      "loss": 0.107,
+      "step": 11147
+    },
+    {
+      "epoch": 3.344812063920774,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.2836270522669074e-05,
+      "loss": 0.085,
+      "step": 11148
+    },
+    {
+      "epoch": 3.345112161452472,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.2821208013255014e-05,
+      "loss": 0.103,
+      "step": 11149
+    },
+    {
+      "epoch": 3.3454122589841697,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.280614550384094e-05,
+      "loss": 0.1002,
+      "step": 11150
+    },
+    {
+      "epoch": 3.3457123565158677,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.2791082994426874e-05,
+      "loss": 0.0815,
+      "step": 11151
+    },
+    {
+      "epoch": 3.3460124540475653,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.27760204850128e-05,
+      "loss": 0.1029,
+      "step": 11152
+    },
+    {
+      "epoch": 3.3463125515792633,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.276095797559874e-05,
+      "loss": 0.0997,
+      "step": 11153
+    },
+    {
+      "epoch": 3.346612649110961,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.274589546618467e-05,
+      "loss": 0.0768,
+      "step": 11154
+    },
+    {
+      "epoch": 3.346912746642659,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.27308329567706e-05,
+      "loss": 0.1234,
+      "step": 11155
+    },
+    {
+      "epoch": 3.3472128441743565,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.271577044735653e-05,
+      "loss": 0.095,
+      "step": 11156
+    },
+    {
+      "epoch": 3.3475129417060545,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.270070793794247e-05,
+      "loss": 0.1019,
+      "step": 11157
+    },
+    {
+      "epoch": 3.347813039237752,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.2685645428528395e-05,
+      "loss": 0.087,
+      "step": 11158
+    },
+    {
+      "epoch": 3.34811313676945,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.267058291911433e-05,
+      "loss": 0.1205,
+      "step": 11159
+    },
+    {
+      "epoch": 3.348413234301148,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.2655520409700256e-05,
+      "loss": 0.1167,
+      "step": 11160
+    },
+    {
+      "epoch": 3.3487133318328457,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.264045790028619e-05,
+      "loss": 0.1001,
+      "step": 11161
+    },
+    {
+      "epoch": 3.3490134293645433,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.262539539087212e-05,
+      "loss": 0.1249,
+      "step": 11162
+    },
+    {
+      "epoch": 3.3493135268962413,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.261033288145805e-05,
+      "loss": 0.1011,
+      "step": 11163
+    },
+    {
+      "epoch": 3.3496136244279393,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.259527037204398e-05,
+      "loss": 0.1081,
+      "step": 11164
+    },
+    {
+      "epoch": 3.349913721959637,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.2580207862629917e-05,
+      "loss": 0.1096,
+      "step": 11165
+    },
+    {
+      "epoch": 3.3502138194913345,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.256514535321585e-05,
+      "loss": 0.1285,
+      "step": 11166
+    },
+    {
+      "epoch": 3.3505139170230325,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.255008284380178e-05,
+      "loss": 0.111,
+      "step": 11167
+    },
+    {
+      "epoch": 3.3508140145547305,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.253502033438771e-05,
+      "loss": 0.1175,
+      "step": 11168
+    },
+    {
+      "epoch": 3.351114112086428,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.251995782497364e-05,
+      "loss": 0.0748,
+      "step": 11169
+    },
+    {
+      "epoch": 3.351414209618126,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.250489531555958e-05,
+      "loss": 0.1218,
+      "step": 11170
+    },
+    {
+      "epoch": 3.3517143071498237,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.2489832806145504e-05,
+      "loss": 0.0986,
+      "step": 11171
+    },
+    {
+      "epoch": 3.3520144046815217,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.247477029673144e-05,
+      "loss": 0.0804,
+      "step": 11172
+    },
+    {
+      "epoch": 3.3523145022132192,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.2459707787317364e-05,
+      "loss": 0.1194,
+      "step": 11173
+    },
+    {
+      "epoch": 3.3526145997449173,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.2444645277903305e-05,
+      "loss": 0.0875,
+      "step": 11174
+    },
+    {
+      "epoch": 3.352914697276615,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.242958276848923e-05,
+      "loss": 0.1517,
+      "step": 11175
+    },
+    {
+      "epoch": 3.353214794808313,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.2414520259075165e-05,
+      "loss": 0.0979,
+      "step": 11176
+    },
+    {
+      "epoch": 3.3535148923400104,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.239945774966109e-05,
+      "loss": 0.0958,
+      "step": 11177
+    },
+    {
+      "epoch": 3.3538149898717085,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.2384395240247025e-05,
+      "loss": 0.0942,
+      "step": 11178
+    },
+    {
+      "epoch": 3.354115087403406,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.236933273083296e-05,
+      "loss": 0.1243,
+      "step": 11179
+    },
+    {
+      "epoch": 3.354415184935104,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.235427022141889e-05,
+      "loss": 0.1053,
+      "step": 11180
+    },
+    {
+      "epoch": 3.3547152824668016,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.233920771200482e-05,
+      "loss": 0.1161,
+      "step": 11181
+    },
+    {
+      "epoch": 3.3550153799984996,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.232414520259075e-05,
+      "loss": 0.1211,
+      "step": 11182
+    },
+    {
+      "epoch": 3.355315477530197,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.2309082693176686e-05,
+      "loss": 0.0943,
+      "step": 11183
+    },
+    {
+      "epoch": 3.3556155750618952,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.229402018376262e-05,
+      "loss": 0.1197,
+      "step": 11184
+    },
+    {
+      "epoch": 3.355915672593593,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.2278957674348546e-05,
+      "loss": 0.1021,
+      "step": 11185
+    },
+    {
+      "epoch": 3.356215770125291,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.226389516493448e-05,
+      "loss": 0.1202,
+      "step": 11186
+    },
+    {
+      "epoch": 3.3565158676569884,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.2248832655520414e-05,
+      "loss": 0.1109,
+      "step": 11187
+    },
+    {
+      "epoch": 3.3568159651886864,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.223377014610634e-05,
+      "loss": 0.1131,
+      "step": 11188
+    },
+    {
+      "epoch": 3.357116062720384,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.2218707636692274e-05,
+      "loss": 0.0922,
+      "step": 11189
+    },
+    {
+      "epoch": 3.357416160252082,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.22036451272782e-05,
+      "loss": 0.1267,
+      "step": 11190
+    },
+    {
+      "epoch": 3.3577162577837796,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.218858261786414e-05,
+      "loss": 0.0921,
+      "step": 11191
+    },
+    {
+      "epoch": 3.3580163553154776,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.217352010845007e-05,
+      "loss": 0.1,
+      "step": 11192
+    },
+    {
+      "epoch": 3.358316452847175,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.2158457599036e-05,
+      "loss": 0.092,
+      "step": 11193
+    },
+    {
+      "epoch": 3.358616550378873,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.214339508962193e-05,
+      "loss": 0.0966,
+      "step": 11194
+    },
+    {
+      "epoch": 3.3589166479105708,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.212833258020787e-05,
+      "loss": 0.1015,
+      "step": 11195
+    },
+    {
+      "epoch": 3.359216745442269,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.2113270070793795e-05,
+      "loss": 0.1134,
+      "step": 11196
+    },
+    {
+      "epoch": 3.3595168429739664,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.209820756137973e-05,
+      "loss": 0.1096,
+      "step": 11197
+    },
+    {
+      "epoch": 3.3598169405056644,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.2083145051965655e-05,
+      "loss": 0.0861,
+      "step": 11198
+    },
+    {
+      "epoch": 3.3601170380373624,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.206808254255159e-05,
+      "loss": 0.0957,
+      "step": 11199
+    },
+    {
+      "epoch": 3.36041713556906,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.205302003313752e-05,
+      "loss": 0.0965,
+      "step": 11200
+    },
+    {
+      "epoch": 3.3607172331007575,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.2037957523723456e-05,
+      "loss": 0.0969,
+      "step": 11201
+    },
+    {
+      "epoch": 3.3610173306324556,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.202289501430938e-05,
+      "loss": 0.1068,
+      "step": 11202
+    },
+    {
+      "epoch": 3.3613174281641536,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.2007832504895316e-05,
+      "loss": 0.115,
+      "step": 11203
+    },
+    {
+      "epoch": 3.361617525695851,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.199276999548125e-05,
+      "loss": 0.0962,
+      "step": 11204
+    },
+    {
+      "epoch": 3.3619176232275487,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.197770748606718e-05,
+      "loss": 0.108,
+      "step": 11205
+    },
+    {
+      "epoch": 3.3622177207592467,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.196264497665311e-05,
+      "loss": 0.1277,
+      "step": 11206
+    },
+    {
+      "epoch": 3.3625178182909448,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.1947582467239043e-05,
+      "loss": 0.103,
+      "step": 11207
+    },
+    {
+      "epoch": 3.3628179158226423,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.193251995782498e-05,
+      "loss": 0.1003,
+      "step": 11208
+    },
+    {
+      "epoch": 3.3631180133543404,
+      "grad_norm": 0.875,
+      "learning_rate": 3.191745744841091e-05,
+      "loss": 0.12,
+      "step": 11209
+    },
+    {
+      "epoch": 3.363418110886038,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.190239493899684e-05,
+      "loss": 0.0993,
+      "step": 11210
+    },
+    {
+      "epoch": 3.363718208417736,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.188733242958277e-05,
+      "loss": 0.0759,
+      "step": 11211
+    },
+    {
+      "epoch": 3.3640183059494335,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.1872269920168704e-05,
+      "loss": 0.1093,
+      "step": 11212
+    },
+    {
+      "epoch": 3.3643184034811315,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.185720741075464e-05,
+      "loss": 0.1108,
+      "step": 11213
+    },
+    {
+      "epoch": 3.364618501012829,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.1842144901340565e-05,
+      "loss": 0.1023,
+      "step": 11214
+    },
+    {
+      "epoch": 3.364918598544527,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.182708239192649e-05,
+      "loss": 0.1228,
+      "step": 11215
+    },
+    {
+      "epoch": 3.3652186960762247,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.181201988251243e-05,
+      "loss": 0.1095,
+      "step": 11216
+    },
+    {
+      "epoch": 3.3655187936079227,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.179695737309836e-05,
+      "loss": 0.1046,
+      "step": 11217
+    },
+    {
+      "epoch": 3.3658188911396203,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.178189486368429e-05,
+      "loss": 0.1099,
+      "step": 11218
+    },
+    {
+      "epoch": 3.3661189886713183,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.176683235427022e-05,
+      "loss": 0.1226,
+      "step": 11219
+    },
+    {
+      "epoch": 3.366419086203016,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.175176984485615e-05,
+      "loss": 0.1066,
+      "step": 11220
+    },
+    {
+      "epoch": 3.366719183734714,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.1736707335442086e-05,
+      "loss": 0.1444,
+      "step": 11221
+    },
+    {
+      "epoch": 3.3670192812664115,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.172164482602802e-05,
+      "loss": 0.1219,
+      "step": 11222
+    },
+    {
+      "epoch": 3.3673193787981095,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.1706582316613946e-05,
+      "loss": 0.1302,
+      "step": 11223
+    },
+    {
+      "epoch": 3.367619476329807,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.169151980719988e-05,
+      "loss": 0.1015,
+      "step": 11224
+    },
+    {
+      "epoch": 3.367919573861505,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.167645729778581e-05,
+      "loss": 0.1096,
+      "step": 11225
+    },
+    {
+      "epoch": 3.3682196713932027,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.166139478837175e-05,
+      "loss": 0.1039,
+      "step": 11226
+    },
+    {
+      "epoch": 3.3685197689249007,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.164633227895767e-05,
+      "loss": 0.1065,
+      "step": 11227
+    },
+    {
+      "epoch": 3.3688198664565983,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.163126976954361e-05,
+      "loss": 0.0889,
+      "step": 11228
+    },
+    {
+      "epoch": 3.3691199639882963,
+      "grad_norm": 0.75,
+      "learning_rate": 3.161620726012954e-05,
+      "loss": 0.0924,
+      "step": 11229
+    },
+    {
+      "epoch": 3.369420061519994,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.1601144750715474e-05,
+      "loss": 0.1079,
+      "step": 11230
+    },
+    {
+      "epoch": 3.369720159051692,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.15860822413014e-05,
+      "loss": 0.0917,
+      "step": 11231
+    },
+    {
+      "epoch": 3.3700202565833894,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.1571019731887334e-05,
+      "loss": 0.0954,
+      "step": 11232
+    },
+    {
+      "epoch": 3.3703203541150875,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.155595722247327e-05,
+      "loss": 0.1237,
+      "step": 11233
+    },
+    {
+      "epoch": 3.370620451646785,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.15408947130592e-05,
+      "loss": 0.1288,
+      "step": 11234
+    },
+    {
+      "epoch": 3.370920549178483,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.152583220364513e-05,
+      "loss": 0.1036,
+      "step": 11235
+    },
+    {
+      "epoch": 3.3712206467101806,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.151076969423106e-05,
+      "loss": 0.0754,
+      "step": 11236
+    },
+    {
+      "epoch": 3.3715207442418786,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.149570718481699e-05,
+      "loss": 0.1127,
+      "step": 11237
+    },
+    {
+      "epoch": 3.3718208417735767,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.148064467540293e-05,
+      "loss": 0.0918,
+      "step": 11238
+    },
+    {
+      "epoch": 3.3721209393052742,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.1465582165988855e-05,
+      "loss": 0.1039,
+      "step": 11239
+    },
+    {
+      "epoch": 3.372421036836972,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.145051965657479e-05,
+      "loss": 0.1179,
+      "step": 11240
+    },
+    {
+      "epoch": 3.37272113436867,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.1435457147160716e-05,
+      "loss": 0.1023,
+      "step": 11241
+    },
+    {
+      "epoch": 3.373021231900368,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.142039463774665e-05,
+      "loss": 0.1034,
+      "step": 11242
+    },
+    {
+      "epoch": 3.3733213294320654,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.140533212833258e-05,
+      "loss": 0.104,
+      "step": 11243
+    },
+    {
+      "epoch": 3.373621426963763,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.139026961891851e-05,
+      "loss": 0.0923,
+      "step": 11244
+    },
+    {
+      "epoch": 3.373921524495461,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.137520710950444e-05,
+      "loss": 0.1023,
+      "step": 11245
+    },
+    {
+      "epoch": 3.374221622027159,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.1360144600090377e-05,
+      "loss": 0.1099,
+      "step": 11246
+    },
+    {
+      "epoch": 3.3745217195588566,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.134508209067631e-05,
+      "loss": 0.1071,
+      "step": 11247
+    },
+    {
+      "epoch": 3.3748218170905546,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.133001958126224e-05,
+      "loss": 0.0781,
+      "step": 11248
+    },
+    {
+      "epoch": 3.375121914622252,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.131495707184817e-05,
+      "loss": 0.1082,
+      "step": 11249
+    },
+    {
+      "epoch": 3.37542201215395,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.1299894562434104e-05,
+      "loss": 0.1019,
+      "step": 11250
+    },
+    {
+      "epoch": 3.375722109685648,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.128483205302004e-05,
+      "loss": 0.0888,
+      "step": 11251
+    },
+    {
+      "epoch": 3.376022207217346,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.1269769543605964e-05,
+      "loss": 0.1183,
+      "step": 11252
+    },
+    {
+      "epoch": 3.3763223047490434,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.12547070341919e-05,
+      "loss": 0.112,
+      "step": 11253
+    },
+    {
+      "epoch": 3.3766224022807414,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.123964452477783e-05,
+      "loss": 0.1016,
+      "step": 11254
+    },
+    {
+      "epoch": 3.376922499812439,
+      "grad_norm": 0.875,
+      "learning_rate": 3.1224582015363765e-05,
+      "loss": 0.1591,
+      "step": 11255
+    },
+    {
+      "epoch": 3.377222597344137,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.120951950594969e-05,
+      "loss": 0.1166,
+      "step": 11256
+    },
+    {
+      "epoch": 3.3775226948758346,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.1194456996535625e-05,
+      "loss": 0.0975,
+      "step": 11257
+    },
+    {
+      "epoch": 3.3778227924075326,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.117939448712155e-05,
+      "loss": 0.1057,
+      "step": 11258
+    },
+    {
+      "epoch": 3.37812288993923,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.116433197770749e-05,
+      "loss": 0.1039,
+      "step": 11259
+    },
+    {
+      "epoch": 3.378422987470928,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.114926946829342e-05,
+      "loss": 0.1008,
+      "step": 11260
+    },
+    {
+      "epoch": 3.3787230850026257,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.113420695887935e-05,
+      "loss": 0.0968,
+      "step": 11261
+    },
+    {
+      "epoch": 3.3790231825343238,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.111914444946528e-05,
+      "loss": 0.1037,
+      "step": 11262
+    },
+    {
+      "epoch": 3.3793232800660213,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.110408194005122e-05,
+      "loss": 0.0898,
+      "step": 11263
+    },
+    {
+      "epoch": 3.3796233775977194,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.1089019430637146e-05,
+      "loss": 0.0955,
+      "step": 11264
+    },
+    {
+      "epoch": 3.379923475129417,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.107395692122308e-05,
+      "loss": 0.1198,
+      "step": 11265
+    },
+    {
+      "epoch": 3.380223572661115,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.1058894411809006e-05,
+      "loss": 0.1075,
+      "step": 11266
+    },
+    {
+      "epoch": 3.3805236701928125,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.104383190239494e-05,
+      "loss": 0.1126,
+      "step": 11267
+    },
+    {
+      "epoch": 3.3808237677245105,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.1028769392980874e-05,
+      "loss": 0.1114,
+      "step": 11268
+    },
+    {
+      "epoch": 3.381123865256208,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.10137068835668e-05,
+      "loss": 0.1068,
+      "step": 11269
+    },
+    {
+      "epoch": 3.381423962787906,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.0998644374152734e-05,
+      "loss": 0.1262,
+      "step": 11270
+    },
+    {
+      "epoch": 3.3817240603196037,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.098358186473867e-05,
+      "loss": 0.1011,
+      "step": 11271
+    },
+    {
+      "epoch": 3.3820241578513017,
+      "grad_norm": 0.75,
+      "learning_rate": 3.09685193553246e-05,
+      "loss": 0.0975,
+      "step": 11272
+    },
+    {
+      "epoch": 3.3823242553829993,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.095345684591053e-05,
+      "loss": 0.1007,
+      "step": 11273
+    },
+    {
+      "epoch": 3.3826243529146973,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.093839433649646e-05,
+      "loss": 0.1276,
+      "step": 11274
+    },
+    {
+      "epoch": 3.382924450446395,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.0923331827082395e-05,
+      "loss": 0.0931,
+      "step": 11275
+    },
+    {
+      "epoch": 3.383224547978093,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.090826931766833e-05,
+      "loss": 0.1077,
+      "step": 11276
+    },
+    {
+      "epoch": 3.383524645509791,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.0893206808254255e-05,
+      "loss": 0.0825,
+      "step": 11277
+    },
+    {
+      "epoch": 3.3838247430414885,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.087814429884019e-05,
+      "loss": 0.1095,
+      "step": 11278
+    },
+    {
+      "epoch": 3.384124840573186,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.0863081789426115e-05,
+      "loss": 0.0966,
+      "step": 11279
+    },
+    {
+      "epoch": 3.384424938104884,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.0848019280012056e-05,
+      "loss": 0.0802,
+      "step": 11280
+    },
+    {
+      "epoch": 3.384725035636582,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.083295677059798e-05,
+      "loss": 0.0891,
+      "step": 11281
+    },
+    {
+      "epoch": 3.3850251331682797,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.0817894261183916e-05,
+      "loss": 0.1243,
+      "step": 11282
+    },
+    {
+      "epoch": 3.3853252306999773,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.080283175176984e-05,
+      "loss": 0.1053,
+      "step": 11283
+    },
+    {
+      "epoch": 3.3856253282316753,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.078776924235578e-05,
+      "loss": 0.1063,
+      "step": 11284
+    },
+    {
+      "epoch": 3.3859254257633733,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.077270673294171e-05,
+      "loss": 0.0932,
+      "step": 11285
+    },
+    {
+      "epoch": 3.386225523295071,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.075764422352764e-05,
+      "loss": 0.1062,
+      "step": 11286
+    },
+    {
+      "epoch": 3.386525620826769,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.074258171411357e-05,
+      "loss": 0.1149,
+      "step": 11287
+    },
+    {
+      "epoch": 3.3868257183584665,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.0727519204699503e-05,
+      "loss": 0.1062,
+      "step": 11288
+    },
+    {
+      "epoch": 3.3871258158901645,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.071245669528544e-05,
+      "loss": 0.1446,
+      "step": 11289
+    },
+    {
+      "epoch": 3.387425913421862,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.069739418587137e-05,
+      "loss": 0.1377,
+      "step": 11290
+    },
+    {
+      "epoch": 3.38772601095356,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.06823316764573e-05,
+      "loss": 0.1217,
+      "step": 11291
+    },
+    {
+      "epoch": 3.3880261084852576,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.066726916704323e-05,
+      "loss": 0.0894,
+      "step": 11292
+    },
+    {
+      "epoch": 3.3883262060169557,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.0652206657629164e-05,
+      "loss": 0.0971,
+      "step": 11293
+    },
+    {
+      "epoch": 3.3886263035486532,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.06371441482151e-05,
+      "loss": 0.1086,
+      "step": 11294
+    },
+    {
+      "epoch": 3.3889264010803513,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.0622081638801025e-05,
+      "loss": 0.1104,
+      "step": 11295
+    },
+    {
+      "epoch": 3.389226498612049,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.060701912938695e-05,
+      "loss": 0.1048,
+      "step": 11296
+    },
+    {
+      "epoch": 3.389526596143747,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.059195661997289e-05,
+      "loss": 0.1224,
+      "step": 11297
+    },
+    {
+      "epoch": 3.3898266936754444,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.057689411055882e-05,
+      "loss": 0.0855,
+      "step": 11298
+    },
+    {
+      "epoch": 3.3901267912071424,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.056183160114475e-05,
+      "loss": 0.1911,
+      "step": 11299
+    },
+    {
+      "epoch": 3.39042688873884,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.054676909173068e-05,
+      "loss": 0.0995,
+      "step": 11300
+    },
+    {
+      "epoch": 3.390726986270538,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.053170658231662e-05,
+      "loss": 0.0874,
+      "step": 11301
+    },
+    {
+      "epoch": 3.3910270838022356,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.0516644072902546e-05,
+      "loss": 0.1126,
+      "step": 11302
+    },
+    {
+      "epoch": 3.3913271813339336,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.050158156348848e-05,
+      "loss": 0.1039,
+      "step": 11303
+    },
+    {
+      "epoch": 3.391627278865631,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.048651905407441e-05,
+      "loss": 0.1058,
+      "step": 11304
+    },
+    {
+      "epoch": 3.391927376397329,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.0471456544660343e-05,
+      "loss": 0.1246,
+      "step": 11305
+    },
+    {
+      "epoch": 3.392227473929027,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.0456394035246273e-05,
+      "loss": 0.0936,
+      "step": 11306
+    },
+    {
+      "epoch": 3.392527571460725,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.0441331525832207e-05,
+      "loss": 0.1121,
+      "step": 11307
+    },
+    {
+      "epoch": 3.3928276689924224,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.0426269016418137e-05,
+      "loss": 0.1011,
+      "step": 11308
+    },
+    {
+      "epoch": 3.3931277665241204,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.041120650700407e-05,
+      "loss": 0.0881,
+      "step": 11309
+    },
+    {
+      "epoch": 3.393427864055818,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.0396143997589997e-05,
+      "loss": 0.1174,
+      "step": 11310
+    },
+    {
+      "epoch": 3.393727961587516,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.0381081488175934e-05,
+      "loss": 0.1282,
+      "step": 11311
+    },
+    {
+      "epoch": 3.3940280591192136,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.036601897876186e-05,
+      "loss": 0.1082,
+      "step": 11312
+    },
+    {
+      "epoch": 3.3943281566509116,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.0350956469347798e-05,
+      "loss": 0.083,
+      "step": 11313
+    },
+    {
+      "epoch": 3.394628254182609,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.0335893959933724e-05,
+      "loss": 0.1152,
+      "step": 11314
+    },
+    {
+      "epoch": 3.394928351714307,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.032083145051966e-05,
+      "loss": 0.1111,
+      "step": 11315
+    },
+    {
+      "epoch": 3.395228449246005,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.0305768941105588e-05,
+      "loss": 0.1144,
+      "step": 11316
+    },
+    {
+      "epoch": 3.3955285467777028,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.0290706431691525e-05,
+      "loss": 0.1249,
+      "step": 11317
+    },
+    {
+      "epoch": 3.3958286443094003,
+      "grad_norm": 0.875,
+      "learning_rate": 3.0275643922277452e-05,
+      "loss": 0.1078,
+      "step": 11318
+    },
+    {
+      "epoch": 3.3961287418410984,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.0260581412863385e-05,
+      "loss": 0.1133,
+      "step": 11319
+    },
+    {
+      "epoch": 3.3964288393727964,
+      "grad_norm": 0.875,
+      "learning_rate": 3.0245518903449315e-05,
+      "loss": 0.1257,
+      "step": 11320
+    },
+    {
+      "epoch": 3.396728936904494,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.023045639403525e-05,
+      "loss": 0.1036,
+      "step": 11321
+    },
+    {
+      "epoch": 3.3970290344361915,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.021539388462118e-05,
+      "loss": 0.0918,
+      "step": 11322
+    },
+    {
+      "epoch": 3.3973291319678895,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.020033137520711e-05,
+      "loss": 0.1,
+      "step": 11323
+    },
+    {
+      "epoch": 3.3976292294995876,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.0185268865793043e-05,
+      "loss": 0.1446,
+      "step": 11324
+    },
+    {
+      "epoch": 3.397929327031285,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.0170206356378973e-05,
+      "loss": 0.0813,
+      "step": 11325
+    },
+    {
+      "epoch": 3.398229424562983,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.0155143846964906e-05,
+      "loss": 0.1152,
+      "step": 11326
+    },
+    {
+      "epoch": 3.3985295220946807,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.0140081337550837e-05,
+      "loss": 0.1014,
+      "step": 11327
+    },
+    {
+      "epoch": 3.3988296196263788,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.012501882813677e-05,
+      "loss": 0.1119,
+      "step": 11328
+    },
+    {
+      "epoch": 3.3991297171580763,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.0109956318722697e-05,
+      "loss": 0.118,
+      "step": 11329
+    },
+    {
+      "epoch": 3.3994298146897743,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.0094893809308634e-05,
+      "loss": 0.1066,
+      "step": 11330
+    },
+    {
+      "epoch": 3.399729912221472,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.007983129989456e-05,
+      "loss": 0.0958,
+      "step": 11331
+    },
+    {
+      "epoch": 3.40003000975317,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.0064768790480497e-05,
+      "loss": 0.1242,
+      "step": 11332
+    },
+    {
+      "epoch": 3.4003301072848675,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.0049706281066424e-05,
+      "loss": 0.1416,
+      "step": 11333
+    },
+    {
+      "epoch": 3.4006302048165655,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.003464377165236e-05,
+      "loss": 0.1104,
+      "step": 11334
+    },
+    {
+      "epoch": 3.400930302348263,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.0019581262238288e-05,
+      "loss": 0.1284,
+      "step": 11335
+    },
+    {
+      "epoch": 3.401230399879961,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.0004518752824225e-05,
+      "loss": 0.0925,
+      "step": 11336
+    },
+    {
+      "epoch": 3.4015304974116587,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.998945624341015e-05,
+      "loss": 0.113,
+      "step": 11337
+    },
+    {
+      "epoch": 3.4018305949433567,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.997439373399609e-05,
+      "loss": 0.1205,
+      "step": 11338
+    },
+    {
+      "epoch": 3.4021306924750543,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.9959331224582015e-05,
+      "loss": 0.0963,
+      "step": 11339
+    },
+    {
+      "epoch": 3.4024307900067523,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.994426871516795e-05,
+      "loss": 0.0936,
+      "step": 11340
+    },
+    {
+      "epoch": 3.40273088753845,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.992920620575388e-05,
+      "loss": 0.1437,
+      "step": 11341
+    },
+    {
+      "epoch": 3.403030985070148,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.9914143696339812e-05,
+      "loss": 0.088,
+      "step": 11342
+    },
+    {
+      "epoch": 3.4033310826018455,
+      "grad_norm": 0.875,
+      "learning_rate": 2.9899081186925743e-05,
+      "loss": 0.1108,
+      "step": 11343
+    },
+    {
+      "epoch": 3.4036311801335435,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.9884018677511676e-05,
+      "loss": 0.0811,
+      "step": 11344
+    },
+    {
+      "epoch": 3.403931277665241,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.9868956168097606e-05,
+      "loss": 0.1047,
+      "step": 11345
+    },
+    {
+      "epoch": 3.404231375196939,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.985389365868354e-05,
+      "loss": 0.1138,
+      "step": 11346
+    },
+    {
+      "epoch": 3.4045314727286367,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.983883114926947e-05,
+      "loss": 0.1011,
+      "step": 11347
+    },
+    {
+      "epoch": 3.4048315702603347,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.9823768639855403e-05,
+      "loss": 0.1192,
+      "step": 11348
+    },
+    {
+      "epoch": 3.4051316677920322,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.9808706130441334e-05,
+      "loss": 0.0941,
+      "step": 11349
+    },
+    {
+      "epoch": 3.4054317653237303,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.979364362102726e-05,
+      "loss": 0.123,
+      "step": 11350
+    },
+    {
+      "epoch": 3.405731862855428,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.9778581111613197e-05,
+      "loss": 0.1341,
+      "step": 11351
+    },
+    {
+      "epoch": 3.406031960387126,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.9763518602199124e-05,
+      "loss": 0.166,
+      "step": 11352
+    },
+    {
+      "epoch": 3.4063320579188234,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.974845609278506e-05,
+      "loss": 0.2405,
+      "step": 11353
+    },
+    {
+      "epoch": 3.4066321554505214,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.9733393583370988e-05,
+      "loss": 0.1076,
+      "step": 11354
+    },
+    {
+      "epoch": 3.4069322529822195,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.9718331073956925e-05,
+      "loss": 0.1246,
+      "step": 11355
+    },
+    {
+      "epoch": 3.407232350513917,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.970326856454285e-05,
+      "loss": 0.1181,
+      "step": 11356
+    },
+    {
+      "epoch": 3.4075324480456146,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.9688206055128788e-05,
+      "loss": 0.0992,
+      "step": 11357
+    },
+    {
+      "epoch": 3.4078325455773126,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.9673143545714715e-05,
+      "loss": 0.2024,
+      "step": 11358
+    },
+    {
+      "epoch": 3.4081326431090107,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.965808103630065e-05,
+      "loss": 0.1042,
+      "step": 11359
+    },
+    {
+      "epoch": 3.4084327406407082,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.964301852688658e-05,
+      "loss": 0.106,
+      "step": 11360
+    },
+    {
+      "epoch": 3.408732838172406,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.9627956017472512e-05,
+      "loss": 0.117,
+      "step": 11361
+    },
+    {
+      "epoch": 3.409032935704104,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.9612893508058442e-05,
+      "loss": 0.1055,
+      "step": 11362
+    },
+    {
+      "epoch": 3.409333033235802,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.9597830998644376e-05,
+      "loss": 0.1,
+      "step": 11363
+    },
+    {
+      "epoch": 3.4096331307674994,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.9582768489230306e-05,
+      "loss": 0.1389,
+      "step": 11364
+    },
+    {
+      "epoch": 3.4099332282991974,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.956770597981624e-05,
+      "loss": 0.1002,
+      "step": 11365
+    },
+    {
+      "epoch": 3.410233325830895,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.955264347040217e-05,
+      "loss": 0.1164,
+      "step": 11366
+    },
+    {
+      "epoch": 3.410533423362593,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9537580960988103e-05,
+      "loss": 0.1073,
+      "step": 11367
+    },
+    {
+      "epoch": 3.4108335208942906,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.9522518451574033e-05,
+      "loss": 0.1329,
+      "step": 11368
+    },
+    {
+      "epoch": 3.4111336184259886,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.9507455942159967e-05,
+      "loss": 0.106,
+      "step": 11369
+    },
+    {
+      "epoch": 3.411433715957686,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.9492393432745897e-05,
+      "loss": 0.0875,
+      "step": 11370
+    },
+    {
+      "epoch": 3.411733813489384,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.947733092333183e-05,
+      "loss": 0.1142,
+      "step": 11371
+    },
+    {
+      "epoch": 3.4120339110210818,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.946226841391776e-05,
+      "loss": 0.0911,
+      "step": 11372
+    },
+    {
+      "epoch": 3.41233400855278,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.9447205904503694e-05,
+      "loss": 0.0943,
+      "step": 11373
+    },
+    {
+      "epoch": 3.4126341060844774,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.9432143395089624e-05,
+      "loss": 0.0932,
+      "step": 11374
+    },
+    {
+      "epoch": 3.4129342036161754,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.9417080885675558e-05,
+      "loss": 0.1123,
+      "step": 11375
+    },
+    {
+      "epoch": 3.413234301147873,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.9402018376261488e-05,
+      "loss": 0.1008,
+      "step": 11376
+    },
+    {
+      "epoch": 3.413534398679571,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.9386955866847415e-05,
+      "loss": 0.1161,
+      "step": 11377
+    },
+    {
+      "epoch": 3.4138344962112686,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9371893357433348e-05,
+      "loss": 0.1011,
+      "step": 11378
+    },
+    {
+      "epoch": 3.4141345937429666,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.935683084801928e-05,
+      "loss": 0.1558,
+      "step": 11379
+    },
+    {
+      "epoch": 3.414434691274664,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.9341768338605212e-05,
+      "loss": 0.1151,
+      "step": 11380
+    },
+    {
+      "epoch": 3.414734788806362,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.9326705829191142e-05,
+      "loss": 0.1148,
+      "step": 11381
+    },
+    {
+      "epoch": 3.4150348863380597,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.9311643319777076e-05,
+      "loss": 0.1096,
+      "step": 11382
+    },
+    {
+      "epoch": 3.4153349838697578,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.9296580810363006e-05,
+      "loss": 0.0899,
+      "step": 11383
+    },
+    {
+      "epoch": 3.4156350814014553,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.928151830094894e-05,
+      "loss": 0.0985,
+      "step": 11384
+    },
+    {
+      "epoch": 3.4159351789331533,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.926645579153487e-05,
+      "loss": 0.0856,
+      "step": 11385
+    },
+    {
+      "epoch": 3.416235276464851,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.9251393282120803e-05,
+      "loss": 0.0957,
+      "step": 11386
+    },
+    {
+      "epoch": 3.416535373996549,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.9236330772706733e-05,
+      "loss": 0.096,
+      "step": 11387
+    },
+    {
+      "epoch": 3.4168354715282465,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.9221268263292667e-05,
+      "loss": 0.1362,
+      "step": 11388
+    },
+    {
+      "epoch": 3.4171355690599445,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.9206205753878597e-05,
+      "loss": 0.0921,
+      "step": 11389
+    },
+    {
+      "epoch": 3.417435666591642,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.919114324446453e-05,
+      "loss": 0.0924,
+      "step": 11390
+    },
+    {
+      "epoch": 3.41773576412334,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.917608073505046e-05,
+      "loss": 0.0926,
+      "step": 11391
+    },
+    {
+      "epoch": 3.4180358616550377,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.9161018225636394e-05,
+      "loss": 0.1168,
+      "step": 11392
+    },
+    {
+      "epoch": 3.4183359591867357,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.9145955716222324e-05,
+      "loss": 0.0932,
+      "step": 11393
+    },
+    {
+      "epoch": 3.4186360567184337,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.9130893206808258e-05,
+      "loss": 0.1413,
+      "step": 11394
+    },
+    {
+      "epoch": 3.4189361542501313,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.9115830697394188e-05,
+      "loss": 0.1013,
+      "step": 11395
+    },
+    {
+      "epoch": 3.419236251781829,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.910076818798012e-05,
+      "loss": 0.1618,
+      "step": 11396
+    },
+    {
+      "epoch": 3.419536349313527,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.908570567856605e-05,
+      "loss": 0.0991,
+      "step": 11397
+    },
+    {
+      "epoch": 3.419836446845225,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.9070643169151985e-05,
+      "loss": 0.093,
+      "step": 11398
+    },
+    {
+      "epoch": 3.4201365443769225,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.9055580659737912e-05,
+      "loss": 0.0645,
+      "step": 11399
+    },
+    {
+      "epoch": 3.42043664190862,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.904051815032385e-05,
+      "loss": 0.1128,
+      "step": 11400
+    },
+    {
+      "epoch": 3.420736739440318,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.9025455640909775e-05,
+      "loss": 0.0941,
+      "step": 11401
+    },
+    {
+      "epoch": 3.421036836972016,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.9010393131495712e-05,
+      "loss": 0.1198,
+      "step": 11402
+    },
+    {
+      "epoch": 3.4213369345037137,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.899533062208164e-05,
+      "loss": 0.1153,
+      "step": 11403
+    },
+    {
+      "epoch": 3.4216370320354117,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.8980268112667576e-05,
+      "loss": 0.1015,
+      "step": 11404
+    },
+    {
+      "epoch": 3.4219371295671093,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.8965205603253503e-05,
+      "loss": 0.0878,
+      "step": 11405
+    },
+    {
+      "epoch": 3.4222372270988073,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.8950143093839433e-05,
+      "loss": 0.1213,
+      "step": 11406
+    },
+    {
+      "epoch": 3.422537324630505,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.8935080584425366e-05,
+      "loss": 0.1038,
+      "step": 11407
+    },
+    {
+      "epoch": 3.422837422162203,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.8920018075011297e-05,
+      "loss": 0.0882,
+      "step": 11408
+    },
+    {
+      "epoch": 3.4231375196939005,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.890495556559723e-05,
+      "loss": 0.0878,
+      "step": 11409
+    },
+    {
+      "epoch": 3.4234376172255985,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.888989305618316e-05,
+      "loss": 0.1033,
+      "step": 11410
+    },
+    {
+      "epoch": 3.423737714757296,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.8874830546769094e-05,
+      "loss": 0.1143,
+      "step": 11411
+    },
+    {
+      "epoch": 3.424037812288994,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.8859768037355024e-05,
+      "loss": 0.112,
+      "step": 11412
+    },
+    {
+      "epoch": 3.4243379098206916,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.8844705527940957e-05,
+      "loss": 0.1339,
+      "step": 11413
+    },
+    {
+      "epoch": 3.4246380073523897,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.8829643018526888e-05,
+      "loss": 0.0989,
+      "step": 11414
+    },
+    {
+      "epoch": 3.4249381048840872,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.881458050911282e-05,
+      "loss": 0.1179,
+      "step": 11415
+    },
+    {
+      "epoch": 3.4252382024157852,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.879951799969875e-05,
+      "loss": 0.1074,
+      "step": 11416
+    },
+    {
+      "epoch": 3.425538299947483,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.8784455490284685e-05,
+      "loss": 0.1079,
+      "step": 11417
+    },
+    {
+      "epoch": 3.425838397479181,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.876939298087061e-05,
+      "loss": 0.0987,
+      "step": 11418
+    },
+    {
+      "epoch": 3.4261384950108784,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.875433047145655e-05,
+      "loss": 0.0976,
+      "step": 11419
+    },
+    {
+      "epoch": 3.4264385925425764,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.8739267962042475e-05,
+      "loss": 0.101,
+      "step": 11420
+    },
+    {
+      "epoch": 3.426738690074274,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.8724205452628412e-05,
+      "loss": 0.1222,
+      "step": 11421
+    },
+    {
+      "epoch": 3.427038787605972,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.870914294321434e-05,
+      "loss": 0.0921,
+      "step": 11422
+    },
+    {
+      "epoch": 3.4273388851376696,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.8694080433800276e-05,
+      "loss": 0.1113,
+      "step": 11423
+    },
+    {
+      "epoch": 3.4276389826693676,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.8679017924386203e-05,
+      "loss": 0.0868,
+      "step": 11424
+    },
+    {
+      "epoch": 3.427939080201065,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.866395541497214e-05,
+      "loss": 0.1059,
+      "step": 11425
+    },
+    {
+      "epoch": 3.428239177732763,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.8648892905558066e-05,
+      "loss": 0.0933,
+      "step": 11426
+    },
+    {
+      "epoch": 3.428539275264461,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.8633830396144e-05,
+      "loss": 0.0951,
+      "step": 11427
+    },
+    {
+      "epoch": 3.428839372796159,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.861876788672993e-05,
+      "loss": 0.0935,
+      "step": 11428
+    },
+    {
+      "epoch": 3.4291394703278564,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.8603705377315863e-05,
+      "loss": 0.107,
+      "step": 11429
+    },
+    {
+      "epoch": 3.4294395678595544,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.8588642867901794e-05,
+      "loss": 0.1408,
+      "step": 11430
+    },
+    {
+      "epoch": 3.429739665391252,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.8573580358487727e-05,
+      "loss": 0.1066,
+      "step": 11431
+    },
+    {
+      "epoch": 3.43003976292295,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.8558517849073657e-05,
+      "loss": 0.113,
+      "step": 11432
+    },
+    {
+      "epoch": 3.430339860454648,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.8543455339659587e-05,
+      "loss": 0.1164,
+      "step": 11433
+    },
+    {
+      "epoch": 3.4306399579863456,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.852839283024552e-05,
+      "loss": 0.1077,
+      "step": 11434
+    },
+    {
+      "epoch": 3.430940055518043,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.851333032083145e-05,
+      "loss": 0.1105,
+      "step": 11435
+    },
+    {
+      "epoch": 3.431240153049741,
+      "grad_norm": 0.875,
+      "learning_rate": 2.8498267811417385e-05,
+      "loss": 0.1073,
+      "step": 11436
+    },
+    {
+      "epoch": 3.431540250581439,
+      "grad_norm": 0.75,
+      "learning_rate": 2.848320530200331e-05,
+      "loss": 0.0931,
+      "step": 11437
+    },
+    {
+      "epoch": 3.4318403481131368,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.8468142792589248e-05,
+      "loss": 0.0994,
+      "step": 11438
+    },
+    {
+      "epoch": 3.4321404456448343,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.8453080283175175e-05,
+      "loss": 0.1116,
+      "step": 11439
+    },
+    {
+      "epoch": 3.4324405431765324,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.8438017773761112e-05,
+      "loss": 0.1056,
+      "step": 11440
+    },
+    {
+      "epoch": 3.4327406407082304,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.842295526434704e-05,
+      "loss": 0.1046,
+      "step": 11441
+    },
+    {
+      "epoch": 3.433040738239928,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.8407892754932976e-05,
+      "loss": 0.1105,
+      "step": 11442
+    },
+    {
+      "epoch": 3.433340835771626,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.8392830245518902e-05,
+      "loss": 0.1106,
+      "step": 11443
+    },
+    {
+      "epoch": 3.4336409333033235,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.837776773610484e-05,
+      "loss": 0.1099,
+      "step": 11444
+    },
+    {
+      "epoch": 3.4339410308350216,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.8362705226690766e-05,
+      "loss": 0.1002,
+      "step": 11445
+    },
+    {
+      "epoch": 3.434241128366719,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.83476427172767e-05,
+      "loss": 0.1052,
+      "step": 11446
+    },
+    {
+      "epoch": 3.434541225898417,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.833258020786263e-05,
+      "loss": 0.1105,
+      "step": 11447
+    },
+    {
+      "epoch": 3.4348413234301147,
+      "grad_norm": 0.875,
+      "learning_rate": 2.8317517698448563e-05,
+      "loss": 0.106,
+      "step": 11448
+    },
+    {
+      "epoch": 3.4351414209618127,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.8302455189034493e-05,
+      "loss": 0.1014,
+      "step": 11449
+    },
+    {
+      "epoch": 3.4354415184935103,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.8287392679620427e-05,
+      "loss": 0.1068,
+      "step": 11450
+    },
+    {
+      "epoch": 3.4357416160252083,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.8272330170206357e-05,
+      "loss": 0.0908,
+      "step": 11451
+    },
+    {
+      "epoch": 3.436041713556906,
+      "grad_norm": 0.875,
+      "learning_rate": 2.825726766079229e-05,
+      "loss": 0.1159,
+      "step": 11452
+    },
+    {
+      "epoch": 3.436341811088604,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.824220515137822e-05,
+      "loss": 0.1186,
+      "step": 11453
+    },
+    {
+      "epoch": 3.4366419086203015,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.8227142641964154e-05,
+      "loss": 0.1159,
+      "step": 11454
+    },
+    {
+      "epoch": 3.4369420061519995,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.8212080132550084e-05,
+      "loss": 0.1112,
+      "step": 11455
+    },
+    {
+      "epoch": 3.437242103683697,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.8197017623136018e-05,
+      "loss": 0.1193,
+      "step": 11456
+    },
+    {
+      "epoch": 3.437542201215395,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.8181955113721948e-05,
+      "loss": 0.1109,
+      "step": 11457
+    },
+    {
+      "epoch": 3.4378422987470927,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.816689260430788e-05,
+      "loss": 0.1296,
+      "step": 11458
+    },
+    {
+      "epoch": 3.4381423962787907,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.8151830094893812e-05,
+      "loss": 0.1212,
+      "step": 11459
+    },
+    {
+      "epoch": 3.4384424938104883,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.813676758547974e-05,
+      "loss": 0.1005,
+      "step": 11460
+    },
+    {
+      "epoch": 3.4387425913421863,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.8121705076065675e-05,
+      "loss": 0.15,
+      "step": 11461
+    },
+    {
+      "epoch": 3.439042688873884,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.8106642566651602e-05,
+      "loss": 0.1053,
+      "step": 11462
+    },
+    {
+      "epoch": 3.439342786405582,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.809158005723754e-05,
+      "loss": 0.0951,
+      "step": 11463
+    },
+    {
+      "epoch": 3.4396428839372795,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.8076517547823466e-05,
+      "loss": 0.0989,
+      "step": 11464
+    },
+    {
+      "epoch": 3.4399429814689775,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.8061455038409403e-05,
+      "loss": 0.0962,
+      "step": 11465
+    },
+    {
+      "epoch": 3.440243079000675,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.804639252899533e-05,
+      "loss": 0.0927,
+      "step": 11466
+    },
+    {
+      "epoch": 3.440543176532373,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.8031330019581263e-05,
+      "loss": 0.0925,
+      "step": 11467
+    },
+    {
+      "epoch": 3.4408432740640706,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.8016267510167193e-05,
+      "loss": 0.1141,
+      "step": 11468
+    },
+    {
+      "epoch": 3.4411433715957687,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.8001205000753127e-05,
+      "loss": 0.1221,
+      "step": 11469
+    },
+    {
+      "epoch": 3.4414434691274662,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.7986142491339057e-05,
+      "loss": 0.1075,
+      "step": 11470
+    },
+    {
+      "epoch": 3.4417435666591643,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.797107998192499e-05,
+      "loss": 0.1023,
+      "step": 11471
+    },
+    {
+      "epoch": 3.4420436641908623,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.795601747251092e-05,
+      "loss": 0.0998,
+      "step": 11472
+    },
+    {
+      "epoch": 3.44234376172256,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.7940954963096854e-05,
+      "loss": 0.0939,
+      "step": 11473
+    },
+    {
+      "epoch": 3.4426438592542574,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.7925892453682784e-05,
+      "loss": 0.0899,
+      "step": 11474
+    },
+    {
+      "epoch": 3.4429439567859554,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.7910829944268718e-05,
+      "loss": 0.1227,
+      "step": 11475
+    },
+    {
+      "epoch": 3.4432440543176535,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.7895767434854648e-05,
+      "loss": 0.1239,
+      "step": 11476
+    },
+    {
+      "epoch": 3.443544151849351,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.788070492544058e-05,
+      "loss": 0.1068,
+      "step": 11477
+    },
+    {
+      "epoch": 3.4438442493810486,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.786564241602651e-05,
+      "loss": 0.1191,
+      "step": 11478
+    },
+    {
+      "epoch": 3.4441443469127466,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.7850579906612445e-05,
+      "loss": 0.0994,
+      "step": 11479
+    },
+    {
+      "epoch": 3.4444444444444446,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.7835517397198375e-05,
+      "loss": 0.0717,
+      "step": 11480
+    },
+    {
+      "epoch": 3.444744541976142,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.782045488778431e-05,
+      "loss": 0.09,
+      "step": 11481
+    },
+    {
+      "epoch": 3.4450446395078402,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.780539237837024e-05,
+      "loss": 0.0752,
+      "step": 11482
+    },
+    {
+      "epoch": 3.445344737039538,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.7790329868956172e-05,
+      "loss": 0.0985,
+      "step": 11483
+    },
+    {
+      "epoch": 3.445644834571236,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.7775267359542102e-05,
+      "loss": 0.1143,
+      "step": 11484
+    },
+    {
+      "epoch": 3.4459449321029334,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.7760204850128036e-05,
+      "loss": 0.1148,
+      "step": 11485
+    },
+    {
+      "epoch": 3.4462450296346314,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.7745142340713963e-05,
+      "loss": 0.0952,
+      "step": 11486
+    },
+    {
+      "epoch": 3.446545127166329,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.7730079831299893e-05,
+      "loss": 0.1052,
+      "step": 11487
+    },
+    {
+      "epoch": 3.446845224698027,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.7715017321885826e-05,
+      "loss": 0.0768,
+      "step": 11488
+    },
+    {
+      "epoch": 3.4471453222297246,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.7699954812471757e-05,
+      "loss": 0.1198,
+      "step": 11489
+    },
+    {
+      "epoch": 3.4474454197614226,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.768489230305769e-05,
+      "loss": 0.0953,
+      "step": 11490
+    },
+    {
+      "epoch": 3.44774551729312,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.766982979364362e-05,
+      "loss": 0.0931,
+      "step": 11491
+    },
+    {
+      "epoch": 3.448045614824818,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7654767284229554e-05,
+      "loss": 0.1108,
+      "step": 11492
+    },
+    {
+      "epoch": 3.4483457123565158,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.7639704774815484e-05,
+      "loss": 0.1105,
+      "step": 11493
+    },
+    {
+      "epoch": 3.448645809888214,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.7624642265401417e-05,
+      "loss": 0.1186,
+      "step": 11494
+    },
+    {
+      "epoch": 3.4489459074199114,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.7609579755987348e-05,
+      "loss": 0.1076,
+      "step": 11495
+    },
+    {
+      "epoch": 3.4492460049516094,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.759451724657328e-05,
+      "loss": 0.1145,
+      "step": 11496
+    },
+    {
+      "epoch": 3.449546102483307,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.757945473715921e-05,
+      "loss": 0.1149,
+      "step": 11497
+    },
+    {
+      "epoch": 3.449846200015005,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7564392227745145e-05,
+      "loss": 0.0877,
+      "step": 11498
+    },
+    {
+      "epoch": 3.4501462975467025,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.7549329718331075e-05,
+      "loss": 0.0943,
+      "step": 11499
+    },
+    {
+      "epoch": 3.4504463950784006,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.753426720891701e-05,
+      "loss": 0.1235,
+      "step": 11500
+    },
+    {
+      "epoch": 3.450746492610098,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.751920469950294e-05,
+      "loss": 0.1415,
+      "step": 11501
+    },
+    {
+      "epoch": 3.451046590141796,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.7504142190088872e-05,
+      "loss": 0.1099,
+      "step": 11502
+    },
+    {
+      "epoch": 3.4513466876734937,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.7489079680674802e-05,
+      "loss": 0.1058,
+      "step": 11503
+    },
+    {
+      "epoch": 3.4516467852051917,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.7474017171260736e-05,
+      "loss": 0.0953,
+      "step": 11504
+    },
+    {
+      "epoch": 3.4519468827368893,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.7458954661846663e-05,
+      "loss": 0.1022,
+      "step": 11505
+    },
+    {
+      "epoch": 3.4522469802685873,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.74438921524326e-05,
+      "loss": 0.0872,
+      "step": 11506
+    },
+    {
+      "epoch": 3.452547077800285,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.7428829643018526e-05,
+      "loss": 0.0942,
+      "step": 11507
+    },
+    {
+      "epoch": 3.452847175331983,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7413767133604463e-05,
+      "loss": 0.1094,
+      "step": 11508
+    },
+    {
+      "epoch": 3.4531472728636805,
+      "grad_norm": 0.875,
+      "learning_rate": 2.739870462419039e-05,
+      "loss": 0.1043,
+      "step": 11509
+    },
+    {
+      "epoch": 3.4534473703953785,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7383642114776327e-05,
+      "loss": 0.0908,
+      "step": 11510
+    },
+    {
+      "epoch": 3.4537474679270765,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.7368579605362254e-05,
+      "loss": 0.1006,
+      "step": 11511
+    },
+    {
+      "epoch": 3.454047565458774,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.735351709594819e-05,
+      "loss": 0.0846,
+      "step": 11512
+    },
+    {
+      "epoch": 3.4543476629904717,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.7338454586534117e-05,
+      "loss": 0.0967,
+      "step": 11513
+    },
+    {
+      "epoch": 3.4546477605221697,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.7323392077120047e-05,
+      "loss": 0.1276,
+      "step": 11514
+    },
+    {
+      "epoch": 3.4549478580538677,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.730832956770598e-05,
+      "loss": 0.1109,
+      "step": 11515
+    },
+    {
+      "epoch": 3.4552479555855653,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.729326705829191e-05,
+      "loss": 0.0839,
+      "step": 11516
+    },
+    {
+      "epoch": 3.455548053117263,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.7278204548877845e-05,
+      "loss": 0.0988,
+      "step": 11517
+    },
+    {
+      "epoch": 3.455848150648961,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.7263142039463775e-05,
+      "loss": 0.1229,
+      "step": 11518
+    },
+    {
+      "epoch": 3.456148248180659,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.7248079530049708e-05,
+      "loss": 0.1085,
+      "step": 11519
+    },
+    {
+      "epoch": 3.4564483457123565,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.723301702063564e-05,
+      "loss": 0.0911,
+      "step": 11520
+    },
+    {
+      "epoch": 3.4567484432440545,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.7217954511221572e-05,
+      "loss": 0.1142,
+      "step": 11521
+    },
+    {
+      "epoch": 3.457048540775752,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.7202892001807502e-05,
+      "loss": 0.1087,
+      "step": 11522
+    },
+    {
+      "epoch": 3.45734863830745,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.7187829492393436e-05,
+      "loss": 0.1139,
+      "step": 11523
+    },
+    {
+      "epoch": 3.4576487358391477,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.7172766982979366e-05,
+      "loss": 0.1087,
+      "step": 11524
+    },
+    {
+      "epoch": 3.4579488333708457,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.71577044735653e-05,
+      "loss": 0.1046,
+      "step": 11525
+    },
+    {
+      "epoch": 3.4582489309025433,
+      "grad_norm": 0.875,
+      "learning_rate": 2.7142641964151226e-05,
+      "loss": 0.1132,
+      "step": 11526
+    },
+    {
+      "epoch": 3.4585490284342413,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.7127579454737163e-05,
+      "loss": 0.1035,
+      "step": 11527
+    },
+    {
+      "epoch": 3.458849125965939,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.711251694532309e-05,
+      "loss": 0.1157,
+      "step": 11528
+    },
+    {
+      "epoch": 3.459149223497637,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7097454435909027e-05,
+      "loss": 0.1059,
+      "step": 11529
+    },
+    {
+      "epoch": 3.4594493210293344,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.7082391926494953e-05,
+      "loss": 0.1021,
+      "step": 11530
+    },
+    {
+      "epoch": 3.4597494185610325,
+      "grad_norm": 0.875,
+      "learning_rate": 2.706732941708089e-05,
+      "loss": 0.1073,
+      "step": 11531
+    },
+    {
+      "epoch": 3.46004951609273,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.7052266907666817e-05,
+      "loss": 0.0774,
+      "step": 11532
+    },
+    {
+      "epoch": 3.460349613624428,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.7037204398252754e-05,
+      "loss": 0.0936,
+      "step": 11533
+    },
+    {
+      "epoch": 3.4606497111561256,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.702214188883868e-05,
+      "loss": 0.0781,
+      "step": 11534
+    },
+    {
+      "epoch": 3.4609498086878236,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.7007079379424614e-05,
+      "loss": 0.0961,
+      "step": 11535
+    },
+    {
+      "epoch": 3.461249906219521,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.6992016870010544e-05,
+      "loss": 0.1142,
+      "step": 11536
+    },
+    {
+      "epoch": 3.4615500037512192,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.6976954360596478e-05,
+      "loss": 0.1081,
+      "step": 11537
+    },
+    {
+      "epoch": 3.461850101282917,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.6961891851182408e-05,
+      "loss": 0.086,
+      "step": 11538
+    },
+    {
+      "epoch": 3.462150198814615,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.694682934176834e-05,
+      "loss": 0.0811,
+      "step": 11539
+    },
+    {
+      "epoch": 3.4624502963463124,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.6931766832354272e-05,
+      "loss": 0.1383,
+      "step": 11540
+    },
+    {
+      "epoch": 3.4627503938780104,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.6916704322940202e-05,
+      "loss": 0.1126,
+      "step": 11541
+    },
+    {
+      "epoch": 3.463050491409708,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.6901641813526135e-05,
+      "loss": 0.1125,
+      "step": 11542
+    },
+    {
+      "epoch": 3.463350588941406,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.6886579304112066e-05,
+      "loss": 0.108,
+      "step": 11543
+    },
+    {
+      "epoch": 3.4636506864731036,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.6871516794698e-05,
+      "loss": 0.0964,
+      "step": 11544
+    },
+    {
+      "epoch": 3.4639507840048016,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.6856454285283926e-05,
+      "loss": 0.0873,
+      "step": 11545
+    },
+    {
+      "epoch": 3.464250881536499,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.6841391775869863e-05,
+      "loss": 0.0907,
+      "step": 11546
+    },
+    {
+      "epoch": 3.464550979068197,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.682632926645579e-05,
+      "loss": 0.1058,
+      "step": 11547
+    },
+    {
+      "epoch": 3.4648510765998948,
+      "grad_norm": 0.875,
+      "learning_rate": 2.6811266757041726e-05,
+      "loss": 0.1069,
+      "step": 11548
+    },
+    {
+      "epoch": 3.465151174131593,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.6796204247627653e-05,
+      "loss": 0.1593,
+      "step": 11549
+    },
+    {
+      "epoch": 3.465451271663291,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.678114173821359e-05,
+      "loss": 0.1241,
+      "step": 11550
+    },
+    {
+      "epoch": 3.4657513691949884,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.6766079228799517e-05,
+      "loss": 0.113,
+      "step": 11551
+    },
+    {
+      "epoch": 3.466051466726686,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.6751016719385454e-05,
+      "loss": 0.0865,
+      "step": 11552
+    },
+    {
+      "epoch": 3.466351564258384,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.673595420997138e-05,
+      "loss": 0.1029,
+      "step": 11553
+    },
+    {
+      "epoch": 3.466651661790082,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.6720891700557314e-05,
+      "loss": 0.098,
+      "step": 11554
+    },
+    {
+      "epoch": 3.4669517593217796,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.6705829191143244e-05,
+      "loss": 0.1349,
+      "step": 11555
+    },
+    {
+      "epoch": 3.467251856853477,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.6690766681729178e-05,
+      "loss": 0.0985,
+      "step": 11556
+    },
+    {
+      "epoch": 3.467551954385175,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6675704172315108e-05,
+      "loss": 0.0691,
+      "step": 11557
+    },
+    {
+      "epoch": 3.467852051916873,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.666064166290104e-05,
+      "loss": 0.0978,
+      "step": 11558
+    },
+    {
+      "epoch": 3.4681521494485708,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.664557915348697e-05,
+      "loss": 0.0961,
+      "step": 11559
+    },
+    {
+      "epoch": 3.4684522469802688,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.6630516644072905e-05,
+      "loss": 0.0796,
+      "step": 11560
+    },
+    {
+      "epoch": 3.4687523445119663,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.6615454134658835e-05,
+      "loss": 0.1106,
+      "step": 11561
+    },
+    {
+      "epoch": 3.4690524420436644,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.660039162524477e-05,
+      "loss": 0.1059,
+      "step": 11562
+    },
+    {
+      "epoch": 3.469352539575362,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.65853291158307e-05,
+      "loss": 0.1057,
+      "step": 11563
+    },
+    {
+      "epoch": 3.46965263710706,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.6570266606416632e-05,
+      "loss": 0.1099,
+      "step": 11564
+    },
+    {
+      "epoch": 3.4699527346387575,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.6555204097002563e-05,
+      "loss": 0.0983,
+      "step": 11565
+    },
+    {
+      "epoch": 3.4702528321704555,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.6540141587588496e-05,
+      "loss": 0.1057,
+      "step": 11566
+    },
+    {
+      "epoch": 3.470552929702153,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.6525079078174426e-05,
+      "loss": 0.0902,
+      "step": 11567
+    },
+    {
+      "epoch": 3.470853027233851,
+      "grad_norm": 0.875,
+      "learning_rate": 2.6510016568760353e-05,
+      "loss": 0.0988,
+      "step": 11568
+    },
+    {
+      "epoch": 3.4711531247655487,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.649495405934629e-05,
+      "loss": 0.1138,
+      "step": 11569
+    },
+    {
+      "epoch": 3.4714532222972467,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.6479891549932217e-05,
+      "loss": 0.1188,
+      "step": 11570
+    },
+    {
+      "epoch": 3.4717533198289443,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.6464829040518154e-05,
+      "loss": 0.0858,
+      "step": 11571
+    },
+    {
+      "epoch": 3.4720534173606423,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.644976653110408e-05,
+      "loss": 0.0936,
+      "step": 11572
+    },
+    {
+      "epoch": 3.47235351489234,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.6434704021690014e-05,
+      "loss": 0.0931,
+      "step": 11573
+    },
+    {
+      "epoch": 3.472653612424038,
+      "grad_norm": 0.875,
+      "learning_rate": 2.6419641512275944e-05,
+      "loss": 0.1053,
+      "step": 11574
+    },
+    {
+      "epoch": 3.4729537099557355,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.6404579002861877e-05,
+      "loss": 0.1181,
+      "step": 11575
+    },
+    {
+      "epoch": 3.4732538074874335,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.6389516493447808e-05,
+      "loss": 0.0963,
+      "step": 11576
+    },
+    {
+      "epoch": 3.473553905019131,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.637445398403374e-05,
+      "loss": 0.1138,
+      "step": 11577
+    },
+    {
+      "epoch": 3.473854002550829,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.635939147461967e-05,
+      "loss": 0.1128,
+      "step": 11578
+    },
+    {
+      "epoch": 3.4741541000825267,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.6344328965205605e-05,
+      "loss": 0.0985,
+      "step": 11579
+    },
+    {
+      "epoch": 3.4744541976142247,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.6329266455791535e-05,
+      "loss": 0.103,
+      "step": 11580
+    },
+    {
+      "epoch": 3.4747542951459223,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.631420394637747e-05,
+      "loss": 0.0999,
+      "step": 11581
+    },
+    {
+      "epoch": 3.4750543926776203,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.62991414369634e-05,
+      "loss": 0.0848,
+      "step": 11582
+    },
+    {
+      "epoch": 3.475354490209318,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.6284078927549332e-05,
+      "loss": 0.1048,
+      "step": 11583
+    },
+    {
+      "epoch": 3.475654587741016,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.6269016418135262e-05,
+      "loss": 0.1136,
+      "step": 11584
+    },
+    {
+      "epoch": 3.4759546852727135,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.6253953908721196e-05,
+      "loss": 0.1203,
+      "step": 11585
+    },
+    {
+      "epoch": 3.4762547828044115,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.6238891399307126e-05,
+      "loss": 0.1197,
+      "step": 11586
+    },
+    {
+      "epoch": 3.476554880336109,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.622382888989306e-05,
+      "loss": 0.1072,
+      "step": 11587
+    },
+    {
+      "epoch": 3.476854977867807,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.620876638047899e-05,
+      "loss": 0.0988,
+      "step": 11588
+    },
+    {
+      "epoch": 3.477155075399505,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.6193703871064923e-05,
+      "loss": 0.0897,
+      "step": 11589
+    },
+    {
+      "epoch": 3.4774551729312027,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.6178641361650853e-05,
+      "loss": 0.1153,
+      "step": 11590
+    },
+    {
+      "epoch": 3.4777552704629002,
+      "grad_norm": 0.75,
+      "learning_rate": 2.6163578852236787e-05,
+      "loss": 0.0942,
+      "step": 11591
+    },
+    {
+      "epoch": 3.4780553679945982,
+      "grad_norm": 0.875,
+      "learning_rate": 2.6148516342822717e-05,
+      "loss": 0.1108,
+      "step": 11592
+    },
+    {
+      "epoch": 3.4783554655262963,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.613345383340865e-05,
+      "loss": 0.0995,
+      "step": 11593
+    },
+    {
+      "epoch": 3.478655563057994,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.6118391323994577e-05,
+      "loss": 0.0947,
+      "step": 11594
+    },
+    {
+      "epoch": 3.4789556605896914,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.6103328814580507e-05,
+      "loss": 0.1173,
+      "step": 11595
+    },
+    {
+      "epoch": 3.4792557581213894,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.608826630516644e-05,
+      "loss": 0.1108,
+      "step": 11596
+    },
+    {
+      "epoch": 3.4795558556530874,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.607320379575237e-05,
+      "loss": 0.0975,
+      "step": 11597
+    },
+    {
+      "epoch": 3.479855953184785,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.6058141286338305e-05,
+      "loss": 0.1485,
+      "step": 11598
+    },
+    {
+      "epoch": 3.480156050716483,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.6043078776924235e-05,
+      "loss": 0.1007,
+      "step": 11599
+    },
+    {
+      "epoch": 3.4804561482481806,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.6028016267510168e-05,
+      "loss": 0.1008,
+      "step": 11600
+    },
+    {
+      "epoch": 3.4807562457798786,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.60129537580961e-05,
+      "loss": 0.1029,
+      "step": 11601
+    },
+    {
+      "epoch": 3.481056343311576,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.5997891248682032e-05,
+      "loss": 0.1,
+      "step": 11602
+    },
+    {
+      "epoch": 3.4813564408432742,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.5982828739267962e-05,
+      "loss": 0.1225,
+      "step": 11603
+    },
+    {
+      "epoch": 3.481656538374972,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.5967766229853896e-05,
+      "loss": 0.0967,
+      "step": 11604
+    },
+    {
+      "epoch": 3.48195663590667,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.5952703720439826e-05,
+      "loss": 0.1071,
+      "step": 11605
+    },
+    {
+      "epoch": 3.4822567334383674,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.593764121102576e-05,
+      "loss": 0.0945,
+      "step": 11606
+    },
+    {
+      "epoch": 3.4825568309700654,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.592257870161169e-05,
+      "loss": 0.0906,
+      "step": 11607
+    },
+    {
+      "epoch": 3.482856928501763,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.5907516192197623e-05,
+      "loss": 0.132,
+      "step": 11608
+    },
+    {
+      "epoch": 3.483157026033461,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.5892453682783553e-05,
+      "loss": 0.0973,
+      "step": 11609
+    },
+    {
+      "epoch": 3.4834571235651586,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.5877391173369487e-05,
+      "loss": 0.0887,
+      "step": 11610
+    },
+    {
+      "epoch": 3.4837572210968566,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.5862328663955417e-05,
+      "loss": 0.1159,
+      "step": 11611
+    },
+    {
+      "epoch": 3.484057318628554,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.584726615454135e-05,
+      "loss": 0.0915,
+      "step": 11612
+    },
+    {
+      "epoch": 3.484357416160252,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.5832203645127277e-05,
+      "loss": 0.1215,
+      "step": 11613
+    },
+    {
+      "epoch": 3.4846575136919498,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.5817141135713214e-05,
+      "loss": 0.1026,
+      "step": 11614
+    },
+    {
+      "epoch": 3.4849576112236478,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.580207862629914e-05,
+      "loss": 0.1159,
+      "step": 11615
+    },
+    {
+      "epoch": 3.4852577087553454,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.5787016116885078e-05,
+      "loss": 0.1235,
+      "step": 11616
+    },
+    {
+      "epoch": 3.4855578062870434,
+      "grad_norm": 0.875,
+      "learning_rate": 2.5771953607471004e-05,
+      "loss": 0.1184,
+      "step": 11617
+    },
+    {
+      "epoch": 3.485857903818741,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.575689109805694e-05,
+      "loss": 0.0906,
+      "step": 11618
+    },
+    {
+      "epoch": 3.486158001350439,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.5741828588642868e-05,
+      "loss": 0.1069,
+      "step": 11619
+    },
+    {
+      "epoch": 3.4864580988821365,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.5726766079228805e-05,
+      "loss": 0.1209,
+      "step": 11620
+    },
+    {
+      "epoch": 3.4867581964138346,
+      "grad_norm": 0.875,
+      "learning_rate": 2.5711703569814732e-05,
+      "loss": 0.1161,
+      "step": 11621
+    },
+    {
+      "epoch": 3.487058293945532,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.5696641060400662e-05,
+      "loss": 0.1265,
+      "step": 11622
+    },
+    {
+      "epoch": 3.48735839147723,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.5681578550986595e-05,
+      "loss": 0.0956,
+      "step": 11623
+    },
+    {
+      "epoch": 3.4876584890089277,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.5666516041572526e-05,
+      "loss": 0.1082,
+      "step": 11624
+    },
+    {
+      "epoch": 3.4879585865406257,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.565145353215846e-05,
+      "loss": 0.0987,
+      "step": 11625
+    },
+    {
+      "epoch": 3.4882586840723233,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.563639102274439e-05,
+      "loss": 0.1119,
+      "step": 11626
+    },
+    {
+      "epoch": 3.4885587816040213,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.5621328513330323e-05,
+      "loss": 0.0955,
+      "step": 11627
+    },
+    {
+      "epoch": 3.4888588791357193,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.5606266003916253e-05,
+      "loss": 0.0985,
+      "step": 11628
+    },
+    {
+      "epoch": 3.489158976667417,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.5591203494502186e-05,
+      "loss": 0.0959,
+      "step": 11629
+    },
+    {
+      "epoch": 3.4894590741991145,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.5576140985088117e-05,
+      "loss": 0.1098,
+      "step": 11630
+    },
+    {
+      "epoch": 3.4897591717308125,
+      "grad_norm": 1.0,
+      "learning_rate": 2.556107847567405e-05,
+      "loss": 0.1416,
+      "step": 11631
+    },
+    {
+      "epoch": 3.4900592692625105,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.5546015966259977e-05,
+      "loss": 0.125,
+      "step": 11632
+    },
+    {
+      "epoch": 3.490359366794208,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.5530953456845914e-05,
+      "loss": 0.0925,
+      "step": 11633
+    },
+    {
+      "epoch": 3.4906594643259057,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.551589094743184e-05,
+      "loss": 0.1116,
+      "step": 11634
+    },
+    {
+      "epoch": 3.4909595618576037,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.5500828438017777e-05,
+      "loss": 0.0971,
+      "step": 11635
+    },
+    {
+      "epoch": 3.4912596593893017,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.5485765928603704e-05,
+      "loss": 0.0965,
+      "step": 11636
+    },
+    {
+      "epoch": 3.4915597569209993,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.547070341918964e-05,
+      "loss": 0.1016,
+      "step": 11637
+    },
+    {
+      "epoch": 3.4918598544526973,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.5455640909775568e-05,
+      "loss": 0.1176,
+      "step": 11638
+    },
+    {
+      "epoch": 3.492159951984395,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.5440578400361505e-05,
+      "loss": 0.0986,
+      "step": 11639
+    },
+    {
+      "epoch": 3.492460049516093,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.542551589094743e-05,
+      "loss": 0.0946,
+      "step": 11640
+    },
+    {
+      "epoch": 3.4927601470477905,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.541045338153337e-05,
+      "loss": 0.0834,
+      "step": 11641
+    },
+    {
+      "epoch": 3.4930602445794885,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.5395390872119295e-05,
+      "loss": 0.0841,
+      "step": 11642
+    },
+    {
+      "epoch": 3.493360342111186,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.538032836270523e-05,
+      "loss": 0.0975,
+      "step": 11643
+    },
+    {
+      "epoch": 3.493660439642884,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.536526585329116e-05,
+      "loss": 0.0786,
+      "step": 11644
+    },
+    {
+      "epoch": 3.4939605371745817,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.5350203343877092e-05,
+      "loss": 0.0864,
+      "step": 11645
+    },
+    {
+      "epoch": 3.4942606347062797,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.5335140834463023e-05,
+      "loss": 0.0896,
+      "step": 11646
+    },
+    {
+      "epoch": 3.4945607322379773,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.5320078325048956e-05,
+      "loss": 0.0761,
+      "step": 11647
+    },
+    {
+      "epoch": 3.4948608297696753,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.5305015815634886e-05,
+      "loss": 0.0919,
+      "step": 11648
+    },
+    {
+      "epoch": 3.495160927301373,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.5289953306220816e-05,
+      "loss": 0.1129,
+      "step": 11649
+    },
+    {
+      "epoch": 3.495461024833071,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.527489079680675e-05,
+      "loss": 0.1102,
+      "step": 11650
+    },
+    {
+      "epoch": 3.4957611223647684,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.525982828739268e-05,
+      "loss": 0.0923,
+      "step": 11651
+    },
+    {
+      "epoch": 3.4960612198964665,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.5244765777978614e-05,
+      "loss": 0.1275,
+      "step": 11652
+    },
+    {
+      "epoch": 3.496361317428164,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.522970326856454e-05,
+      "loss": 0.0978,
+      "step": 11653
+    },
+    {
+      "epoch": 3.496661414959862,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.5214640759150477e-05,
+      "loss": 0.0967,
+      "step": 11654
+    },
+    {
+      "epoch": 3.4969615124915596,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.5199578249736404e-05,
+      "loss": 0.1585,
+      "step": 11655
+    },
+    {
+      "epoch": 3.4972616100232576,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.518451574032234e-05,
+      "loss": 0.1026,
+      "step": 11656
+    },
+    {
+      "epoch": 3.497561707554955,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.5169453230908268e-05,
+      "loss": 0.1102,
+      "step": 11657
+    },
+    {
+      "epoch": 3.4978618050866532,
+      "grad_norm": 0.75,
+      "learning_rate": 2.5154390721494205e-05,
+      "loss": 0.0988,
+      "step": 11658
+    },
+    {
+      "epoch": 3.498161902618351,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.513932821208013e-05,
+      "loss": 0.1235,
+      "step": 11659
+    },
+    {
+      "epoch": 3.498462000150049,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.5124265702666068e-05,
+      "loss": 0.0811,
+      "step": 11660
+    },
+    {
+      "epoch": 3.4987620976817464,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.5109203193251995e-05,
+      "loss": 0.0943,
+      "step": 11661
+    },
+    {
+      "epoch": 3.4990621952134444,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.509414068383793e-05,
+      "loss": 0.0972,
+      "step": 11662
+    },
+    {
+      "epoch": 3.499362292745142,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.507907817442386e-05,
+      "loss": 0.1012,
+      "step": 11663
+    },
+    {
+      "epoch": 3.49966239027684,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.5064015665009792e-05,
+      "loss": 0.123,
+      "step": 11664
+    },
+    {
+      "epoch": 3.4999624878085376,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.5048953155595722e-05,
+      "loss": 0.1052,
+      "step": 11665
+    },
+    {
+      "epoch": 3.5002625853402356,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.5033890646181656e-05,
+      "loss": 0.1126,
+      "step": 11666
+    },
+    {
+      "epoch": 3.5005626828719336,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.5018828136767586e-05,
+      "loss": 0.1154,
+      "step": 11667
+    },
+    {
+      "epoch": 3.500862780403631,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.500376562735352e-05,
+      "loss": 0.0891,
+      "step": 11668
+    },
+    {
+      "epoch": 3.5011628779353288,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.498870311793945e-05,
+      "loss": 0.1212,
+      "step": 11669
+    },
+    {
+      "epoch": 3.501462975467027,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.497364060852538e-05,
+      "loss": 0.0956,
+      "step": 11670
+    },
+    {
+      "epoch": 3.501763072998725,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.4958578099111313e-05,
+      "loss": 0.108,
+      "step": 11671
+    },
+    {
+      "epoch": 3.5020631705304224,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.4943515589697243e-05,
+      "loss": 0.1243,
+      "step": 11672
+    },
+    {
+      "epoch": 3.50236326806212,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.4928453080283177e-05,
+      "loss": 0.1029,
+      "step": 11673
+    },
+    {
+      "epoch": 3.502663365593818,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.4913390570869107e-05,
+      "loss": 0.1059,
+      "step": 11674
+    },
+    {
+      "epoch": 3.502963463125516,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.489832806145504e-05,
+      "loss": 0.1081,
+      "step": 11675
+    },
+    {
+      "epoch": 3.5032635606572136,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.488326555204097e-05,
+      "loss": 0.1259,
+      "step": 11676
+    },
+    {
+      "epoch": 3.503563658188911,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.4868203042626904e-05,
+      "loss": 0.1128,
+      "step": 11677
+    },
+    {
+      "epoch": 3.503863755720609,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.4853140533212834e-05,
+      "loss": 0.097,
+      "step": 11678
+    },
+    {
+      "epoch": 3.504163853252307,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.4838078023798768e-05,
+      "loss": 0.0953,
+      "step": 11679
+    },
+    {
+      "epoch": 3.5044639507840047,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.4823015514384698e-05,
+      "loss": 0.102,
+      "step": 11680
+    },
+    {
+      "epoch": 3.5047640483157028,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.4807953004970628e-05,
+      "loss": 0.0985,
+      "step": 11681
+    },
+    {
+      "epoch": 3.5050641458474003,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.4792890495556562e-05,
+      "loss": 0.1009,
+      "step": 11682
+    },
+    {
+      "epoch": 3.5053642433790984,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.4777827986142492e-05,
+      "loss": 0.1215,
+      "step": 11683
+    },
+    {
+      "epoch": 3.505664340910796,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.4762765476728425e-05,
+      "loss": 0.1391,
+      "step": 11684
+    },
+    {
+      "epoch": 3.505964438442494,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.4747702967314356e-05,
+      "loss": 0.0926,
+      "step": 11685
+    },
+    {
+      "epoch": 3.5062645359741915,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.473264045790029e-05,
+      "loss": 0.1022,
+      "step": 11686
+    },
+    {
+      "epoch": 3.5065646335058895,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.471757794848622e-05,
+      "loss": 0.0926,
+      "step": 11687
+    },
+    {
+      "epoch": 3.506864731037587,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.4702515439072153e-05,
+      "loss": 0.1154,
+      "step": 11688
+    },
+    {
+      "epoch": 3.507164828569285,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.468745292965808e-05,
+      "loss": 0.1307,
+      "step": 11689
+    },
+    {
+      "epoch": 3.5074649261009827,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.4672390420244013e-05,
+      "loss": 0.1166,
+      "step": 11690
+    },
+    {
+      "epoch": 3.5077650236326807,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.4657327910829943e-05,
+      "loss": 0.0945,
+      "step": 11691
+    },
+    {
+      "epoch": 3.5080651211643783,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.4642265401415877e-05,
+      "loss": 0.1109,
+      "step": 11692
+    },
+    {
+      "epoch": 3.5083652186960763,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.4627202892001807e-05,
+      "loss": 0.0929,
+      "step": 11693
+    },
+    {
+      "epoch": 3.508665316227774,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.461214038258774e-05,
+      "loss": 0.0912,
+      "step": 11694
+    },
+    {
+      "epoch": 3.508965413759472,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.459707787317367e-05,
+      "loss": 0.1813,
+      "step": 11695
+    },
+    {
+      "epoch": 3.5092655112911695,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.4582015363759604e-05,
+      "loss": 0.138,
+      "step": 11696
+    },
+    {
+      "epoch": 3.5095656088228675,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.4566952854345534e-05,
+      "loss": 0.103,
+      "step": 11697
+    },
+    {
+      "epoch": 3.509865706354565,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.4551890344931468e-05,
+      "loss": 0.1077,
+      "step": 11698
+    },
+    {
+      "epoch": 3.510165803886263,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.4536827835517398e-05,
+      "loss": 0.1014,
+      "step": 11699
+    },
+    {
+      "epoch": 3.5104659014179607,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.452176532610333e-05,
+      "loss": 0.1053,
+      "step": 11700
+    },
+    {
+      "epoch": 3.5107659989496587,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.450670281668926e-05,
+      "loss": 0.0981,
+      "step": 11701
+    },
+    {
+      "epoch": 3.5110660964813567,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.4491640307275192e-05,
+      "loss": 0.0996,
+      "step": 11702
+    },
+    {
+      "epoch": 3.5113661940130543,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.4476577797861125e-05,
+      "loss": 0.1014,
+      "step": 11703
+    },
+    {
+      "epoch": 3.511666291544752,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.4461515288447055e-05,
+      "loss": 0.0984,
+      "step": 11704
+    },
+    {
+      "epoch": 3.51196638907645,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.444645277903299e-05,
+      "loss": 0.1021,
+      "step": 11705
+    },
+    {
+      "epoch": 3.512266486608148,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.443139026961892e-05,
+      "loss": 0.0869,
+      "step": 11706
+    },
+    {
+      "epoch": 3.5125665841398455,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.4416327760204853e-05,
+      "loss": 0.1095,
+      "step": 11707
+    },
+    {
+      "epoch": 3.512866681671543,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.4401265250790783e-05,
+      "loss": 0.0847,
+      "step": 11708
+    },
+    {
+      "epoch": 3.513166779203241,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.4386202741376716e-05,
+      "loss": 0.1327,
+      "step": 11709
+    },
+    {
+      "epoch": 3.513466876734939,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.4371140231962646e-05,
+      "loss": 0.0904,
+      "step": 11710
+    },
+    {
+      "epoch": 3.5137669742666366,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.435607772254858e-05,
+      "loss": 0.0871,
+      "step": 11711
+    },
+    {
+      "epoch": 3.514067071798334,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.434101521313451e-05,
+      "loss": 0.1029,
+      "step": 11712
+    },
+    {
+      "epoch": 3.5143671693300322,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.4325952703720444e-05,
+      "loss": 0.0907,
+      "step": 11713
+    },
+    {
+      "epoch": 3.5146672668617303,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.4310890194306374e-05,
+      "loss": 0.1021,
+      "step": 11714
+    },
+    {
+      "epoch": 3.514967364393428,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.4295827684892304e-05,
+      "loss": 0.0955,
+      "step": 11715
+    },
+    {
+      "epoch": 3.5152674619251254,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.4280765175478234e-05,
+      "loss": 0.0854,
+      "step": 11716
+    },
+    {
+      "epoch": 3.5155675594568234,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.4265702666064168e-05,
+      "loss": 0.0831,
+      "step": 11717
+    },
+    {
+      "epoch": 3.5158676569885214,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.4250640156650098e-05,
+      "loss": 0.1237,
+      "step": 11718
+    },
+    {
+      "epoch": 3.516167754520219,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.423557764723603e-05,
+      "loss": 0.1186,
+      "step": 11719
+    },
+    {
+      "epoch": 3.516467852051917,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.422051513782196e-05,
+      "loss": 0.0944,
+      "step": 11720
+    },
+    {
+      "epoch": 3.5167679495836146,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.420545262840789e-05,
+      "loss": 0.0904,
+      "step": 11721
+    },
+    {
+      "epoch": 3.5170680471153126,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.4190390118993825e-05,
+      "loss": 0.1058,
+      "step": 11722
+    },
+    {
+      "epoch": 3.51736814464701,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.4175327609579755e-05,
+      "loss": 0.1095,
+      "step": 11723
+    },
+    {
+      "epoch": 3.517668242178708,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.416026510016569e-05,
+      "loss": 0.1037,
+      "step": 11724
+    },
+    {
+      "epoch": 3.517968339710406,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.414520259075162e-05,
+      "loss": 0.0824,
+      "step": 11725
+    },
+    {
+      "epoch": 3.518268437242104,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.4130140081337552e-05,
+      "loss": 0.0928,
+      "step": 11726
+    },
+    {
+      "epoch": 3.5185685347738014,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.4115077571923483e-05,
+      "loss": 0.1183,
+      "step": 11727
+    },
+    {
+      "epoch": 3.5188686323054994,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.4100015062509416e-05,
+      "loss": 0.0956,
+      "step": 11728
+    },
+    {
+      "epoch": 3.519168729837197,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.4084952553095346e-05,
+      "loss": 0.0977,
+      "step": 11729
+    },
+    {
+      "epoch": 3.519468827368895,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.406989004368128e-05,
+      "loss": 0.1187,
+      "step": 11730
+    },
+    {
+      "epoch": 3.5197689249005926,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.405482753426721e-05,
+      "loss": 0.1022,
+      "step": 11731
+    },
+    {
+      "epoch": 3.5200690224322906,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.4039765024853143e-05,
+      "loss": 0.1127,
+      "step": 11732
+    },
+    {
+      "epoch": 3.520369119963988,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.4024702515439074e-05,
+      "loss": 0.1145,
+      "step": 11733
+    },
+    {
+      "epoch": 3.520669217495686,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.4009640006025007e-05,
+      "loss": 0.1241,
+      "step": 11734
+    },
+    {
+      "epoch": 3.5209693150273838,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.3994577496610937e-05,
+      "loss": 0.1243,
+      "step": 11735
+    },
+    {
+      "epoch": 3.5212694125590818,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.3979514987196867e-05,
+      "loss": 0.1408,
+      "step": 11736
+    },
+    {
+      "epoch": 3.5215695100907793,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.39644524777828e-05,
+      "loss": 0.1209,
+      "step": 11737
+    },
+    {
+      "epoch": 3.5218696076224774,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.394938996836873e-05,
+      "loss": 0.1037,
+      "step": 11738
+    },
+    {
+      "epoch": 3.522169705154175,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.3934327458954665e-05,
+      "loss": 0.0985,
+      "step": 11739
+    },
+    {
+      "epoch": 3.522469802685873,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.3919264949540595e-05,
+      "loss": 0.0907,
+      "step": 11740
+    },
+    {
+      "epoch": 3.522769900217571,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.3904202440126528e-05,
+      "loss": 0.0933,
+      "step": 11741
+    },
+    {
+      "epoch": 3.5230699977492685,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.388913993071246e-05,
+      "loss": 0.0836,
+      "step": 11742
+    },
+    {
+      "epoch": 3.523370095280966,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.387407742129839e-05,
+      "loss": 0.093,
+      "step": 11743
+    },
+    {
+      "epoch": 3.523670192812664,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.385901491188432e-05,
+      "loss": 0.1236,
+      "step": 11744
+    },
+    {
+      "epoch": 3.523970290344362,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.3843952402470252e-05,
+      "loss": 0.1107,
+      "step": 11745
+    },
+    {
+      "epoch": 3.5242703878760597,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.3828889893056182e-05,
+      "loss": 0.0947,
+      "step": 11746
+    },
+    {
+      "epoch": 3.5245704854077573,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.3813827383642116e-05,
+      "loss": 0.1138,
+      "step": 11747
+    },
+    {
+      "epoch": 3.5248705829394553,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.3798764874228046e-05,
+      "loss": 0.1329,
+      "step": 11748
+    },
+    {
+      "epoch": 3.5251706804711533,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.378370236481398e-05,
+      "loss": 0.1,
+      "step": 11749
+    },
+    {
+      "epoch": 3.525470778002851,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.376863985539991e-05,
+      "loss": 0.1044,
+      "step": 11750
+    },
+    {
+      "epoch": 3.5257708755345485,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.3753577345985843e-05,
+      "loss": 0.1121,
+      "step": 11751
+    },
+    {
+      "epoch": 3.5260709730662465,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.3738514836571773e-05,
+      "loss": 0.1018,
+      "step": 11752
+    },
+    {
+      "epoch": 3.5263710705979445,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.3723452327157707e-05,
+      "loss": 0.1031,
+      "step": 11753
+    },
+    {
+      "epoch": 3.526671168129642,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.3708389817743637e-05,
+      "loss": 0.114,
+      "step": 11754
+    },
+    {
+      "epoch": 3.5269712656613397,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.3693327308329567e-05,
+      "loss": 0.1172,
+      "step": 11755
+    },
+    {
+      "epoch": 3.5272713631930377,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.36782647989155e-05,
+      "loss": 0.0808,
+      "step": 11756
+    },
+    {
+      "epoch": 3.5275714607247357,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.366320228950143e-05,
+      "loss": 0.0883,
+      "step": 11757
+    },
+    {
+      "epoch": 3.5278715582564333,
+      "grad_norm": 1.0,
+      "learning_rate": 2.3648139780087364e-05,
+      "loss": 0.1226,
+      "step": 11758
+    },
+    {
+      "epoch": 3.5281716557881313,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.3633077270673294e-05,
+      "loss": 0.0794,
+      "step": 11759
+    },
+    {
+      "epoch": 3.528471753319829,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.3618014761259228e-05,
+      "loss": 0.0952,
+      "step": 11760
+    },
+    {
+      "epoch": 3.528771850851527,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.3602952251845158e-05,
+      "loss": 0.1032,
+      "step": 11761
+    },
+    {
+      "epoch": 3.5290719483832245,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.358788974243109e-05,
+      "loss": 0.1105,
+      "step": 11762
+    },
+    {
+      "epoch": 3.5293720459149225,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.3572827233017022e-05,
+      "loss": 0.0964,
+      "step": 11763
+    },
+    {
+      "epoch": 3.52967214344662,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.3557764723602955e-05,
+      "loss": 0.0848,
+      "step": 11764
+    },
+    {
+      "epoch": 3.529972240978318,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3542702214188885e-05,
+      "loss": 0.0991,
+      "step": 11765
+    },
+    {
+      "epoch": 3.5302723385100157,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.352763970477482e-05,
+      "loss": 0.1108,
+      "step": 11766
+    },
+    {
+      "epoch": 3.5305724360417137,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.351257719536075e-05,
+      "loss": 0.0904,
+      "step": 11767
+    },
+    {
+      "epoch": 3.5308725335734112,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.3497514685946683e-05,
+      "loss": 0.088,
+      "step": 11768
+    },
+    {
+      "epoch": 3.5311726311051093,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.3482452176532613e-05,
+      "loss": 0.1133,
+      "step": 11769
+    },
+    {
+      "epoch": 3.531472728636807,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.3467389667118543e-05,
+      "loss": 0.0936,
+      "step": 11770
+    },
+    {
+      "epoch": 3.531772826168505,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.3452327157704473e-05,
+      "loss": 0.0857,
+      "step": 11771
+    },
+    {
+      "epoch": 3.5320729237002024,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.3437264648290407e-05,
+      "loss": 0.1352,
+      "step": 11772
+    },
+    {
+      "epoch": 3.5323730212319004,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.3422202138876337e-05,
+      "loss": 0.1178,
+      "step": 11773
+    },
+    {
+      "epoch": 3.532673118763598,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.3407139629462267e-05,
+      "loss": 0.0804,
+      "step": 11774
+    },
+    {
+      "epoch": 3.532973216295296,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.33920771200482e-05,
+      "loss": 0.0975,
+      "step": 11775
+    },
+    {
+      "epoch": 3.5332733138269936,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.337701461063413e-05,
+      "loss": 0.0967,
+      "step": 11776
+    },
+    {
+      "epoch": 3.5335734113586916,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.3361952101220064e-05,
+      "loss": 0.1231,
+      "step": 11777
+    },
+    {
+      "epoch": 3.533873508890389,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.3346889591805994e-05,
+      "loss": 0.1226,
+      "step": 11778
+    },
+    {
+      "epoch": 3.534173606422087,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.3331827082391928e-05,
+      "loss": 0.0855,
+      "step": 11779
+    },
+    {
+      "epoch": 3.5344737039537852,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.3316764572977858e-05,
+      "loss": 0.1016,
+      "step": 11780
+    },
+    {
+      "epoch": 3.534773801485483,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.330170206356379e-05,
+      "loss": 0.1017,
+      "step": 11781
+    },
+    {
+      "epoch": 3.5350738990171804,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.328663955414972e-05,
+      "loss": 0.0873,
+      "step": 11782
+    },
+    {
+      "epoch": 3.5353739965488784,
+      "grad_norm": 0.875,
+      "learning_rate": 2.3271577044735655e-05,
+      "loss": 0.1054,
+      "step": 11783
+    },
+    {
+      "epoch": 3.5356740940805764,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.3256514535321585e-05,
+      "loss": 0.1185,
+      "step": 11784
+    },
+    {
+      "epoch": 3.535974191612274,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.324145202590752e-05,
+      "loss": 0.0901,
+      "step": 11785
+    },
+    {
+      "epoch": 3.5362742891439716,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.322638951649345e-05,
+      "loss": 0.1205,
+      "step": 11786
+    },
+    {
+      "epoch": 3.5365743866756696,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.3211327007079382e-05,
+      "loss": 0.1097,
+      "step": 11787
+    },
+    {
+      "epoch": 3.5368744842073676,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.3196264497665313e-05,
+      "loss": 0.1191,
+      "step": 11788
+    },
+    {
+      "epoch": 3.537174581739065,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.3181201988251243e-05,
+      "loss": 0.0989,
+      "step": 11789
+    },
+    {
+      "epoch": 3.5374746792707628,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3166139478837176e-05,
+      "loss": 0.1031,
+      "step": 11790
+    },
+    {
+      "epoch": 3.5377747768024608,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.3151076969423106e-05,
+      "loss": 0.1102,
+      "step": 11791
+    },
+    {
+      "epoch": 3.538074874334159,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.313601446000904e-05,
+      "loss": 0.1058,
+      "step": 11792
+    },
+    {
+      "epoch": 3.5383749718658564,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.312095195059497e-05,
+      "loss": 0.1166,
+      "step": 11793
+    },
+    {
+      "epoch": 3.538675069397554,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.3105889441180904e-05,
+      "loss": 0.1296,
+      "step": 11794
+    },
+    {
+      "epoch": 3.538975166929252,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.3090826931766834e-05,
+      "loss": 0.1067,
+      "step": 11795
+    },
+    {
+      "epoch": 3.53927526446095,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.3075764422352767e-05,
+      "loss": 0.1054,
+      "step": 11796
+    },
+    {
+      "epoch": 3.5395753619926476,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.3060701912938694e-05,
+      "loss": 0.0999,
+      "step": 11797
+    },
+    {
+      "epoch": 3.5398754595243456,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.3045639403524628e-05,
+      "loss": 0.09,
+      "step": 11798
+    },
+    {
+      "epoch": 3.540175557056043,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.3030576894110558e-05,
+      "loss": 0.0839,
+      "step": 11799
+    },
+    {
+      "epoch": 3.540475654587741,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.301551438469649e-05,
+      "loss": 0.0899,
+      "step": 11800
+    },
+    {
+      "epoch": 3.5407757521194387,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.300045187528242e-05,
+      "loss": 0.1235,
+      "step": 11801
+    },
+    {
+      "epoch": 3.5410758496511368,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.2985389365868355e-05,
+      "loss": 0.0955,
+      "step": 11802
+    },
+    {
+      "epoch": 3.5413759471828343,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.2970326856454285e-05,
+      "loss": 0.0906,
+      "step": 11803
+    },
+    {
+      "epoch": 3.5416760447145323,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.295526434704022e-05,
+      "loss": 0.1111,
+      "step": 11804
+    },
+    {
+      "epoch": 3.54197614224623,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.294020183762615e-05,
+      "loss": 0.0994,
+      "step": 11805
+    },
+    {
+      "epoch": 3.542276239777928,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.2925139328212082e-05,
+      "loss": 0.0857,
+      "step": 11806
+    },
+    {
+      "epoch": 3.5425763373096255,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.2910076818798012e-05,
+      "loss": 0.1132,
+      "step": 11807
+    },
+    {
+      "epoch": 3.5428764348413235,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.2895014309383943e-05,
+      "loss": 0.1126,
+      "step": 11808
+    },
+    {
+      "epoch": 3.543176532373021,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.2879951799969876e-05,
+      "loss": 0.0918,
+      "step": 11809
+    },
+    {
+      "epoch": 3.543476629904719,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.2864889290555806e-05,
+      "loss": 0.119,
+      "step": 11810
+    },
+    {
+      "epoch": 3.5437767274364167,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.284982678114174e-05,
+      "loss": 0.0881,
+      "step": 11811
+    },
+    {
+      "epoch": 3.5440768249681147,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.283476427172767e-05,
+      "loss": 0.1073,
+      "step": 11812
+    },
+    {
+      "epoch": 3.5443769224998123,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.2819701762313603e-05,
+      "loss": 0.109,
+      "step": 11813
+    },
+    {
+      "epoch": 3.5446770200315103,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.2804639252899534e-05,
+      "loss": 0.0953,
+      "step": 11814
+    },
+    {
+      "epoch": 3.544977117563208,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.2789576743485467e-05,
+      "loss": 0.0944,
+      "step": 11815
+    },
+    {
+      "epoch": 3.545277215094906,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.2774514234071397e-05,
+      "loss": 0.0998,
+      "step": 11816
+    },
+    {
+      "epoch": 3.5455773126266035,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.275945172465733e-05,
+      "loss": 0.0999,
+      "step": 11817
+    },
+    {
+      "epoch": 3.5458774101583015,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.274438921524326e-05,
+      "loss": 0.1067,
+      "step": 11818
+    },
+    {
+      "epoch": 3.5461775076899995,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.2729326705829194e-05,
+      "loss": 0.1288,
+      "step": 11819
+    },
+    {
+      "epoch": 3.546477605221697,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.2714264196415125e-05,
+      "loss": 0.104,
+      "step": 11820
+    },
+    {
+      "epoch": 3.5467777027533947,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.2699201687001058e-05,
+      "loss": 0.1136,
+      "step": 11821
+    },
+    {
+      "epoch": 3.5470778002850927,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2684139177586988e-05,
+      "loss": 0.0806,
+      "step": 11822
+    },
+    {
+      "epoch": 3.5473778978167907,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.266907666817292e-05,
+      "loss": 0.1166,
+      "step": 11823
+    },
+    {
+      "epoch": 3.5476779953484883,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.2654014158758852e-05,
+      "loss": 0.131,
+      "step": 11824
+    },
+    {
+      "epoch": 3.547978092880186,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.2638951649344782e-05,
+      "loss": 0.0739,
+      "step": 11825
+    },
+    {
+      "epoch": 3.548278190411884,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.2623889139930712e-05,
+      "loss": 0.0985,
+      "step": 11826
+    },
+    {
+      "epoch": 3.548578287943582,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.2608826630516646e-05,
+      "loss": 0.1106,
+      "step": 11827
+    },
+    {
+      "epoch": 3.5488783854752795,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.2593764121102576e-05,
+      "loss": 0.0906,
+      "step": 11828
+    },
+    {
+      "epoch": 3.549178483006977,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.2578701611688506e-05,
+      "loss": 0.0887,
+      "step": 11829
+    },
+    {
+      "epoch": 3.549478580538675,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.256363910227444e-05,
+      "loss": 0.0834,
+      "step": 11830
+    },
+    {
+      "epoch": 3.549778678070373,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.254857659286037e-05,
+      "loss": 0.1073,
+      "step": 11831
+    },
+    {
+      "epoch": 3.5500787756020706,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.2533514083446303e-05,
+      "loss": 0.1195,
+      "step": 11832
+    },
+    {
+      "epoch": 3.550378873133768,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.2518451574032233e-05,
+      "loss": 0.1058,
+      "step": 11833
+    },
+    {
+      "epoch": 3.5506789706654662,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.2503389064618167e-05,
+      "loss": 0.1116,
+      "step": 11834
+    },
+    {
+      "epoch": 3.5509790681971642,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.2488326555204097e-05,
+      "loss": 0.0867,
+      "step": 11835
+    },
+    {
+      "epoch": 3.551279165728862,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.247326404579003e-05,
+      "loss": 0.1037,
+      "step": 11836
+    },
+    {
+      "epoch": 3.55157926326056,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.245820153637596e-05,
+      "loss": 0.1413,
+      "step": 11837
+    },
+    {
+      "epoch": 3.5518793607922574,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.2443139026961894e-05,
+      "loss": 0.1132,
+      "step": 11838
+    },
+    {
+      "epoch": 3.5521794583239554,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.2428076517547824e-05,
+      "loss": 0.0907,
+      "step": 11839
+    },
+    {
+      "epoch": 3.552479555855653,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.2413014008133758e-05,
+      "loss": 0.0848,
+      "step": 11840
+    },
+    {
+      "epoch": 3.552779653387351,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.2397951498719688e-05,
+      "loss": 0.1308,
+      "step": 11841
+    },
+    {
+      "epoch": 3.5530797509190486,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.2382888989305618e-05,
+      "loss": 0.1019,
+      "step": 11842
+    },
+    {
+      "epoch": 3.5533798484507466,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.2367826479891552e-05,
+      "loss": 0.2023,
+      "step": 11843
+    },
+    {
+      "epoch": 3.553679945982444,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.2352763970477482e-05,
+      "loss": 0.0703,
+      "step": 11844
+    },
+    {
+      "epoch": 3.553980043514142,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.2337701461063415e-05,
+      "loss": 0.0785,
+      "step": 11845
+    },
+    {
+      "epoch": 3.55428014104584,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.2322638951649346e-05,
+      "loss": 0.0817,
+      "step": 11846
+    },
+    {
+      "epoch": 3.554580238577538,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.230757644223528e-05,
+      "loss": 0.0929,
+      "step": 11847
+    },
+    {
+      "epoch": 3.5548803361092354,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.229251393282121e-05,
+      "loss": 0.0946,
+      "step": 11848
+    },
+    {
+      "epoch": 3.5551804336409334,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.2277451423407143e-05,
+      "loss": 0.0994,
+      "step": 11849
+    },
+    {
+      "epoch": 3.555480531172631,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.2262388913993073e-05,
+      "loss": 0.0998,
+      "step": 11850
+    },
+    {
+      "epoch": 3.555780628704329,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.2247326404579006e-05,
+      "loss": 0.1057,
+      "step": 11851
+    },
+    {
+      "epoch": 3.5560807262360266,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.2232263895164933e-05,
+      "loss": 0.0824,
+      "step": 11852
+    },
+    {
+      "epoch": 3.5563808237677246,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.2217201385750867e-05,
+      "loss": 0.0868,
+      "step": 11853
+    },
+    {
+      "epoch": 3.556680921299422,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.2202138876336797e-05,
+      "loss": 0.1362,
+      "step": 11854
+    },
+    {
+      "epoch": 3.55698101883112,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.218707636692273e-05,
+      "loss": 0.0972,
+      "step": 11855
+    },
+    {
+      "epoch": 3.5572811163628177,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.217201385750866e-05,
+      "loss": 0.0812,
+      "step": 11856
+    },
+    {
+      "epoch": 3.5575812138945158,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.2156951348094594e-05,
+      "loss": 0.1074,
+      "step": 11857
+    },
+    {
+      "epoch": 3.5578813114262138,
+      "grad_norm": 0.875,
+      "learning_rate": 2.2141888838680524e-05,
+      "loss": 0.1205,
+      "step": 11858
+    },
+    {
+      "epoch": 3.5581814089579114,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.2126826329266458e-05,
+      "loss": 0.1038,
+      "step": 11859
+    },
+    {
+      "epoch": 3.558481506489609,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.2111763819852388e-05,
+      "loss": 0.0949,
+      "step": 11860
+    },
+    {
+      "epoch": 3.558781604021307,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.209670131043832e-05,
+      "loss": 0.1102,
+      "step": 11861
+    },
+    {
+      "epoch": 3.559081701553005,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.208163880102425e-05,
+      "loss": 0.0877,
+      "step": 11862
+    },
+    {
+      "epoch": 3.5593817990847025,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.206657629161018e-05,
+      "loss": 0.1271,
+      "step": 11863
+    },
+    {
+      "epoch": 3.5596818966164,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.2051513782196115e-05,
+      "loss": 0.0813,
+      "step": 11864
+    },
+    {
+      "epoch": 3.559981994148098,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.2036451272782045e-05,
+      "loss": 0.0848,
+      "step": 11865
+    },
+    {
+      "epoch": 3.560282091679796,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.202138876336798e-05,
+      "loss": 0.0896,
+      "step": 11866
+    },
+    {
+      "epoch": 3.5605821892114937,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.200632625395391e-05,
+      "loss": 0.084,
+      "step": 11867
+    },
+    {
+      "epoch": 3.5608822867431913,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.1991263744539843e-05,
+      "loss": 0.1145,
+      "step": 11868
+    },
+    {
+      "epoch": 3.5611823842748893,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.1976201235125773e-05,
+      "loss": 0.117,
+      "step": 11869
+    },
+    {
+      "epoch": 3.5614824818065873,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.1961138725711706e-05,
+      "loss": 0.1308,
+      "step": 11870
+    },
+    {
+      "epoch": 3.561782579338285,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.1946076216297636e-05,
+      "loss": 0.1156,
+      "step": 11871
+    },
+    {
+      "epoch": 3.5620826768699825,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.193101370688357e-05,
+      "loss": 0.1212,
+      "step": 11872
+    },
+    {
+      "epoch": 3.5623827744016805,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.19159511974695e-05,
+      "loss": 0.142,
+      "step": 11873
+    },
+    {
+      "epoch": 3.5626828719333785,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.1900888688055434e-05,
+      "loss": 0.1242,
+      "step": 11874
+    },
+    {
+      "epoch": 3.562982969465076,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.1885826178641364e-05,
+      "loss": 0.1006,
+      "step": 11875
+    },
+    {
+      "epoch": 3.563283066996774,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.1870763669227294e-05,
+      "loss": 0.0886,
+      "step": 11876
+    },
+    {
+      "epoch": 3.5635831645284717,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.1855701159813227e-05,
+      "loss": 0.1288,
+      "step": 11877
+    },
+    {
+      "epoch": 3.5638832620601697,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.1840638650399157e-05,
+      "loss": 0.1034,
+      "step": 11878
+    },
+    {
+      "epoch": 3.5641833595918673,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.1825576140985088e-05,
+      "loss": 0.0703,
+      "step": 11879
+    },
+    {
+      "epoch": 3.5644834571235653,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.181051363157102e-05,
+      "loss": 0.0899,
+      "step": 11880
+    },
+    {
+      "epoch": 3.564783554655263,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.179545112215695e-05,
+      "loss": 0.0768,
+      "step": 11881
+    },
+    {
+      "epoch": 3.565083652186961,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.178038861274288e-05,
+      "loss": 0.0972,
+      "step": 11882
+    },
+    {
+      "epoch": 3.5653837497186585,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.1765326103328815e-05,
+      "loss": 0.0884,
+      "step": 11883
+    },
+    {
+      "epoch": 3.5656838472503565,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.1750263593914745e-05,
+      "loss": 0.0837,
+      "step": 11884
+    },
+    {
+      "epoch": 3.565983944782054,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.173520108450068e-05,
+      "loss": 0.0979,
+      "step": 11885
+    },
+    {
+      "epoch": 3.566284042313752,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.172013857508661e-05,
+      "loss": 0.1113,
+      "step": 11886
+    },
+    {
+      "epoch": 3.5665841398454496,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.1705076065672542e-05,
+      "loss": 0.0924,
+      "step": 11887
+    },
+    {
+      "epoch": 3.5668842373771477,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.1690013556258472e-05,
+      "loss": 0.089,
+      "step": 11888
+    },
+    {
+      "epoch": 3.5671843349088452,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.1674951046844406e-05,
+      "loss": 0.1152,
+      "step": 11889
+    },
+    {
+      "epoch": 3.5674844324405433,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.1659888537430336e-05,
+      "loss": 0.1029,
+      "step": 11890
+    },
+    {
+      "epoch": 3.567784529972241,
+      "grad_norm": 0.875,
+      "learning_rate": 2.164482602801627e-05,
+      "loss": 0.1125,
+      "step": 11891
+    },
+    {
+      "epoch": 3.568084627503939,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.16297635186022e-05,
+      "loss": 0.0983,
+      "step": 11892
+    },
+    {
+      "epoch": 3.5683847250356364,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.1614701009188133e-05,
+      "loss": 0.1288,
+      "step": 11893
+    },
+    {
+      "epoch": 3.5686848225673344,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.1599638499774063e-05,
+      "loss": 0.1141,
+      "step": 11894
+    },
+    {
+      "epoch": 3.568984920099032,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.1584575990359997e-05,
+      "loss": 0.0967,
+      "step": 11895
+    },
+    {
+      "epoch": 3.56928501763073,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.1569513480945927e-05,
+      "loss": 0.0962,
+      "step": 11896
+    },
+    {
+      "epoch": 3.569585115162428,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.1554450971531857e-05,
+      "loss": 0.1359,
+      "step": 11897
+    },
+    {
+      "epoch": 3.5698852126941256,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.153938846211779e-05,
+      "loss": 0.1055,
+      "step": 11898
+    },
+    {
+      "epoch": 3.570185310225823,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.152432595270372e-05,
+      "loss": 0.1011,
+      "step": 11899
+    },
+    {
+      "epoch": 3.570485407757521,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.1509263443289654e-05,
+      "loss": 0.0785,
+      "step": 11900
+    },
+    {
+      "epoch": 3.5707855052892192,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.1494200933875585e-05,
+      "loss": 0.1085,
+      "step": 11901
+    },
+    {
+      "epoch": 3.571085602820917,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.1479138424461518e-05,
+      "loss": 0.0973,
+      "step": 11902
+    },
+    {
+      "epoch": 3.5713857003526144,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.1464075915047448e-05,
+      "loss": 0.0764,
+      "step": 11903
+    },
+    {
+      "epoch": 3.5716857978843124,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.1449013405633382e-05,
+      "loss": 0.0832,
+      "step": 11904
+    },
+    {
+      "epoch": 3.5719858954160104,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.1433950896219312e-05,
+      "loss": 0.0883,
+      "step": 11905
+    },
+    {
+      "epoch": 3.572285992947708,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.1418888386805242e-05,
+      "loss": 0.1205,
+      "step": 11906
+    },
+    {
+      "epoch": 3.5725860904794056,
+      "grad_norm": 0.75,
+      "learning_rate": 2.1403825877391172e-05,
+      "loss": 0.0852,
+      "step": 11907
+    },
+    {
+      "epoch": 3.5728861880111036,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.1388763367977106e-05,
+      "loss": 0.1048,
+      "step": 11908
+    },
+    {
+      "epoch": 3.5731862855428016,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.1373700858563036e-05,
+      "loss": 0.1392,
+      "step": 11909
+    },
+    {
+      "epoch": 3.573486383074499,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.135863834914897e-05,
+      "loss": 0.1012,
+      "step": 11910
+    },
+    {
+      "epoch": 3.5737864806061967,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.13435758397349e-05,
+      "loss": 0.0857,
+      "step": 11911
+    },
+    {
+      "epoch": 3.5740865781378948,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.1328513330320833e-05,
+      "loss": 0.094,
+      "step": 11912
+    },
+    {
+      "epoch": 3.574386675669593,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.1313450820906763e-05,
+      "loss": 0.097,
+      "step": 11913
+    },
+    {
+      "epoch": 3.5746867732012904,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.1298388311492697e-05,
+      "loss": 0.1083,
+      "step": 11914
+    },
+    {
+      "epoch": 3.5749868707329884,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.1283325802078627e-05,
+      "loss": 0.1152,
+      "step": 11915
+    },
+    {
+      "epoch": 3.575286968264686,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.1268263292664557e-05,
+      "loss": 0.1079,
+      "step": 11916
+    },
+    {
+      "epoch": 3.575587065796384,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.125320078325049e-05,
+      "loss": 0.0971,
+      "step": 11917
+    },
+    {
+      "epoch": 3.5758871633280815,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.123813827383642e-05,
+      "loss": 0.1021,
+      "step": 11918
+    },
+    {
+      "epoch": 3.5761872608597796,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.1223075764422354e-05,
+      "loss": 0.1013,
+      "step": 11919
+    },
+    {
+      "epoch": 3.576487358391477,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.1208013255008284e-05,
+      "loss": 0.086,
+      "step": 11920
+    },
+    {
+      "epoch": 3.576787455923175,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.1192950745594218e-05,
+      "loss": 0.1048,
+      "step": 11921
+    },
+    {
+      "epoch": 3.5770875534548727,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.1177888236180148e-05,
+      "loss": 0.1504,
+      "step": 11922
+    },
+    {
+      "epoch": 3.5773876509865707,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.116282572676608e-05,
+      "loss": 0.0937,
+      "step": 11923
+    },
+    {
+      "epoch": 3.5776877485182683,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.1147763217352012e-05,
+      "loss": 0.0772,
+      "step": 11924
+    },
+    {
+      "epoch": 3.5779878460499663,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.1132700707937945e-05,
+      "loss": 0.1334,
+      "step": 11925
+    },
+    {
+      "epoch": 3.578287943581664,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.1117638198523875e-05,
+      "loss": 0.1002,
+      "step": 11926
+    },
+    {
+      "epoch": 3.578588041113362,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.110257568910981e-05,
+      "loss": 0.1048,
+      "step": 11927
+    },
+    {
+      "epoch": 3.5788881386450595,
+      "grad_norm": 0.875,
+      "learning_rate": 2.108751317969574e-05,
+      "loss": 0.1094,
+      "step": 11928
+    },
+    {
+      "epoch": 3.5791882361767575,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.1072450670281673e-05,
+      "loss": 0.0966,
+      "step": 11929
+    },
+    {
+      "epoch": 3.579488333708455,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.1057388160867603e-05,
+      "loss": 0.1112,
+      "step": 11930
+    },
+    {
+      "epoch": 3.579788431240153,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.1042325651453533e-05,
+      "loss": 0.098,
+      "step": 11931
+    },
+    {
+      "epoch": 3.5800885287718507,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.1027263142039466e-05,
+      "loss": 0.0995,
+      "step": 11932
+    },
+    {
+      "epoch": 3.5803886263035487,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.1012200632625397e-05,
+      "loss": 0.1069,
+      "step": 11933
+    },
+    {
+      "epoch": 3.5806887238352463,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.0997138123211327e-05,
+      "loss": 0.1,
+      "step": 11934
+    },
+    {
+      "epoch": 3.5809888213669443,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.0982075613797257e-05,
+      "loss": 0.1425,
+      "step": 11935
+    },
+    {
+      "epoch": 3.5812889188986423,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.096701310438319e-05,
+      "loss": 0.0893,
+      "step": 11936
+    },
+    {
+      "epoch": 3.58158901643034,
+      "grad_norm": 0.75,
+      "learning_rate": 2.095195059496912e-05,
+      "loss": 0.0999,
+      "step": 11937
+    },
+    {
+      "epoch": 3.5818891139620375,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.0936888085555054e-05,
+      "loss": 0.0916,
+      "step": 11938
+    },
+    {
+      "epoch": 3.5821892114937355,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.0921825576140984e-05,
+      "loss": 0.1235,
+      "step": 11939
+    },
+    {
+      "epoch": 3.5824893090254335,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0906763066726918e-05,
+      "loss": 0.0969,
+      "step": 11940
+    },
+    {
+      "epoch": 3.582789406557131,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.0891700557312848e-05,
+      "loss": 0.0914,
+      "step": 11941
+    },
+    {
+      "epoch": 3.5830895040888286,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.087663804789878e-05,
+      "loss": 0.1173,
+      "step": 11942
+    },
+    {
+      "epoch": 3.5833896016205267,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.086157553848471e-05,
+      "loss": 0.0845,
+      "step": 11943
+    },
+    {
+      "epoch": 3.5836896991522247,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.0846513029070645e-05,
+      "loss": 0.09,
+      "step": 11944
+    },
+    {
+      "epoch": 3.5839897966839223,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.0831450519656575e-05,
+      "loss": 0.1083,
+      "step": 11945
+    },
+    {
+      "epoch": 3.58428989421562,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.081638801024251e-05,
+      "loss": 0.1001,
+      "step": 11946
+    },
+    {
+      "epoch": 3.584589991747318,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.080132550082844e-05,
+      "loss": 0.1742,
+      "step": 11947
+    },
+    {
+      "epoch": 3.584890089279016,
+      "grad_norm": 0.875,
+      "learning_rate": 2.0786262991414372e-05,
+      "loss": 0.1064,
+      "step": 11948
+    },
+    {
+      "epoch": 3.5851901868107134,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.0771200482000303e-05,
+      "loss": 0.0976,
+      "step": 11949
+    },
+    {
+      "epoch": 3.585490284342411,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.0756137972586233e-05,
+      "loss": 0.1089,
+      "step": 11950
+    },
+    {
+      "epoch": 3.585790381874109,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.0741075463172166e-05,
+      "loss": 0.1033,
+      "step": 11951
+    },
+    {
+      "epoch": 3.586090479405807,
+      "grad_norm": 0.75,
+      "learning_rate": 2.0726012953758096e-05,
+      "loss": 0.0957,
+      "step": 11952
+    },
+    {
+      "epoch": 3.5863905769375046,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.071095044434403e-05,
+      "loss": 0.0913,
+      "step": 11953
+    },
+    {
+      "epoch": 3.5866906744692026,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.069588793492996e-05,
+      "loss": 0.0927,
+      "step": 11954
+    },
+    {
+      "epoch": 3.5869907720009,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.0680825425515894e-05,
+      "loss": 0.1113,
+      "step": 11955
+    },
+    {
+      "epoch": 3.5872908695325982,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.0665762916101824e-05,
+      "loss": 0.1011,
+      "step": 11956
+    },
+    {
+      "epoch": 3.587590967064296,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.0650700406687757e-05,
+      "loss": 0.0834,
+      "step": 11957
+    },
+    {
+      "epoch": 3.587891064595994,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.0635637897273687e-05,
+      "loss": 0.1156,
+      "step": 11958
+    },
+    {
+      "epoch": 3.5881911621276914,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.062057538785962e-05,
+      "loss": 0.0881,
+      "step": 11959
+    },
+    {
+      "epoch": 3.5884912596593894,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.0605512878445548e-05,
+      "loss": 0.132,
+      "step": 11960
+    },
+    {
+      "epoch": 3.588791357191087,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.059045036903148e-05,
+      "loss": 0.1325,
+      "step": 11961
+    },
+    {
+      "epoch": 3.589091454722785,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.057538785961741e-05,
+      "loss": 0.0801,
+      "step": 11962
+    },
+    {
+      "epoch": 3.5893915522544826,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.0560325350203345e-05,
+      "loss": 0.1069,
+      "step": 11963
+    },
+    {
+      "epoch": 3.5896916497861806,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.0545262840789275e-05,
+      "loss": 0.0925,
+      "step": 11964
+    },
+    {
+      "epoch": 3.589991747317878,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.053020033137521e-05,
+      "loss": 0.1098,
+      "step": 11965
+    },
+    {
+      "epoch": 3.590291844849576,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.051513782196114e-05,
+      "loss": 0.1053,
+      "step": 11966
+    },
+    {
+      "epoch": 3.5905919423812738,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.0500075312547072e-05,
+      "loss": 0.1057,
+      "step": 11967
+    },
+    {
+      "epoch": 3.590892039912972,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.0485012803133002e-05,
+      "loss": 0.1208,
+      "step": 11968
+    },
+    {
+      "epoch": 3.5911921374446694,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.0469950293718932e-05,
+      "loss": 0.1074,
+      "step": 11969
+    },
+    {
+      "epoch": 3.5914922349763674,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.0454887784304866e-05,
+      "loss": 0.0972,
+      "step": 11970
+    },
+    {
+      "epoch": 3.591792332508065,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.0439825274890796e-05,
+      "loss": 0.0836,
+      "step": 11971
+    },
+    {
+      "epoch": 3.592092430039763,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.042476276547673e-05,
+      "loss": 0.0929,
+      "step": 11972
+    },
+    {
+      "epoch": 3.5923925275714605,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.040970025606266e-05,
+      "loss": 0.1049,
+      "step": 11973
+    },
+    {
+      "epoch": 3.5926926251031586,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.0394637746648593e-05,
+      "loss": 0.1325,
+      "step": 11974
+    },
+    {
+      "epoch": 3.5929927226348566,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.0379575237234523e-05,
+      "loss": 0.1106,
+      "step": 11975
+    },
+    {
+      "epoch": 3.593292820166554,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.0364512727820457e-05,
+      "loss": 0.145,
+      "step": 11976
+    },
+    {
+      "epoch": 3.5935929176982517,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.0349450218406387e-05,
+      "loss": 0.1333,
+      "step": 11977
+    },
+    {
+      "epoch": 3.5938930152299498,
+      "grad_norm": 0.75,
+      "learning_rate": 2.033438770899232e-05,
+      "loss": 0.0987,
+      "step": 11978
+    },
+    {
+      "epoch": 3.5941931127616478,
+      "grad_norm": 0.625,
+      "learning_rate": 2.031932519957825e-05,
+      "loss": 0.0763,
+      "step": 11979
+    },
+    {
+      "epoch": 3.5944932102933453,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.0304262690164184e-05,
+      "loss": 0.0946,
+      "step": 11980
+    },
+    {
+      "epoch": 3.594793307825043,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.0289200180750114e-05,
+      "loss": 0.094,
+      "step": 11981
+    },
+    {
+      "epoch": 3.595093405356741,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.0274137671336048e-05,
+      "loss": 0.1085,
+      "step": 11982
+    },
+    {
+      "epoch": 3.595393502888439,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.0259075161921978e-05,
+      "loss": 0.1237,
+      "step": 11983
+    },
+    {
+      "epoch": 3.5956936004201365,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.0244012652507908e-05,
+      "loss": 0.0885,
+      "step": 11984
+    },
+    {
+      "epoch": 3.595993697951834,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.0228950143093842e-05,
+      "loss": 0.1115,
+      "step": 11985
+    },
+    {
+      "epoch": 3.596293795483532,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.0213887633679772e-05,
+      "loss": 0.0963,
+      "step": 11986
+    },
+    {
+      "epoch": 3.59659389301523,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.0198825124265702e-05,
+      "loss": 0.1352,
+      "step": 11987
+    },
+    {
+      "epoch": 3.5968939905469277,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.0183762614851636e-05,
+      "loss": 0.1534,
+      "step": 11988
+    },
+    {
+      "epoch": 3.5971940880786253,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.0168700105437566e-05,
+      "loss": 0.11,
+      "step": 11989
+    },
+    {
+      "epoch": 3.5974941856103233,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.0153637596023496e-05,
+      "loss": 0.074,
+      "step": 11990
+    },
+    {
+      "epoch": 3.5977942831420213,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.013857508660943e-05,
+      "loss": 0.1345,
+      "step": 11991
+    },
+    {
+      "epoch": 3.598094380673719,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.012351257719536e-05,
+      "loss": 0.1131,
+      "step": 11992
+    },
+    {
+      "epoch": 3.598394478205417,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0108450067781293e-05,
+      "loss": 0.1014,
+      "step": 11993
+    },
+    {
+      "epoch": 3.5986945757371145,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.0093387558367223e-05,
+      "loss": 0.1056,
+      "step": 11994
+    },
+    {
+      "epoch": 3.5989946732688125,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.0078325048953157e-05,
+      "loss": 0.1053,
+      "step": 11995
+    },
+    {
+      "epoch": 3.59929477080051,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.0063262539539087e-05,
+      "loss": 0.1264,
+      "step": 11996
+    },
+    {
+      "epoch": 3.599594868332208,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.004820003012502e-05,
+      "loss": 0.0949,
+      "step": 11997
+    },
+    {
+      "epoch": 3.5998949658639057,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.003313752071095e-05,
+      "loss": 0.1016,
+      "step": 11998
+    },
+    {
+      "epoch": 3.6001950633956037,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0018075011296884e-05,
+      "loss": 0.094,
+      "step": 11999
+    },
+    {
+      "epoch": 3.6004951609273013,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.0003012501882814e-05,
+      "loss": 0.0942,
+      "step": 12000
+    },
+    {
+      "epoch": 3.6007952584589993,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9987949992468748e-05,
+      "loss": 0.1068,
+      "step": 12001
+    },
+    {
+      "epoch": 3.601095355990697,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9972887483054678e-05,
+      "loss": 0.0973,
+      "step": 12002
+    },
+    {
+      "epoch": 3.601395453522395,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.995782497364061e-05,
+      "loss": 0.0981,
+      "step": 12003
+    },
+    {
+      "epoch": 3.6016955510540924,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.994276246422654e-05,
+      "loss": 0.1279,
+      "step": 12004
+    },
+    {
+      "epoch": 3.6019956485857905,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9927699954812472e-05,
+      "loss": 0.0929,
+      "step": 12005
+    },
+    {
+      "epoch": 3.602295746117488,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9912637445398405e-05,
+      "loss": 0.0947,
+      "step": 12006
+    },
+    {
+      "epoch": 3.602595843649186,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9897574935984335e-05,
+      "loss": 0.1117,
+      "step": 12007
+    },
+    {
+      "epoch": 3.6028959411808836,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.988251242657027e-05,
+      "loss": 0.0904,
+      "step": 12008
+    },
+    {
+      "epoch": 3.6031960387125817,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.98674499171562e-05,
+      "loss": 0.0964,
+      "step": 12009
+    },
+    {
+      "epoch": 3.6034961362442792,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9852387407742133e-05,
+      "loss": 0.0866,
+      "step": 12010
+    },
+    {
+      "epoch": 3.6037962337759772,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9837324898328063e-05,
+      "loss": 0.1055,
+      "step": 12011
+    },
+    {
+      "epoch": 3.604096331307675,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9822262388913996e-05,
+      "loss": 0.0988,
+      "step": 12012
+    },
+    {
+      "epoch": 3.604396428839373,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9807199879499926e-05,
+      "loss": 0.1187,
+      "step": 12013
+    },
+    {
+      "epoch": 3.604696526371071,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9792137370085857e-05,
+      "loss": 0.0887,
+      "step": 12014
+    },
+    {
+      "epoch": 3.6049966239027684,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9777074860671787e-05,
+      "loss": 0.1558,
+      "step": 12015
+    },
+    {
+      "epoch": 3.605296721434466,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.976201235125772e-05,
+      "loss": 0.1064,
+      "step": 12016
+    },
+    {
+      "epoch": 3.605596818966164,
+      "grad_norm": 0.625,
+      "learning_rate": 1.974694984184365e-05,
+      "loss": 0.0824,
+      "step": 12017
+    },
+    {
+      "epoch": 3.605896916497862,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9731887332429584e-05,
+      "loss": 0.12,
+      "step": 12018
+    },
+    {
+      "epoch": 3.6061970140295596,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9716824823015514e-05,
+      "loss": 0.1003,
+      "step": 12019
+    },
+    {
+      "epoch": 3.606497111561257,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9701762313601448e-05,
+      "loss": 0.1138,
+      "step": 12020
+    },
+    {
+      "epoch": 3.606797209092955,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9686699804187378e-05,
+      "loss": 0.1223,
+      "step": 12021
+    },
+    {
+      "epoch": 3.607097306624653,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.967163729477331e-05,
+      "loss": 0.1134,
+      "step": 12022
+    },
+    {
+      "epoch": 3.607397404156351,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.965657478535924e-05,
+      "loss": 0.0901,
+      "step": 12023
+    },
+    {
+      "epoch": 3.6076975016880484,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.964151227594517e-05,
+      "loss": 0.1362,
+      "step": 12024
+    },
+    {
+      "epoch": 3.6079975992197464,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9626449766531105e-05,
+      "loss": 0.1017,
+      "step": 12025
+    },
+    {
+      "epoch": 3.6082976967514444,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.9611387257117035e-05,
+      "loss": 0.108,
+      "step": 12026
+    },
+    {
+      "epoch": 3.608597794283142,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.959632474770297e-05,
+      "loss": 0.1227,
+      "step": 12027
+    },
+    {
+      "epoch": 3.6088978918148396,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.95812622382889e-05,
+      "loss": 0.081,
+      "step": 12028
+    },
+    {
+      "epoch": 3.6091979893465376,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9566199728874832e-05,
+      "loss": 0.0961,
+      "step": 12029
+    },
+    {
+      "epoch": 3.6094980868782356,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9551137219460763e-05,
+      "loss": 0.107,
+      "step": 12030
+    },
+    {
+      "epoch": 3.609798184409933,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9536074710046696e-05,
+      "loss": 0.0987,
+      "step": 12031
+    },
+    {
+      "epoch": 3.610098281941631,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9521012200632626e-05,
+      "loss": 0.0981,
+      "step": 12032
+    },
+    {
+      "epoch": 3.6103983794733288,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.950594969121856e-05,
+      "loss": 0.0949,
+      "step": 12033
+    },
+    {
+      "epoch": 3.6106984770050268,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.949088718180449e-05,
+      "loss": 0.0791,
+      "step": 12034
+    },
+    {
+      "epoch": 3.6109985745367243,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9475824672390423e-05,
+      "loss": 0.0864,
+      "step": 12035
+    },
+    {
+      "epoch": 3.6112986720684224,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9460762162976354e-05,
+      "loss": 0.1187,
+      "step": 12036
+    },
+    {
+      "epoch": 3.61159876960012,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9445699653562287e-05,
+      "loss": 0.1034,
+      "step": 12037
+    },
+    {
+      "epoch": 3.611898867131818,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.9430637144148217e-05,
+      "loss": 0.126,
+      "step": 12038
+    },
+    {
+      "epoch": 3.6121989646635155,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9415574634734147e-05,
+      "loss": 0.1049,
+      "step": 12039
+    },
+    {
+      "epoch": 3.6124990621952136,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.940051212532008e-05,
+      "loss": 0.1104,
+      "step": 12040
+    },
+    {
+      "epoch": 3.612799159726911,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.938544961590601e-05,
+      "loss": 0.0885,
+      "step": 12041
+    },
+    {
+      "epoch": 3.613099257258609,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.937038710649194e-05,
+      "loss": 0.1141,
+      "step": 12042
+    },
+    {
+      "epoch": 3.6133993547903067,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.935532459707787e-05,
+      "loss": 0.0998,
+      "step": 12043
+    },
+    {
+      "epoch": 3.6136994523220047,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9340262087663805e-05,
+      "loss": 0.1003,
+      "step": 12044
+    },
+    {
+      "epoch": 3.6139995498537023,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9325199578249735e-05,
+      "loss": 0.1414,
+      "step": 12045
+    },
+    {
+      "epoch": 3.6142996473854003,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.931013706883567e-05,
+      "loss": 0.0985,
+      "step": 12046
+    },
+    {
+      "epoch": 3.614599744917098,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.92950745594216e-05,
+      "loss": 0.1153,
+      "step": 12047
+    },
+    {
+      "epoch": 3.614899842448796,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9280012050007532e-05,
+      "loss": 0.1119,
+      "step": 12048
+    },
+    {
+      "epoch": 3.615199939980494,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9264949540593462e-05,
+      "loss": 0.081,
+      "step": 12049
+    },
+    {
+      "epoch": 3.6155000375121915,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9249887031179396e-05,
+      "loss": 0.1029,
+      "step": 12050
+    },
+    {
+      "epoch": 3.615800135043889,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9234824521765326e-05,
+      "loss": 0.0958,
+      "step": 12051
+    },
+    {
+      "epoch": 3.616100232575587,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.921976201235126e-05,
+      "loss": 0.1133,
+      "step": 12052
+    },
+    {
+      "epoch": 3.616400330107285,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.920469950293719e-05,
+      "loss": 0.0976,
+      "step": 12053
+    },
+    {
+      "epoch": 3.6167004276389827,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9189636993523123e-05,
+      "loss": 0.1202,
+      "step": 12054
+    },
+    {
+      "epoch": 3.6170005251706803,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9174574484109053e-05,
+      "loss": 0.1162,
+      "step": 12055
+    },
+    {
+      "epoch": 3.6173006227023783,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9159511974694987e-05,
+      "loss": 0.0743,
+      "step": 12056
+    },
+    {
+      "epoch": 3.6176007202340763,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9144449465280917e-05,
+      "loss": 0.1134,
+      "step": 12057
+    },
+    {
+      "epoch": 3.617900817765774,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9129386955866847e-05,
+      "loss": 0.1062,
+      "step": 12058
+    },
+    {
+      "epoch": 3.6182009152974715,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.911432444645278e-05,
+      "loss": 0.0997,
+      "step": 12059
+    },
+    {
+      "epoch": 3.6185010128291695,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.909926193703871e-05,
+      "loss": 0.1102,
+      "step": 12060
+    },
+    {
+      "epoch": 3.6188011103608675,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9084199427624644e-05,
+      "loss": 0.1006,
+      "step": 12061
+    },
+    {
+      "epoch": 3.619101207892565,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9069136918210574e-05,
+      "loss": 0.1148,
+      "step": 12062
+    },
+    {
+      "epoch": 3.6194013054242626,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9054074408796508e-05,
+      "loss": 0.1025,
+      "step": 12063
+    },
+    {
+      "epoch": 3.6197014029559607,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9039011899382438e-05,
+      "loss": 0.0794,
+      "step": 12064
+    },
+    {
+      "epoch": 3.6200015004876587,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.902394938996837e-05,
+      "loss": 0.1342,
+      "step": 12065
+    },
+    {
+      "epoch": 3.6203015980193562,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9008886880554302e-05,
+      "loss": 0.0931,
+      "step": 12066
+    },
+    {
+      "epoch": 3.620601695551054,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.8993824371140235e-05,
+      "loss": 0.1147,
+      "step": 12067
+    },
+    {
+      "epoch": 3.620901793082752,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8978761861726162e-05,
+      "loss": 0.0843,
+      "step": 12068
+    },
+    {
+      "epoch": 3.62120189061445,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8963699352312096e-05,
+      "loss": 0.1245,
+      "step": 12069
+    },
+    {
+      "epoch": 3.6215019881461474,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.8948636842898026e-05,
+      "loss": 0.1017,
+      "step": 12070
+    },
+    {
+      "epoch": 3.6218020856778455,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.893357433348396e-05,
+      "loss": 0.0818,
+      "step": 12071
+    },
+    {
+      "epoch": 3.622102183209543,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.891851182406989e-05,
+      "loss": 0.1327,
+      "step": 12072
+    },
+    {
+      "epoch": 3.622402280741241,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8903449314655823e-05,
+      "loss": 0.0916,
+      "step": 12073
+    },
+    {
+      "epoch": 3.6227023782729386,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8888386805241753e-05,
+      "loss": 0.0982,
+      "step": 12074
+    },
+    {
+      "epoch": 3.6230024758046366,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.8873324295827687e-05,
+      "loss": 0.117,
+      "step": 12075
+    },
+    {
+      "epoch": 3.623302573336334,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8858261786413617e-05,
+      "loss": 0.1222,
+      "step": 12076
+    },
+    {
+      "epoch": 3.6236026708680322,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8843199276999547e-05,
+      "loss": 0.1077,
+      "step": 12077
+    },
+    {
+      "epoch": 3.62390276839973,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.882813676758548e-05,
+      "loss": 0.0902,
+      "step": 12078
+    },
+    {
+      "epoch": 3.624202865931428,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.881307425817141e-05,
+      "loss": 0.0969,
+      "step": 12079
+    },
+    {
+      "epoch": 3.6245029634631254,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8798011748757344e-05,
+      "loss": 0.1066,
+      "step": 12080
+    },
+    {
+      "epoch": 3.6248030609948234,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8782949239343274e-05,
+      "loss": 0.0946,
+      "step": 12081
+    },
+    {
+      "epoch": 3.625103158526521,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.8767886729929208e-05,
+      "loss": 0.1269,
+      "step": 12082
+    },
+    {
+      "epoch": 3.625403256058219,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8752824220515138e-05,
+      "loss": 0.0822,
+      "step": 12083
+    },
+    {
+      "epoch": 3.6257033535899166,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.873776171110107e-05,
+      "loss": 0.1025,
+      "step": 12084
+    },
+    {
+      "epoch": 3.6260034511216146,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8722699201687e-05,
+      "loss": 0.0933,
+      "step": 12085
+    },
+    {
+      "epoch": 3.626303548653312,
+      "grad_norm": 0.875,
+      "learning_rate": 1.8707636692272935e-05,
+      "loss": 0.1107,
+      "step": 12086
+    },
+    {
+      "epoch": 3.62660364618501,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8692574182858865e-05,
+      "loss": 0.1116,
+      "step": 12087
+    },
+    {
+      "epoch": 3.626903743716708,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.86775116734448e-05,
+      "loss": 0.0993,
+      "step": 12088
+    },
+    {
+      "epoch": 3.627203841248406,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.866244916403073e-05,
+      "loss": 0.1018,
+      "step": 12089
+    },
+    {
+      "epoch": 3.6275039387801034,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8647386654616662e-05,
+      "loss": 0.1207,
+      "step": 12090
+    },
+    {
+      "epoch": 3.6278040363118014,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8632324145202593e-05,
+      "loss": 0.1082,
+      "step": 12091
+    },
+    {
+      "epoch": 3.6281041338434994,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8617261635788523e-05,
+      "loss": 0.14,
+      "step": 12092
+    },
+    {
+      "epoch": 3.628404231375197,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.8602199126374456e-05,
+      "loss": 0.0887,
+      "step": 12093
+    },
+    {
+      "epoch": 3.6287043289068945,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8587136616960386e-05,
+      "loss": 0.1152,
+      "step": 12094
+    },
+    {
+      "epoch": 3.6290044264385926,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.857207410754632e-05,
+      "loss": 0.1008,
+      "step": 12095
+    },
+    {
+      "epoch": 3.6293045239702906,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.855701159813225e-05,
+      "loss": 0.074,
+      "step": 12096
+    },
+    {
+      "epoch": 3.629604621501988,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.854194908871818e-05,
+      "loss": 0.0992,
+      "step": 12097
+    },
+    {
+      "epoch": 3.6299047190336857,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.852688657930411e-05,
+      "loss": 0.1105,
+      "step": 12098
+    },
+    {
+      "epoch": 3.6302048165653837,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.8511824069890044e-05,
+      "loss": 0.1212,
+      "step": 12099
+    },
+    {
+      "epoch": 3.6305049140970818,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8496761560475974e-05,
+      "loss": 0.1173,
+      "step": 12100
+    },
+    {
+      "epoch": 3.6308050116287793,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8481699051061908e-05,
+      "loss": 0.104,
+      "step": 12101
+    },
+    {
+      "epoch": 3.631105109160477,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8466636541647838e-05,
+      "loss": 0.0972,
+      "step": 12102
+    },
+    {
+      "epoch": 3.631405206692175,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.845157403223377e-05,
+      "loss": 0.0984,
+      "step": 12103
+    },
+    {
+      "epoch": 3.631705304223873,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.84365115228197e-05,
+      "loss": 0.1034,
+      "step": 12104
+    },
+    {
+      "epoch": 3.6320054017555705,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.8421449013405635e-05,
+      "loss": 0.1043,
+      "step": 12105
+    },
+    {
+      "epoch": 3.632305499287268,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8406386503991565e-05,
+      "loss": 0.1046,
+      "step": 12106
+    },
+    {
+      "epoch": 3.632605596818966,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.83913239945775e-05,
+      "loss": 0.1095,
+      "step": 12107
+    },
+    {
+      "epoch": 3.632905694350664,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.837626148516343e-05,
+      "loss": 0.1172,
+      "step": 12108
+    },
+    {
+      "epoch": 3.6332057918823617,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8361198975749362e-05,
+      "loss": 0.1028,
+      "step": 12109
+    },
+    {
+      "epoch": 3.6335058894140597,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.8346136466335292e-05,
+      "loss": 0.1192,
+      "step": 12110
+    },
+    {
+      "epoch": 3.6338059869457573,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8331073956921223e-05,
+      "loss": 0.0972,
+      "step": 12111
+    },
+    {
+      "epoch": 3.6341060844774553,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8316011447507156e-05,
+      "loss": 0.0665,
+      "step": 12112
+    },
+    {
+      "epoch": 3.634406182009153,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8300948938093086e-05,
+      "loss": 0.1018,
+      "step": 12113
+    },
+    {
+      "epoch": 3.634706279540851,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.828588642867902e-05,
+      "loss": 0.0956,
+      "step": 12114
+    },
+    {
+      "epoch": 3.6350063770725485,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.827082391926495e-05,
+      "loss": 0.144,
+      "step": 12115
+    },
+    {
+      "epoch": 3.6353064746042465,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8255761409850883e-05,
+      "loss": 0.1302,
+      "step": 12116
+    },
+    {
+      "epoch": 3.635606572135944,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8240698900436814e-05,
+      "loss": 0.0974,
+      "step": 12117
+    },
+    {
+      "epoch": 3.635906669667642,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.8225636391022747e-05,
+      "loss": 0.1037,
+      "step": 12118
+    },
+    {
+      "epoch": 3.6362067671993397,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.8210573881608677e-05,
+      "loss": 0.108,
+      "step": 12119
+    },
+    {
+      "epoch": 3.6365068647310377,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.819551137219461e-05,
+      "loss": 0.0831,
+      "step": 12120
+    },
+    {
+      "epoch": 3.6368069622627353,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.818044886278054e-05,
+      "loss": 0.0901,
+      "step": 12121
+    },
+    {
+      "epoch": 3.6371070597944333,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8165386353366474e-05,
+      "loss": 0.1172,
+      "step": 12122
+    },
+    {
+      "epoch": 3.637407157326131,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.81503238439524e-05,
+      "loss": 0.0856,
+      "step": 12123
+    },
+    {
+      "epoch": 3.637707254857829,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8135261334538335e-05,
+      "loss": 0.102,
+      "step": 12124
+    },
+    {
+      "epoch": 3.6380073523895264,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8120198825124265e-05,
+      "loss": 0.0892,
+      "step": 12125
+    },
+    {
+      "epoch": 3.6383074499212245,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.81051363157102e-05,
+      "loss": 0.0897,
+      "step": 12126
+    },
+    {
+      "epoch": 3.6386075474529225,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.809007380629613e-05,
+      "loss": 0.0957,
+      "step": 12127
+    },
+    {
+      "epoch": 3.63890764498462,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8075011296882062e-05,
+      "loss": 0.0836,
+      "step": 12128
+    },
+    {
+      "epoch": 3.6392077425163176,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8059948787467992e-05,
+      "loss": 0.113,
+      "step": 12129
+    },
+    {
+      "epoch": 3.6395078400480156,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.8044886278053926e-05,
+      "loss": 0.1123,
+      "step": 12130
+    },
+    {
+      "epoch": 3.6398079375797137,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.8029823768639856e-05,
+      "loss": 0.1033,
+      "step": 12131
+    },
+    {
+      "epoch": 3.6401080351114112,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8014761259225786e-05,
+      "loss": 0.108,
+      "step": 12132
+    },
+    {
+      "epoch": 3.640408132643109,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.799969874981172e-05,
+      "loss": 0.1071,
+      "step": 12133
+    },
+    {
+      "epoch": 3.640708230174807,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.798463624039765e-05,
+      "loss": 0.1154,
+      "step": 12134
+    },
+    {
+      "epoch": 3.641008327706505,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7969573730983583e-05,
+      "loss": 0.084,
+      "step": 12135
+    },
+    {
+      "epoch": 3.6413084252382024,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7954511221569513e-05,
+      "loss": 0.1033,
+      "step": 12136
+    },
+    {
+      "epoch": 3.6416085227699,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7939448712155447e-05,
+      "loss": 0.0818,
+      "step": 12137
+    },
+    {
+      "epoch": 3.641908620301598,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.7924386202741377e-05,
+      "loss": 0.1242,
+      "step": 12138
+    },
+    {
+      "epoch": 3.642208717833296,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.790932369332731e-05,
+      "loss": 0.1067,
+      "step": 12139
+    },
+    {
+      "epoch": 3.6425088153649936,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.789426118391324e-05,
+      "loss": 0.1036,
+      "step": 12140
+    },
+    {
+      "epoch": 3.642808912896691,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.7879198674499174e-05,
+      "loss": 0.0867,
+      "step": 12141
+    },
+    {
+      "epoch": 3.643109010428389,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.7864136165085104e-05,
+      "loss": 0.1136,
+      "step": 12142
+    },
+    {
+      "epoch": 3.643409107960087,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7849073655671038e-05,
+      "loss": 0.0968,
+      "step": 12143
+    },
+    {
+      "epoch": 3.643709205491785,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.7834011146256968e-05,
+      "loss": 0.1103,
+      "step": 12144
+    },
+    {
+      "epoch": 3.6440093030234824,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.7818948636842898e-05,
+      "loss": 0.1213,
+      "step": 12145
+    },
+    {
+      "epoch": 3.6443094005551804,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7803886127428832e-05,
+      "loss": 0.0838,
+      "step": 12146
+    },
+    {
+      "epoch": 3.6446094980868784,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.7788823618014762e-05,
+      "loss": 0.1138,
+      "step": 12147
+    },
+    {
+      "epoch": 3.644909595618576,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7773761108600695e-05,
+      "loss": 0.0907,
+      "step": 12148
+    },
+    {
+      "epoch": 3.645209693150274,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.7758698599186626e-05,
+      "loss": 0.1169,
+      "step": 12149
+    },
+    {
+      "epoch": 3.6455097906819716,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7743636089772556e-05,
+      "loss": 0.0971,
+      "step": 12150
+    },
+    {
+      "epoch": 3.6458098882136696,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7728573580358486e-05,
+      "loss": 0.0965,
+      "step": 12151
+    },
+    {
+      "epoch": 3.646109985745367,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.771351107094442e-05,
+      "loss": 0.1019,
+      "step": 12152
+    },
+    {
+      "epoch": 3.646410083277065,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.769844856153035e-05,
+      "loss": 0.1576,
+      "step": 12153
+    },
+    {
+      "epoch": 3.6467101808087627,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.7683386052116283e-05,
+      "loss": 0.1209,
+      "step": 12154
+    },
+    {
+      "epoch": 3.6470102783404608,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7668323542702213e-05,
+      "loss": 0.1243,
+      "step": 12155
+    },
+    {
+      "epoch": 3.6473103758721583,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7653261033288147e-05,
+      "loss": 0.0893,
+      "step": 12156
+    },
+    {
+      "epoch": 3.6476104734038564,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7638198523874077e-05,
+      "loss": 0.0898,
+      "step": 12157
+    },
+    {
+      "epoch": 3.647910570935554,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.762313601446001e-05,
+      "loss": 0.1152,
+      "step": 12158
+    },
+    {
+      "epoch": 3.648210668467252,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.760807350504594e-05,
+      "loss": 0.1047,
+      "step": 12159
+    },
+    {
+      "epoch": 3.6485107659989495,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.7593010995631874e-05,
+      "loss": 0.1278,
+      "step": 12160
+    },
+    {
+      "epoch": 3.6488108635306475,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7577948486217804e-05,
+      "loss": 0.0887,
+      "step": 12161
+    },
+    {
+      "epoch": 3.649110961062345,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7562885976803738e-05,
+      "loss": 0.1171,
+      "step": 12162
+    },
+    {
+      "epoch": 3.649411058594043,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.7547823467389668e-05,
+      "loss": 0.1253,
+      "step": 12163
+    },
+    {
+      "epoch": 3.6497111561257407,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.75327609579756e-05,
+      "loss": 0.118,
+      "step": 12164
+    },
+    {
+      "epoch": 3.6500112536574387,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.751769844856153e-05,
+      "loss": 0.0924,
+      "step": 12165
+    },
+    {
+      "epoch": 3.6503113511891367,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.750263593914746e-05,
+      "loss": 0.1192,
+      "step": 12166
+    },
+    {
+      "epoch": 3.6506114487208343,
+      "grad_norm": 0.875,
+      "learning_rate": 1.7487573429733395e-05,
+      "loss": 0.1224,
+      "step": 12167
+    },
+    {
+      "epoch": 3.650911546252532,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7472510920319325e-05,
+      "loss": 0.0809,
+      "step": 12168
+    },
+    {
+      "epoch": 3.65121164378423,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.745744841090526e-05,
+      "loss": 0.1028,
+      "step": 12169
+    },
+    {
+      "epoch": 3.651511741315928,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.744238590149119e-05,
+      "loss": 0.0944,
+      "step": 12170
+    },
+    {
+      "epoch": 3.6518118388476255,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7427323392077122e-05,
+      "loss": 0.088,
+      "step": 12171
+    },
+    {
+      "epoch": 3.652111936379323,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7412260882663053e-05,
+      "loss": 0.0906,
+      "step": 12172
+    },
+    {
+      "epoch": 3.652412033911021,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7397198373248986e-05,
+      "loss": 0.0818,
+      "step": 12173
+    },
+    {
+      "epoch": 3.652712131442719,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7382135863834916e-05,
+      "loss": 0.0862,
+      "step": 12174
+    },
+    {
+      "epoch": 3.6530122289744167,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.736707335442085e-05,
+      "loss": 0.09,
+      "step": 12175
+    },
+    {
+      "epoch": 3.6533123265061143,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.735201084500678e-05,
+      "loss": 0.0992,
+      "step": 12176
+    },
+    {
+      "epoch": 3.6536124240378123,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.733694833559271e-05,
+      "loss": 0.0806,
+      "step": 12177
+    },
+    {
+      "epoch": 3.6539125215695103,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.732188582617864e-05,
+      "loss": 0.117,
+      "step": 12178
+    },
+    {
+      "epoch": 3.654212619101208,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.7306823316764574e-05,
+      "loss": 0.1417,
+      "step": 12179
+    },
+    {
+      "epoch": 3.6545127166329054,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7291760807350504e-05,
+      "loss": 0.0887,
+      "step": 12180
+    },
+    {
+      "epoch": 3.6548128141646035,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.7276698297936437e-05,
+      "loss": 0.1205,
+      "step": 12181
+    },
+    {
+      "epoch": 3.6551129116963015,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.7261635788522368e-05,
+      "loss": 0.1042,
+      "step": 12182
+    },
+    {
+      "epoch": 3.655413009227999,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.72465732791083e-05,
+      "loss": 0.1006,
+      "step": 12183
+    },
+    {
+      "epoch": 3.6557131067596966,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.723151076969423e-05,
+      "loss": 0.0897,
+      "step": 12184
+    },
+    {
+      "epoch": 3.6560132042913946,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.721644826028016e-05,
+      "loss": 0.1114,
+      "step": 12185
+    },
+    {
+      "epoch": 3.6563133018230927,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.7201385750866095e-05,
+      "loss": 0.1249,
+      "step": 12186
+    },
+    {
+      "epoch": 3.6566133993547902,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.7186323241452025e-05,
+      "loss": 0.1214,
+      "step": 12187
+    },
+    {
+      "epoch": 3.6569134968864883,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.717126073203796e-05,
+      "loss": 0.0944,
+      "step": 12188
+    },
+    {
+      "epoch": 3.657213594418186,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.715619822262389e-05,
+      "loss": 0.1173,
+      "step": 12189
+    },
+    {
+      "epoch": 3.657513691949884,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7141135713209822e-05,
+      "loss": 0.0804,
+      "step": 12190
+    },
+    {
+      "epoch": 3.6578137894815814,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7126073203795752e-05,
+      "loss": 0.1099,
+      "step": 12191
+    },
+    {
+      "epoch": 3.6581138870132794,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.7111010694381686e-05,
+      "loss": 0.112,
+      "step": 12192
+    },
+    {
+      "epoch": 3.658413984544977,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7095948184967616e-05,
+      "loss": 0.0936,
+      "step": 12193
+    },
+    {
+      "epoch": 3.658714082076675,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.708088567555355e-05,
+      "loss": 0.0833,
+      "step": 12194
+    },
+    {
+      "epoch": 3.6590141796083726,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.706582316613948e-05,
+      "loss": 0.097,
+      "step": 12195
+    },
+    {
+      "epoch": 3.6593142771400706,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.7050760656725413e-05,
+      "loss": 0.0955,
+      "step": 12196
+    },
+    {
+      "epoch": 3.659614374671768,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7035698147311343e-05,
+      "loss": 0.0966,
+      "step": 12197
+    },
+    {
+      "epoch": 3.659914472203466,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7020635637897277e-05,
+      "loss": 0.0937,
+      "step": 12198
+    },
+    {
+      "epoch": 3.660214569735164,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7005573128483207e-05,
+      "loss": 0.0884,
+      "step": 12199
+    },
+    {
+      "epoch": 3.660514667266862,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.6990510619069137e-05,
+      "loss": 0.1124,
+      "step": 12200
+    },
+    {
+      "epoch": 3.6608147647985594,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.697544810965507e-05,
+      "loss": 0.1051,
+      "step": 12201
+    },
+    {
+      "epoch": 3.6611148623302574,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6960385600241e-05,
+      "loss": 0.1013,
+      "step": 12202
+    },
+    {
+      "epoch": 3.661414959861955,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6945323090826934e-05,
+      "loss": 0.0864,
+      "step": 12203
+    },
+    {
+      "epoch": 3.661715057393653,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.693026058141286e-05,
+      "loss": 0.0942,
+      "step": 12204
+    },
+    {
+      "epoch": 3.662015154925351,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.6915198071998795e-05,
+      "loss": 0.1191,
+      "step": 12205
+    },
+    {
+      "epoch": 3.6623152524570486,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.6900135562584725e-05,
+      "loss": 0.116,
+      "step": 12206
+    },
+    {
+      "epoch": 3.662615349988746,
+      "grad_norm": 0.75,
+      "learning_rate": 1.688507305317066e-05,
+      "loss": 0.0986,
+      "step": 12207
+    },
+    {
+      "epoch": 3.662915447520444,
+      "grad_norm": 0.875,
+      "learning_rate": 1.687001054375659e-05,
+      "loss": 0.1062,
+      "step": 12208
+    },
+    {
+      "epoch": 3.663215545052142,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.6854948034342522e-05,
+      "loss": 0.115,
+      "step": 12209
+    },
+    {
+      "epoch": 3.6635156425838398,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.6839885524928452e-05,
+      "loss": 0.0948,
+      "step": 12210
+    },
+    {
+      "epoch": 3.6638157401155373,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.6824823015514386e-05,
+      "loss": 0.1005,
+      "step": 12211
+    },
+    {
+      "epoch": 3.6641158376472354,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6809760506100316e-05,
+      "loss": 0.102,
+      "step": 12212
+    },
+    {
+      "epoch": 3.6644159351789334,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.679469799668625e-05,
+      "loss": 0.1005,
+      "step": 12213
+    },
+    {
+      "epoch": 3.664716032710631,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.677963548727218e-05,
+      "loss": 0.0957,
+      "step": 12214
+    },
+    {
+      "epoch": 3.6650161302423285,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6764572977858113e-05,
+      "loss": 0.0858,
+      "step": 12215
+    },
+    {
+      "epoch": 3.6653162277740265,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6749510468444043e-05,
+      "loss": 0.0738,
+      "step": 12216
+    },
+    {
+      "epoch": 3.6656163253057246,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.6734447959029977e-05,
+      "loss": 0.1029,
+      "step": 12217
+    },
+    {
+      "epoch": 3.665916422837422,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6719385449615907e-05,
+      "loss": 0.1129,
+      "step": 12218
+    },
+    {
+      "epoch": 3.6662165203691197,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.6704322940201837e-05,
+      "loss": 0.0837,
+      "step": 12219
+    },
+    {
+      "epoch": 3.6665166179008177,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.668926043078777e-05,
+      "loss": 0.0884,
+      "step": 12220
+    },
+    {
+      "epoch": 3.6668167154325157,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.66741979213737e-05,
+      "loss": 0.0867,
+      "step": 12221
+    },
+    {
+      "epoch": 3.6671168129642133,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6659135411959634e-05,
+      "loss": 0.1355,
+      "step": 12222
+    },
+    {
+      "epoch": 3.667416910495911,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.6644072902545564e-05,
+      "loss": 0.0799,
+      "step": 12223
+    },
+    {
+      "epoch": 3.667717008027609,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6629010393131498e-05,
+      "loss": 0.0948,
+      "step": 12224
+    },
+    {
+      "epoch": 3.668017105559307,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.6613947883717428e-05,
+      "loss": 0.1112,
+      "step": 12225
+    },
+    {
+      "epoch": 3.6683172030910045,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.659888537430336e-05,
+      "loss": 0.0979,
+      "step": 12226
+    },
+    {
+      "epoch": 3.6686173006227025,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6583822864889292e-05,
+      "loss": 0.1059,
+      "step": 12227
+    },
+    {
+      "epoch": 3.6689173981544,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6568760355475225e-05,
+      "loss": 0.0959,
+      "step": 12228
+    },
+    {
+      "epoch": 3.669217495686098,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6553697846061155e-05,
+      "loss": 0.0949,
+      "step": 12229
+    },
+    {
+      "epoch": 3.6695175932177957,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.653863533664709e-05,
+      "loss": 0.0872,
+      "step": 12230
+    },
+    {
+      "epoch": 3.6698176907494937,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6523572827233016e-05,
+      "loss": 0.0949,
+      "step": 12231
+    },
+    {
+      "epoch": 3.6701177882811913,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.650851031781895e-05,
+      "loss": 0.1103,
+      "step": 12232
+    },
+    {
+      "epoch": 3.6704178858128893,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.649344780840488e-05,
+      "loss": 0.1324,
+      "step": 12233
+    },
+    {
+      "epoch": 3.670717983344587,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.6478385298990813e-05,
+      "loss": 0.1518,
+      "step": 12234
+    },
+    {
+      "epoch": 3.671018080876285,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6463322789576743e-05,
+      "loss": 0.0933,
+      "step": 12235
+    },
+    {
+      "epoch": 3.6713181784079825,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.6448260280162677e-05,
+      "loss": 0.0921,
+      "step": 12236
+    },
+    {
+      "epoch": 3.6716182759396805,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.6433197770748607e-05,
+      "loss": 0.096,
+      "step": 12237
+    },
+    {
+      "epoch": 3.671918373471378,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6418135261334537e-05,
+      "loss": 0.0783,
+      "step": 12238
+    },
+    {
+      "epoch": 3.672218471003076,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.640307275192047e-05,
+      "loss": 0.1044,
+      "step": 12239
+    },
+    {
+      "epoch": 3.6725185685347737,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.63880102425064e-05,
+      "loss": 0.1107,
+      "step": 12240
+    },
+    {
+      "epoch": 3.6728186660664717,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6372947733092334e-05,
+      "loss": 0.0936,
+      "step": 12241
+    },
+    {
+      "epoch": 3.6731187635981692,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.6357885223678264e-05,
+      "loss": 0.0979,
+      "step": 12242
+    },
+    {
+      "epoch": 3.6734188611298673,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6342822714264198e-05,
+      "loss": 0.0889,
+      "step": 12243
+    },
+    {
+      "epoch": 3.6737189586615653,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.6327760204850128e-05,
+      "loss": 0.0896,
+      "step": 12244
+    },
+    {
+      "epoch": 3.674019056193263,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.631269769543606e-05,
+      "loss": 0.0882,
+      "step": 12245
+    },
+    {
+      "epoch": 3.6743191537249604,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.629763518602199e-05,
+      "loss": 0.0785,
+      "step": 12246
+    },
+    {
+      "epoch": 3.6746192512566584,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.6282572676607925e-05,
+      "loss": 0.1014,
+      "step": 12247
+    },
+    {
+      "epoch": 3.6749193487883565,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6267510167193855e-05,
+      "loss": 0.1093,
+      "step": 12248
+    },
+    {
+      "epoch": 3.675219446320054,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.625244765777979e-05,
+      "loss": 0.0963,
+      "step": 12249
+    },
+    {
+      "epoch": 3.6755195438517516,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.623738514836572e-05,
+      "loss": 0.0902,
+      "step": 12250
+    },
+    {
+      "epoch": 3.6758196413834496,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6222322638951652e-05,
+      "loss": 0.1009,
+      "step": 12251
+    },
+    {
+      "epoch": 3.6761197389151477,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6207260129537583e-05,
+      "loss": 0.1063,
+      "step": 12252
+    },
+    {
+      "epoch": 3.6764198364468452,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6192197620123513e-05,
+      "loss": 0.1122,
+      "step": 12253
+    },
+    {
+      "epoch": 3.676719933978543,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6177135110709446e-05,
+      "loss": 0.0956,
+      "step": 12254
+    },
+    {
+      "epoch": 3.677020031510241,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6162072601295376e-05,
+      "loss": 0.096,
+      "step": 12255
+    },
+    {
+      "epoch": 3.677320129041939,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.614701009188131e-05,
+      "loss": 0.119,
+      "step": 12256
+    },
+    {
+      "epoch": 3.6776202265736364,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.613194758246724e-05,
+      "loss": 0.0819,
+      "step": 12257
+    },
+    {
+      "epoch": 3.677920324105334,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.611688507305317e-05,
+      "loss": 0.0749,
+      "step": 12258
+    },
+    {
+      "epoch": 3.678220421637032,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.61018225636391e-05,
+      "loss": 0.098,
+      "step": 12259
+    },
+    {
+      "epoch": 3.67852051916873,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6086760054225034e-05,
+      "loss": 0.0962,
+      "step": 12260
+    },
+    {
+      "epoch": 3.6788206167004276,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.6071697544810964e-05,
+      "loss": 0.1015,
+      "step": 12261
+    },
+    {
+      "epoch": 3.679120714232125,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6056635035396897e-05,
+      "loss": 0.1086,
+      "step": 12262
+    },
+    {
+      "epoch": 3.679420811763823,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6041572525982828e-05,
+      "loss": 0.0976,
+      "step": 12263
+    },
+    {
+      "epoch": 3.679720909295521,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.602651001656876e-05,
+      "loss": 0.0888,
+      "step": 12264
+    },
+    {
+      "epoch": 3.6800210068272188,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.601144750715469e-05,
+      "loss": 0.1056,
+      "step": 12265
+    },
+    {
+      "epoch": 3.680321104358917,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.5996384997740625e-05,
+      "loss": 0.1125,
+      "step": 12266
+    },
+    {
+      "epoch": 3.6806212018906144,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.5981322488326555e-05,
+      "loss": 0.1157,
+      "step": 12267
+    },
+    {
+      "epoch": 3.6809212994223124,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.596625997891249e-05,
+      "loss": 0.1125,
+      "step": 12268
+    },
+    {
+      "epoch": 3.68122139695401,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.595119746949842e-05,
+      "loss": 0.0928,
+      "step": 12269
+    },
+    {
+      "epoch": 3.681521494485708,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5936134960084352e-05,
+      "loss": 0.0881,
+      "step": 12270
+    },
+    {
+      "epoch": 3.6818215920174056,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5921072450670282e-05,
+      "loss": 0.0922,
+      "step": 12271
+    },
+    {
+      "epoch": 3.6821216895491036,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5906009941256216e-05,
+      "loss": 0.0991,
+      "step": 12272
+    },
+    {
+      "epoch": 3.682421787080801,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5890947431842146e-05,
+      "loss": 0.137,
+      "step": 12273
+    },
+    {
+      "epoch": 3.682721884612499,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.5875884922428076e-05,
+      "loss": 0.106,
+      "step": 12274
+    },
+    {
+      "epoch": 3.6830219821441967,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.586082241301401e-05,
+      "loss": 0.1064,
+      "step": 12275
+    },
+    {
+      "epoch": 3.6833220796758948,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.584575990359994e-05,
+      "loss": 0.0861,
+      "step": 12276
+    },
+    {
+      "epoch": 3.6836221772075923,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5830697394185873e-05,
+      "loss": 0.1043,
+      "step": 12277
+    },
+    {
+      "epoch": 3.6839222747392903,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5815634884771803e-05,
+      "loss": 0.0798,
+      "step": 12278
+    },
+    {
+      "epoch": 3.684222372270988,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5800572375357737e-05,
+      "loss": 0.0774,
+      "step": 12279
+    },
+    {
+      "epoch": 3.684522469802686,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5785509865943667e-05,
+      "loss": 0.1588,
+      "step": 12280
+    },
+    {
+      "epoch": 3.6848225673343835,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.57704473565296e-05,
+      "loss": 0.1089,
+      "step": 12281
+    },
+    {
+      "epoch": 3.6851226648660815,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.575538484711553e-05,
+      "loss": 0.0988,
+      "step": 12282
+    },
+    {
+      "epoch": 3.6854227623977796,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.5740322337701464e-05,
+      "loss": 0.126,
+      "step": 12283
+    },
+    {
+      "epoch": 3.685722859929477,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5725259828287394e-05,
+      "loss": 0.0887,
+      "step": 12284
+    },
+    {
+      "epoch": 3.6860229574611747,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5710197318873325e-05,
+      "loss": 0.0864,
+      "step": 12285
+    },
+    {
+      "epoch": 3.6863230549928727,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5695134809459255e-05,
+      "loss": 0.1013,
+      "step": 12286
+    },
+    {
+      "epoch": 3.6866231525245707,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.5680072300045188e-05,
+      "loss": 0.1253,
+      "step": 12287
+    },
+    {
+      "epoch": 3.6869232500562683,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.566500979063112e-05,
+      "loss": 0.1005,
+      "step": 12288
+    },
+    {
+      "epoch": 3.687223347587966,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5649947281217052e-05,
+      "loss": 0.1072,
+      "step": 12289
+    },
+    {
+      "epoch": 3.687523445119664,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.5634884771802982e-05,
+      "loss": 0.1073,
+      "step": 12290
+    },
+    {
+      "epoch": 3.687823542651362,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5619822262388916e-05,
+      "loss": 0.1,
+      "step": 12291
+    },
+    {
+      "epoch": 3.6881236401830595,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5604759752974846e-05,
+      "loss": 0.1016,
+      "step": 12292
+    },
+    {
+      "epoch": 3.688423737714757,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5589697243560776e-05,
+      "loss": 0.0951,
+      "step": 12293
+    },
+    {
+      "epoch": 3.688723835246455,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.557463473414671e-05,
+      "loss": 0.0856,
+      "step": 12294
+    },
+    {
+      "epoch": 3.689023932778153,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.555957222473264e-05,
+      "loss": 0.1091,
+      "step": 12295
+    },
+    {
+      "epoch": 3.6893240303098507,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5544509715318573e-05,
+      "loss": 0.1043,
+      "step": 12296
+    },
+    {
+      "epoch": 3.6896241278415483,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.5529447205904503e-05,
+      "loss": 0.1263,
+      "step": 12297
+    },
+    {
+      "epoch": 3.6899242253732463,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5514384696490437e-05,
+      "loss": 0.0913,
+      "step": 12298
+    },
+    {
+      "epoch": 3.6902243229049443,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5499322187076367e-05,
+      "loss": 0.1018,
+      "step": 12299
+    },
+    {
+      "epoch": 3.690524420436642,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.54842596776623e-05,
+      "loss": 0.0809,
+      "step": 12300
+    },
+    {
+      "epoch": 3.6908245179683394,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.546919716824823e-05,
+      "loss": 0.0789,
+      "step": 12301
+    },
+    {
+      "epoch": 3.6911246155000375,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5454134658834164e-05,
+      "loss": 0.0973,
+      "step": 12302
+    },
+    {
+      "epoch": 3.6914247130317355,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5439072149420094e-05,
+      "loss": 0.096,
+      "step": 12303
+    },
+    {
+      "epoch": 3.691724810563433,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5424009640006028e-05,
+      "loss": 0.0816,
+      "step": 12304
+    },
+    {
+      "epoch": 3.692024908095131,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5408947130591958e-05,
+      "loss": 0.0953,
+      "step": 12305
+    },
+    {
+      "epoch": 3.6923250056268286,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.539388462117789e-05,
+      "loss": 0.1018,
+      "step": 12306
+    },
+    {
+      "epoch": 3.6926251031585267,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.537882211176382e-05,
+      "loss": 0.0981,
+      "step": 12307
+    },
+    {
+      "epoch": 3.6929252006902242,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5363759602349752e-05,
+      "loss": 0.1191,
+      "step": 12308
+    },
+    {
+      "epoch": 3.6932252982219222,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5348697092935685e-05,
+      "loss": 0.0969,
+      "step": 12309
+    },
+    {
+      "epoch": 3.69352539575362,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5333634583521615e-05,
+      "loss": 0.0986,
+      "step": 12310
+    },
+    {
+      "epoch": 3.693825493285318,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.531857207410755e-05,
+      "loss": 0.0907,
+      "step": 12311
+    },
+    {
+      "epoch": 3.6941255908170154,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5303509564693476e-05,
+      "loss": 0.0938,
+      "step": 12312
+    },
+    {
+      "epoch": 3.6944256883487134,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.528844705527941e-05,
+      "loss": 0.1045,
+      "step": 12313
+    },
+    {
+      "epoch": 3.694725785880411,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.527338454586534e-05,
+      "loss": 0.1013,
+      "step": 12314
+    },
+    {
+      "epoch": 3.695025883412109,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5258322036451273e-05,
+      "loss": 0.0969,
+      "step": 12315
+    },
+    {
+      "epoch": 3.6953259809438066,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5243259527037205e-05,
+      "loss": 0.0894,
+      "step": 12316
+    },
+    {
+      "epoch": 3.6956260784755046,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5228197017623137e-05,
+      "loss": 0.0977,
+      "step": 12317
+    },
+    {
+      "epoch": 3.695926176007202,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5213134508209068e-05,
+      "loss": 0.1152,
+      "step": 12318
+    },
+    {
+      "epoch": 3.6962262735389,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5198071998794999e-05,
+      "loss": 0.0882,
+      "step": 12319
+    },
+    {
+      "epoch": 3.696526371070598,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.518300948938093e-05,
+      "loss": 0.1222,
+      "step": 12320
+    },
+    {
+      "epoch": 3.696826468602296,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5167946979966862e-05,
+      "loss": 0.0879,
+      "step": 12321
+    },
+    {
+      "epoch": 3.697126566133994,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.5152884470552794e-05,
+      "loss": 0.1004,
+      "step": 12322
+    },
+    {
+      "epoch": 3.6974266636656914,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5137821961138726e-05,
+      "loss": 0.1228,
+      "step": 12323
+    },
+    {
+      "epoch": 3.697726761197389,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.5122759451724658e-05,
+      "loss": 0.099,
+      "step": 12324
+    },
+    {
+      "epoch": 3.698026858729087,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.510769694231059e-05,
+      "loss": 0.1225,
+      "step": 12325
+    },
+    {
+      "epoch": 3.698326956260785,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5092634432896521e-05,
+      "loss": 0.1116,
+      "step": 12326
+    },
+    {
+      "epoch": 3.6986270537924826,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.5077571923482453e-05,
+      "loss": 0.1186,
+      "step": 12327
+    },
+    {
+      "epoch": 3.69892715132418,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5062509414068385e-05,
+      "loss": 0.104,
+      "step": 12328
+    },
+    {
+      "epoch": 3.699227248855878,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5047446904654317e-05,
+      "loss": 0.0956,
+      "step": 12329
+    },
+    {
+      "epoch": 3.699527346387576,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.5032384395240249e-05,
+      "loss": 0.122,
+      "step": 12330
+    },
+    {
+      "epoch": 3.6998274439192738,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.501732188582618e-05,
+      "loss": 0.0919,
+      "step": 12331
+    },
+    {
+      "epoch": 3.7001275414509713,
+      "grad_norm": 0.875,
+      "learning_rate": 1.5002259376412112e-05,
+      "loss": 0.0967,
+      "step": 12332
+    },
+    {
+      "epoch": 3.7004276389826694,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4987196866998044e-05,
+      "loss": 0.0916,
+      "step": 12333
+    },
+    {
+      "epoch": 3.7007277365143674,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.4972134357583974e-05,
+      "loss": 0.1248,
+      "step": 12334
+    },
+    {
+      "epoch": 3.701027834046065,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4957071848169906e-05,
+      "loss": 0.1249,
+      "step": 12335
+    },
+    {
+      "epoch": 3.7013279315777625,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4942009338755838e-05,
+      "loss": 0.1014,
+      "step": 12336
+    },
+    {
+      "epoch": 3.7016280291094605,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.492694682934177e-05,
+      "loss": 0.0915,
+      "step": 12337
+    },
+    {
+      "epoch": 3.7019281266411586,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4911884319927702e-05,
+      "loss": 0.1008,
+      "step": 12338
+    },
+    {
+      "epoch": 3.702228224172856,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.489682181051363e-05,
+      "loss": 0.075,
+      "step": 12339
+    },
+    {
+      "epoch": 3.7025283217045537,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4881759301099562e-05,
+      "loss": 0.0985,
+      "step": 12340
+    },
+    {
+      "epoch": 3.7028284192362517,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4866696791685494e-05,
+      "loss": 0.0982,
+      "step": 12341
+    },
+    {
+      "epoch": 3.7031285167679497,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4851634282271426e-05,
+      "loss": 0.0776,
+      "step": 12342
+    },
+    {
+      "epoch": 3.7034286142996473,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4836571772857357e-05,
+      "loss": 0.109,
+      "step": 12343
+    },
+    {
+      "epoch": 3.7037287118313453,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.482150926344329e-05,
+      "loss": 0.1291,
+      "step": 12344
+    },
+    {
+      "epoch": 3.704028809363043,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4806446754029221e-05,
+      "loss": 0.096,
+      "step": 12345
+    },
+    {
+      "epoch": 3.704328906894741,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4791384244615153e-05,
+      "loss": 0.0846,
+      "step": 12346
+    },
+    {
+      "epoch": 3.7046290044264385,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4776321735201085e-05,
+      "loss": 0.1084,
+      "step": 12347
+    },
+    {
+      "epoch": 3.7049291019581365,
+      "grad_norm": 1.125,
+      "learning_rate": 1.4761259225787017e-05,
+      "loss": 0.1486,
+      "step": 12348
+    },
+    {
+      "epoch": 3.705229199489834,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4746196716372948e-05,
+      "loss": 0.1016,
+      "step": 12349
+    },
+    {
+      "epoch": 3.705529297021532,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.473113420695888e-05,
+      "loss": 0.1026,
+      "step": 12350
+    },
+    {
+      "epoch": 3.7058293945532297,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4716071697544812e-05,
+      "loss": 0.0992,
+      "step": 12351
+    },
+    {
+      "epoch": 3.7061294920849277,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.4701009188130744e-05,
+      "loss": 0.1214,
+      "step": 12352
+    },
+    {
+      "epoch": 3.7064295896166253,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4685946678716674e-05,
+      "loss": 0.098,
+      "step": 12353
+    },
+    {
+      "epoch": 3.7067296871483233,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4670884169302606e-05,
+      "loss": 0.1041,
+      "step": 12354
+    },
+    {
+      "epoch": 3.707029784680021,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.4655821659888538e-05,
+      "loss": 0.109,
+      "step": 12355
+    },
+    {
+      "epoch": 3.707329882211719,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.464075915047447e-05,
+      "loss": 0.0854,
+      "step": 12356
+    },
+    {
+      "epoch": 3.7076299797434165,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4625696641060401e-05,
+      "loss": 0.1005,
+      "step": 12357
+    },
+    {
+      "epoch": 3.7079300772751145,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4610634131646333e-05,
+      "loss": 0.1013,
+      "step": 12358
+    },
+    {
+      "epoch": 3.708230174806812,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.4595571622232265e-05,
+      "loss": 0.0952,
+      "step": 12359
+    },
+    {
+      "epoch": 3.70853027233851,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4580509112818197e-05,
+      "loss": 0.0752,
+      "step": 12360
+    },
+    {
+      "epoch": 3.708830369870208,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4565446603404129e-05,
+      "loss": 0.1097,
+      "step": 12361
+    },
+    {
+      "epoch": 3.7091304674019057,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.455038409399006e-05,
+      "loss": 0.0948,
+      "step": 12362
+    },
+    {
+      "epoch": 3.7094305649336032,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.4535321584575992e-05,
+      "loss": 0.127,
+      "step": 12363
+    },
+    {
+      "epoch": 3.7097306624653013,
+      "grad_norm": 0.875,
+      "learning_rate": 1.4520259075161924e-05,
+      "loss": 0.1101,
+      "step": 12364
+    },
+    {
+      "epoch": 3.7100307599969993,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.4505196565747856e-05,
+      "loss": 0.098,
+      "step": 12365
+    },
+    {
+      "epoch": 3.710330857528697,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4490134056333788e-05,
+      "loss": 0.0833,
+      "step": 12366
+    },
+    {
+      "epoch": 3.7106309550603944,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4475071546919716e-05,
+      "loss": 0.0899,
+      "step": 12367
+    },
+    {
+      "epoch": 3.7109310525920924,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4460009037505648e-05,
+      "loss": 0.1109,
+      "step": 12368
+    },
+    {
+      "epoch": 3.7112311501237905,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.444494652809158e-05,
+      "loss": 0.1237,
+      "step": 12369
+    },
+    {
+      "epoch": 3.711531247655488,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4429884018677512e-05,
+      "loss": 0.1062,
+      "step": 12370
+    },
+    {
+      "epoch": 3.7118313451871856,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.4414821509263444e-05,
+      "loss": 0.1,
+      "step": 12371
+    },
+    {
+      "epoch": 3.7121314427188836,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4399758999849376e-05,
+      "loss": 0.1136,
+      "step": 12372
+    },
+    {
+      "epoch": 3.7124315402505816,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4384696490435306e-05,
+      "loss": 0.0984,
+      "step": 12373
+    },
+    {
+      "epoch": 3.712731637782279,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4369633981021238e-05,
+      "loss": 0.092,
+      "step": 12374
+    },
+    {
+      "epoch": 3.713031735313977,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.435457147160717e-05,
+      "loss": 0.1192,
+      "step": 12375
+    },
+    {
+      "epoch": 3.713331832845675,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4339508962193101e-05,
+      "loss": 0.1262,
+      "step": 12376
+    },
+    {
+      "epoch": 3.713631930377373,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4324446452779033e-05,
+      "loss": 0.104,
+      "step": 12377
+    },
+    {
+      "epoch": 3.7139320279090704,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4309383943364965e-05,
+      "loss": 0.12,
+      "step": 12378
+    },
+    {
+      "epoch": 3.714232125440768,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4294321433950897e-05,
+      "loss": 0.0923,
+      "step": 12379
+    },
+    {
+      "epoch": 3.714532222972466,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4279258924536829e-05,
+      "loss": 0.1194,
+      "step": 12380
+    },
+    {
+      "epoch": 3.714832320504164,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.426419641512276e-05,
+      "loss": 0.096,
+      "step": 12381
+    },
+    {
+      "epoch": 3.7151324180358616,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.4249133905708692e-05,
+      "loss": 0.1157,
+      "step": 12382
+    },
+    {
+      "epoch": 3.7154325155675596,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.4234071396294624e-05,
+      "loss": 0.1292,
+      "step": 12383
+    },
+    {
+      "epoch": 3.715732613099257,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4219008886880556e-05,
+      "loss": 0.0987,
+      "step": 12384
+    },
+    {
+      "epoch": 3.716032710630955,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4203946377466488e-05,
+      "loss": 0.1004,
+      "step": 12385
+    },
+    {
+      "epoch": 3.7163328081626528,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.418888386805242e-05,
+      "loss": 0.1086,
+      "step": 12386
+    },
+    {
+      "epoch": 3.716632905694351,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.417382135863835e-05,
+      "loss": 0.0972,
+      "step": 12387
+    },
+    {
+      "epoch": 3.7169330032260484,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4158758849224282e-05,
+      "loss": 0.1089,
+      "step": 12388
+    },
+    {
+      "epoch": 3.7172331007577464,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4143696339810213e-05,
+      "loss": 0.0992,
+      "step": 12389
+    },
+    {
+      "epoch": 3.717533198289444,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.4128633830396145e-05,
+      "loss": 0.1254,
+      "step": 12390
+    },
+    {
+      "epoch": 3.717833295821142,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.4113571320982077e-05,
+      "loss": 0.1167,
+      "step": 12391
+    },
+    {
+      "epoch": 3.7181333933528395,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4098508811568009e-05,
+      "loss": 0.078,
+      "step": 12392
+    },
+    {
+      "epoch": 3.7184334908845376,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.408344630215394e-05,
+      "loss": 0.0994,
+      "step": 12393
+    },
+    {
+      "epoch": 3.718733588416235,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.406838379273987e-05,
+      "loss": 0.0994,
+      "step": 12394
+    },
+    {
+      "epoch": 3.719033685947933,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4053321283325801e-05,
+      "loss": 0.1249,
+      "step": 12395
+    },
+    {
+      "epoch": 3.7193337834796307,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4038258773911733e-05,
+      "loss": 0.0937,
+      "step": 12396
+    },
+    {
+      "epoch": 3.7196338810113287,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4023196264497665e-05,
+      "loss": 0.0911,
+      "step": 12397
+    },
+    {
+      "epoch": 3.7199339785430263,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.4008133755083597e-05,
+      "loss": 0.181,
+      "step": 12398
+    },
+    {
+      "epoch": 3.7202340760747243,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3993071245669528e-05,
+      "loss": 0.1069,
+      "step": 12399
+    },
+    {
+      "epoch": 3.7205341736064224,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.397800873625546e-05,
+      "loss": 0.1064,
+      "step": 12400
+    },
+    {
+      "epoch": 3.72083427113812,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3962946226841392e-05,
+      "loss": 0.0769,
+      "step": 12401
+    },
+    {
+      "epoch": 3.7211343686698175,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.3947883717427324e-05,
+      "loss": 0.131,
+      "step": 12402
+    },
+    {
+      "epoch": 3.7214344662015155,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3932821208013256e-05,
+      "loss": 0.1175,
+      "step": 12403
+    },
+    {
+      "epoch": 3.7217345637332135,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3917758698599188e-05,
+      "loss": 0.1029,
+      "step": 12404
+    },
+    {
+      "epoch": 3.722034661264911,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.390269618918512e-05,
+      "loss": 0.1099,
+      "step": 12405
+    },
+    {
+      "epoch": 3.7223347587966087,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.3887633679771051e-05,
+      "loss": 0.1167,
+      "step": 12406
+    },
+    {
+      "epoch": 3.7226348563283067,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3872571170356981e-05,
+      "loss": 0.1038,
+      "step": 12407
+    },
+    {
+      "epoch": 3.7229349538600047,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.3857508660942913e-05,
+      "loss": 0.1482,
+      "step": 12408
+    },
+    {
+      "epoch": 3.7232350513917023,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.3842446151528845e-05,
+      "loss": 0.1057,
+      "step": 12409
+    },
+    {
+      "epoch": 3.7235351489234,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3827383642114777e-05,
+      "loss": 0.0873,
+      "step": 12410
+    },
+    {
+      "epoch": 3.723835246455098,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3812321132700709e-05,
+      "loss": 0.0907,
+      "step": 12411
+    },
+    {
+      "epoch": 3.724135343986796,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.379725862328664e-05,
+      "loss": 0.1391,
+      "step": 12412
+    },
+    {
+      "epoch": 3.7244354415184935,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3782196113872572e-05,
+      "loss": 0.0857,
+      "step": 12413
+    },
+    {
+      "epoch": 3.724735539050191,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3767133604458504e-05,
+      "loss": 0.0945,
+      "step": 12414
+    },
+    {
+      "epoch": 3.725035636581889,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3752071095044436e-05,
+      "loss": 0.1138,
+      "step": 12415
+    },
+    {
+      "epoch": 3.725335734113587,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3737008585630368e-05,
+      "loss": 0.0954,
+      "step": 12416
+    },
+    {
+      "epoch": 3.7256358316452847,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.37219460762163e-05,
+      "loss": 0.1005,
+      "step": 12417
+    },
+    {
+      "epoch": 3.7259359291769822,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3706883566802232e-05,
+      "loss": 0.098,
+      "step": 12418
+    },
+    {
+      "epoch": 3.7262360267086803,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3691821057388163e-05,
+      "loss": 0.1156,
+      "step": 12419
+    },
+    {
+      "epoch": 3.7265361242403783,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3676758547974095e-05,
+      "loss": 0.0971,
+      "step": 12420
+    },
+    {
+      "epoch": 3.726836221772076,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3661696038560024e-05,
+      "loss": 0.0903,
+      "step": 12421
+    },
+    {
+      "epoch": 3.727136319303774,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.3646633529145956e-05,
+      "loss": 0.1112,
+      "step": 12422
+    },
+    {
+      "epoch": 3.7274364168354714,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.3631571019731887e-05,
+      "loss": 0.1091,
+      "step": 12423
+    },
+    {
+      "epoch": 3.7277365143671695,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.361650851031782e-05,
+      "loss": 0.0836,
+      "step": 12424
+    },
+    {
+      "epoch": 3.728036611898867,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3601446000903751e-05,
+      "loss": 0.0921,
+      "step": 12425
+    },
+    {
+      "epoch": 3.728336709430565,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3586383491489683e-05,
+      "loss": 0.1023,
+      "step": 12426
+    },
+    {
+      "epoch": 3.7286368069622626,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.3571320982075613e-05,
+      "loss": 0.1955,
+      "step": 12427
+    },
+    {
+      "epoch": 3.7289369044939606,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3556258472661545e-05,
+      "loss": 0.1238,
+      "step": 12428
+    },
+    {
+      "epoch": 3.729237002025658,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3541195963247477e-05,
+      "loss": 0.0934,
+      "step": 12429
+    },
+    {
+      "epoch": 3.7295370995573562,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3526133453833409e-05,
+      "loss": 0.0883,
+      "step": 12430
+    },
+    {
+      "epoch": 3.729837197089054,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.351107094441934e-05,
+      "loss": 0.0848,
+      "step": 12431
+    },
+    {
+      "epoch": 3.730137294620752,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3496008435005272e-05,
+      "loss": 0.0883,
+      "step": 12432
+    },
+    {
+      "epoch": 3.7304373921524494,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.3480945925591204e-05,
+      "loss": 0.0926,
+      "step": 12433
+    },
+    {
+      "epoch": 3.7307374896841474,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.3465883416177136e-05,
+      "loss": 0.1203,
+      "step": 12434
+    },
+    {
+      "epoch": 3.731037587215845,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3450820906763068e-05,
+      "loss": 0.0836,
+      "step": 12435
+    },
+    {
+      "epoch": 3.731337684747543,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3435758397349e-05,
+      "loss": 0.0881,
+      "step": 12436
+    },
+    {
+      "epoch": 3.7316377822792406,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3420695887934931e-05,
+      "loss": 0.1073,
+      "step": 12437
+    },
+    {
+      "epoch": 3.7319378798109386,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3405633378520863e-05,
+      "loss": 0.1294,
+      "step": 12438
+    },
+    {
+      "epoch": 3.7322379773426366,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3390570869106795e-05,
+      "loss": 0.0861,
+      "step": 12439
+    },
+    {
+      "epoch": 3.732538074874334,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.3375508359692727e-05,
+      "loss": 0.1109,
+      "step": 12440
+    },
+    {
+      "epoch": 3.7328381724060318,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3360445850278657e-05,
+      "loss": 0.0888,
+      "step": 12441
+    },
+    {
+      "epoch": 3.73313826993773,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3345383340864589e-05,
+      "loss": 0.1135,
+      "step": 12442
+    },
+    {
+      "epoch": 3.733438367469428,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.333032083145052e-05,
+      "loss": 0.1025,
+      "step": 12443
+    },
+    {
+      "epoch": 3.7337384650011254,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.3315258322036453e-05,
+      "loss": 0.0948,
+      "step": 12444
+    },
+    {
+      "epoch": 3.734038562532823,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3300195812622384e-05,
+      "loss": 0.097,
+      "step": 12445
+    },
+    {
+      "epoch": 3.734338660064521,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.3285133303208316e-05,
+      "loss": 0.1089,
+      "step": 12446
+    },
+    {
+      "epoch": 3.734638757596219,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.3270070793794248e-05,
+      "loss": 0.1167,
+      "step": 12447
+    },
+    {
+      "epoch": 3.7349388551279166,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3255008284380176e-05,
+      "loss": 0.0948,
+      "step": 12448
+    },
+    {
+      "epoch": 3.735238952659614,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3239945774966108e-05,
+      "loss": 0.0952,
+      "step": 12449
+    },
+    {
+      "epoch": 3.735539050191312,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.322488326555204e-05,
+      "loss": 0.0915,
+      "step": 12450
+    },
+    {
+      "epoch": 3.73583914772301,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3209820756137972e-05,
+      "loss": 0.0697,
+      "step": 12451
+    },
+    {
+      "epoch": 3.7361392452547078,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3194758246723904e-05,
+      "loss": 0.0989,
+      "step": 12452
+    },
+    {
+      "epoch": 3.7364393427864053,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3179695737309836e-05,
+      "loss": 0.1088,
+      "step": 12453
+    },
+    {
+      "epoch": 3.7367394403181033,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.3164633227895767e-05,
+      "loss": 0.1212,
+      "step": 12454
+    },
+    {
+      "epoch": 3.7370395378498014,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.31495707184817e-05,
+      "loss": 0.0838,
+      "step": 12455
+    },
+    {
+      "epoch": 3.737339635381499,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3134508209067631e-05,
+      "loss": 0.1034,
+      "step": 12456
+    },
+    {
+      "epoch": 3.737639732913197,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.3119445699653563e-05,
+      "loss": 0.1095,
+      "step": 12457
+    },
+    {
+      "epoch": 3.7379398304448945,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3104383190239495e-05,
+      "loss": 0.0873,
+      "step": 12458
+    },
+    {
+      "epoch": 3.7382399279765925,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.3089320680825427e-05,
+      "loss": 0.0889,
+      "step": 12459
+    },
+    {
+      "epoch": 3.73854002550829,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.3074258171411358e-05,
+      "loss": 0.1149,
+      "step": 12460
+    },
+    {
+      "epoch": 3.738840123039988,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3059195661997289e-05,
+      "loss": 0.1265,
+      "step": 12461
+    },
+    {
+      "epoch": 3.7391402205716857,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.304413315258322e-05,
+      "loss": 0.1195,
+      "step": 12462
+    },
+    {
+      "epoch": 3.7394403181033837,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.3029070643169152e-05,
+      "loss": 0.1087,
+      "step": 12463
+    },
+    {
+      "epoch": 3.7397404156350813,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.3014008133755084e-05,
+      "loss": 0.1149,
+      "step": 12464
+    },
+    {
+      "epoch": 3.7400405131667793,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2998945624341016e-05,
+      "loss": 0.0974,
+      "step": 12465
+    },
+    {
+      "epoch": 3.740340610698477,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2983883114926948e-05,
+      "loss": 0.0792,
+      "step": 12466
+    },
+    {
+      "epoch": 3.740640708230175,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.296882060551288e-05,
+      "loss": 0.085,
+      "step": 12467
+    },
+    {
+      "epoch": 3.7409408057618725,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2953758096098811e-05,
+      "loss": 0.0961,
+      "step": 12468
+    },
+    {
+      "epoch": 3.7412409032935705,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2938695586684743e-05,
+      "loss": 0.0875,
+      "step": 12469
+    },
+    {
+      "epoch": 3.741541000825268,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.2923633077270675e-05,
+      "loss": 0.1,
+      "step": 12470
+    },
+    {
+      "epoch": 3.741841098356966,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2908570567856607e-05,
+      "loss": 0.0868,
+      "step": 12471
+    },
+    {
+      "epoch": 3.7421411958886637,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.2893508058442539e-05,
+      "loss": 0.1065,
+      "step": 12472
+    },
+    {
+      "epoch": 3.7424412934203617,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.287844554902847e-05,
+      "loss": 0.0871,
+      "step": 12473
+    },
+    {
+      "epoch": 3.7427413909520593,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2863383039614402e-05,
+      "loss": 0.1037,
+      "step": 12474
+    },
+    {
+      "epoch": 3.7430414884837573,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.2848320530200331e-05,
+      "loss": 0.1344,
+      "step": 12475
+    },
+    {
+      "epoch": 3.743341586015455,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.2833258020786263e-05,
+      "loss": 0.1343,
+      "step": 12476
+    },
+    {
+      "epoch": 3.743641683547153,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.2818195511372195e-05,
+      "loss": 0.0937,
+      "step": 12477
+    },
+    {
+      "epoch": 3.743941781078851,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.2803133001958126e-05,
+      "loss": 0.0993,
+      "step": 12478
+    },
+    {
+      "epoch": 3.7442418786105485,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2788070492544058e-05,
+      "loss": 0.1098,
+      "step": 12479
+    },
+    {
+      "epoch": 3.744541976142246,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2773007983129988e-05,
+      "loss": 0.1185,
+      "step": 12480
+    },
+    {
+      "epoch": 3.744842073673944,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.275794547371592e-05,
+      "loss": 0.1089,
+      "step": 12481
+    },
+    {
+      "epoch": 3.745142171205642,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2742882964301852e-05,
+      "loss": 0.088,
+      "step": 12482
+    },
+    {
+      "epoch": 3.7454422687373397,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2727820454887784e-05,
+      "loss": 0.0977,
+      "step": 12483
+    },
+    {
+      "epoch": 3.7457423662690372,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2712757945473716e-05,
+      "loss": 0.1115,
+      "step": 12484
+    },
+    {
+      "epoch": 3.7460424638007352,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2697695436059648e-05,
+      "loss": 0.0989,
+      "step": 12485
+    },
+    {
+      "epoch": 3.7463425613324333,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.268263292664558e-05,
+      "loss": 0.1001,
+      "step": 12486
+    },
+    {
+      "epoch": 3.746642658864131,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.2667570417231511e-05,
+      "loss": 0.1176,
+      "step": 12487
+    },
+    {
+      "epoch": 3.7469427563958284,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2652507907817443e-05,
+      "loss": 0.0924,
+      "step": 12488
+    },
+    {
+      "epoch": 3.7472428539275264,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.2637445398403375e-05,
+      "loss": 0.1106,
+      "step": 12489
+    },
+    {
+      "epoch": 3.7475429514592244,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.2622382888989307e-05,
+      "loss": 0.1183,
+      "step": 12490
+    },
+    {
+      "epoch": 3.747843048990922,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.2607320379575239e-05,
+      "loss": 0.099,
+      "step": 12491
+    },
+    {
+      "epoch": 3.7481431465226196,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.259225787016117e-05,
+      "loss": 0.0828,
+      "step": 12492
+    },
+    {
+      "epoch": 3.7484432440543176,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2577195360747102e-05,
+      "loss": 0.0907,
+      "step": 12493
+    },
+    {
+      "epoch": 3.7487433415860156,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2562132851333034e-05,
+      "loss": 0.1027,
+      "step": 12494
+    },
+    {
+      "epoch": 3.749043439117713,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.2547070341918964e-05,
+      "loss": 0.1077,
+      "step": 12495
+    },
+    {
+      "epoch": 3.7493435366494112,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.2532007832504896e-05,
+      "loss": 0.124,
+      "step": 12496
+    },
+    {
+      "epoch": 3.749643634181109,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2516945323090828e-05,
+      "loss": 0.122,
+      "step": 12497
+    },
+    {
+      "epoch": 3.749943731712807,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.250188281367676e-05,
+      "loss": 0.0982,
+      "step": 12498
+    },
+    {
+      "epoch": 3.7502438292445044,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.248682030426269e-05,
+      "loss": 0.0934,
+      "step": 12499
+    },
+    {
+      "epoch": 3.7505439267762024,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2471757794848622e-05,
+      "loss": 0.0762,
+      "step": 12500
+    },
+    {
+      "epoch": 3.7508440243079,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2456695285434554e-05,
+      "loss": 0.0956,
+      "step": 12501
+    },
+    {
+      "epoch": 3.751144121839598,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2441632776020485e-05,
+      "loss": 0.0965,
+      "step": 12502
+    },
+    {
+      "epoch": 3.7514442193712956,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2426570266606417e-05,
+      "loss": 0.0926,
+      "step": 12503
+    },
+    {
+      "epoch": 3.7517443169029936,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.2411507757192349e-05,
+      "loss": 0.0974,
+      "step": 12504
+    },
+    {
+      "epoch": 3.752044414434691,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.2396445247778281e-05,
+      "loss": 0.0979,
+      "step": 12505
+    },
+    {
+      "epoch": 3.752344511966389,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2381382738364213e-05,
+      "loss": 0.1013,
+      "step": 12506
+    },
+    {
+      "epoch": 3.7526446094980868,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.2366320228950145e-05,
+      "loss": 0.124,
+      "step": 12507
+    },
+    {
+      "epoch": 3.7529447070297848,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2351257719536076e-05,
+      "loss": 0.089,
+      "step": 12508
+    },
+    {
+      "epoch": 3.7532448045614824,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2336195210122007e-05,
+      "loss": 0.0599,
+      "step": 12509
+    },
+    {
+      "epoch": 3.7535449020931804,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.2321132700707938e-05,
+      "loss": 0.075,
+      "step": 12510
+    },
+    {
+      "epoch": 3.753844999624878,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.230607019129387e-05,
+      "loss": 0.148,
+      "step": 12511
+    },
+    {
+      "epoch": 3.754145097156576,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2291007681879802e-05,
+      "loss": 0.1111,
+      "step": 12512
+    },
+    {
+      "epoch": 3.7544451946882735,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.2275945172465734e-05,
+      "loss": 0.1163,
+      "step": 12513
+    },
+    {
+      "epoch": 3.7547452922199716,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.2260882663051666e-05,
+      "loss": 0.1173,
+      "step": 12514
+    },
+    {
+      "epoch": 3.755045389751669,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.2245820153637596e-05,
+      "loss": 0.1057,
+      "step": 12515
+    },
+    {
+      "epoch": 3.755345487283367,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.2230757644223528e-05,
+      "loss": 0.0973,
+      "step": 12516
+    },
+    {
+      "epoch": 3.755645584815065,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.221569513480946e-05,
+      "loss": 0.0921,
+      "step": 12517
+    },
+    {
+      "epoch": 3.7559456823467627,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.2200632625395391e-05,
+      "loss": 0.0956,
+      "step": 12518
+    },
+    {
+      "epoch": 3.7562457798784603,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2185570115981323e-05,
+      "loss": 0.0916,
+      "step": 12519
+    },
+    {
+      "epoch": 3.7565458774101583,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.2170507606567255e-05,
+      "loss": 0.1252,
+      "step": 12520
+    },
+    {
+      "epoch": 3.7568459749418563,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.2155445097153187e-05,
+      "loss": 0.1125,
+      "step": 12521
+    },
+    {
+      "epoch": 3.757146072473554,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.2140382587739117e-05,
+      "loss": 0.0972,
+      "step": 12522
+    },
+    {
+      "epoch": 3.7574461700052515,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.2125320078325049e-05,
+      "loss": 0.117,
+      "step": 12523
+    },
+    {
+      "epoch": 3.7577462675369495,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.211025756891098e-05,
+      "loss": 0.1011,
+      "step": 12524
+    },
+    {
+      "epoch": 3.7580463650686475,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2095195059496913e-05,
+      "loss": 0.1095,
+      "step": 12525
+    },
+    {
+      "epoch": 3.758346462600345,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2080132550082844e-05,
+      "loss": 0.1114,
+      "step": 12526
+    },
+    {
+      "epoch": 3.7586465601320427,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2065070040668776e-05,
+      "loss": 0.0966,
+      "step": 12527
+    },
+    {
+      "epoch": 3.7589466576637407,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.2050007531254708e-05,
+      "loss": 0.1154,
+      "step": 12528
+    },
+    {
+      "epoch": 3.7592467551954387,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.203494502184064e-05,
+      "loss": 0.1131,
+      "step": 12529
+    },
+    {
+      "epoch": 3.7595468527271363,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.2019882512426572e-05,
+      "loss": 0.109,
+      "step": 12530
+    },
+    {
+      "epoch": 3.759846950258834,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2004820003012504e-05,
+      "loss": 0.1065,
+      "step": 12531
+    },
+    {
+      "epoch": 3.760147047790532,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1989757493598434e-05,
+      "loss": 0.1194,
+      "step": 12532
+    },
+    {
+      "epoch": 3.76044714532223,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1974694984184366e-05,
+      "loss": 0.1279,
+      "step": 12533
+    },
+    {
+      "epoch": 3.7607472428539275,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.1959632474770297e-05,
+      "loss": 0.0878,
+      "step": 12534
+    },
+    {
+      "epoch": 3.7610473403856255,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.194456996535623e-05,
+      "loss": 0.1266,
+      "step": 12535
+    },
+    {
+      "epoch": 3.761347437917323,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.192950745594216e-05,
+      "loss": 0.1107,
+      "step": 12536
+    },
+    {
+      "epoch": 3.761647535449021,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1914444946528091e-05,
+      "loss": 0.1154,
+      "step": 12537
+    },
+    {
+      "epoch": 3.7619476329807187,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1899382437114023e-05,
+      "loss": 0.1173,
+      "step": 12538
+    },
+    {
+      "epoch": 3.7622477305124167,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1884319927699955e-05,
+      "loss": 0.0802,
+      "step": 12539
+    },
+    {
+      "epoch": 3.7625478280441143,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.1869257418285887e-05,
+      "loss": 0.1055,
+      "step": 12540
+    },
+    {
+      "epoch": 3.7628479255758123,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.1854194908871819e-05,
+      "loss": 0.0868,
+      "step": 12541
+    },
+    {
+      "epoch": 3.76314802310751,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.183913239945775e-05,
+      "loss": 0.0975,
+      "step": 12542
+    },
+    {
+      "epoch": 3.763448120639208,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.1824069890043682e-05,
+      "loss": 0.1059,
+      "step": 12543
+    },
+    {
+      "epoch": 3.7637482181709054,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1809007380629614e-05,
+      "loss": 0.0858,
+      "step": 12544
+    },
+    {
+      "epoch": 3.7640483157026035,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.1793944871215546e-05,
+      "loss": 0.131,
+      "step": 12545
+    },
+    {
+      "epoch": 3.764348413234301,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1778882361801478e-05,
+      "loss": 0.0939,
+      "step": 12546
+    },
+    {
+      "epoch": 3.764648510765999,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.176381985238741e-05,
+      "loss": 0.1132,
+      "step": 12547
+    },
+    {
+      "epoch": 3.7649486082976966,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1748757342973341e-05,
+      "loss": 0.0839,
+      "step": 12548
+    },
+    {
+      "epoch": 3.7652487058293946,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.1733694833559271e-05,
+      "loss": 0.1197,
+      "step": 12549
+    },
+    {
+      "epoch": 3.765548803361092,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1718632324145203e-05,
+      "loss": 0.1104,
+      "step": 12550
+    },
+    {
+      "epoch": 3.7658489008927902,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.1703569814731133e-05,
+      "loss": 0.0942,
+      "step": 12551
+    },
+    {
+      "epoch": 3.766148998424488,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1688507305317065e-05,
+      "loss": 0.0956,
+      "step": 12552
+    },
+    {
+      "epoch": 3.766449095956186,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1673444795902997e-05,
+      "loss": 0.0907,
+      "step": 12553
+    },
+    {
+      "epoch": 3.7667491934878834,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1658382286488929e-05,
+      "loss": 0.0924,
+      "step": 12554
+    },
+    {
+      "epoch": 3.7670492910195814,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.164331977707486e-05,
+      "loss": 0.109,
+      "step": 12555
+    },
+    {
+      "epoch": 3.7673493885512794,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.1628257267660793e-05,
+      "loss": 0.1127,
+      "step": 12556
+    },
+    {
+      "epoch": 3.767649486082977,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.1613194758246724e-05,
+      "loss": 0.1112,
+      "step": 12557
+    },
+    {
+      "epoch": 3.7679495836146746,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1598132248832656e-05,
+      "loss": 0.0806,
+      "step": 12558
+    },
+    {
+      "epoch": 3.7682496811463726,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.1583069739418588e-05,
+      "loss": 0.0752,
+      "step": 12559
+    },
+    {
+      "epoch": 3.7685497786780706,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.156800723000452e-05,
+      "loss": 0.1045,
+      "step": 12560
+    },
+    {
+      "epoch": 3.768849876209768,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1552944720590452e-05,
+      "loss": 0.109,
+      "step": 12561
+    },
+    {
+      "epoch": 3.7691499737414658,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1537882211176384e-05,
+      "loss": 0.0926,
+      "step": 12562
+    },
+    {
+      "epoch": 3.769450071273164,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1522819701762314e-05,
+      "loss": 0.1191,
+      "step": 12563
+    },
+    {
+      "epoch": 3.769750168804862,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1507757192348246e-05,
+      "loss": 0.0912,
+      "step": 12564
+    },
+    {
+      "epoch": 3.7700502663365594,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1492694682934177e-05,
+      "loss": 0.0998,
+      "step": 12565
+    },
+    {
+      "epoch": 3.770350363868257,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.147763217352011e-05,
+      "loss": 0.1152,
+      "step": 12566
+    },
+    {
+      "epoch": 3.770650461399955,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.1462569664106041e-05,
+      "loss": 0.1016,
+      "step": 12567
+    },
+    {
+      "epoch": 3.770950558931653,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1447507154691971e-05,
+      "loss": 0.0755,
+      "step": 12568
+    },
+    {
+      "epoch": 3.7712506564633506,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1432444645277903e-05,
+      "loss": 0.0841,
+      "step": 12569
+    },
+    {
+      "epoch": 3.771550753995048,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1417382135863835e-05,
+      "loss": 0.1194,
+      "step": 12570
+    },
+    {
+      "epoch": 3.771850851526746,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1402319626449767e-05,
+      "loss": 0.1049,
+      "step": 12571
+    },
+    {
+      "epoch": 3.772150949058444,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.1387257117035699e-05,
+      "loss": 0.0984,
+      "step": 12572
+    },
+    {
+      "epoch": 3.7724510465901417,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.137219460762163e-05,
+      "loss": 0.1846,
+      "step": 12573
+    },
+    {
+      "epoch": 3.7727511441218398,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.1357132098207562e-05,
+      "loss": 0.1284,
+      "step": 12574
+    },
+    {
+      "epoch": 3.7730512416535373,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1342069588793494e-05,
+      "loss": 0.0888,
+      "step": 12575
+    },
+    {
+      "epoch": 3.7733513391852354,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.1327007079379426e-05,
+      "loss": 0.1156,
+      "step": 12576
+    },
+    {
+      "epoch": 3.773651436716933,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.1311944569965356e-05,
+      "loss": 0.1318,
+      "step": 12577
+    },
+    {
+      "epoch": 3.773951534248631,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1296882060551288e-05,
+      "loss": 0.0927,
+      "step": 12578
+    },
+    {
+      "epoch": 3.7742516317803285,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.128181955113722e-05,
+      "loss": 0.0997,
+      "step": 12579
+    },
+    {
+      "epoch": 3.7745517293120265,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.1266757041723152e-05,
+      "loss": 0.086,
+      "step": 12580
+    },
+    {
+      "epoch": 3.774851826843724,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.1251694532309083e-05,
+      "loss": 0.1027,
+      "step": 12581
+    },
+    {
+      "epoch": 3.775151924375422,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1236632022895015e-05,
+      "loss": 0.0862,
+      "step": 12582
+    },
+    {
+      "epoch": 3.7754520219071197,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.1221569513480947e-05,
+      "loss": 0.1107,
+      "step": 12583
+    },
+    {
+      "epoch": 3.7757521194388177,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1206507004066879e-05,
+      "loss": 0.111,
+      "step": 12584
+    },
+    {
+      "epoch": 3.7760522169705153,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1191444494652809e-05,
+      "loss": 0.0852,
+      "step": 12585
+    },
+    {
+      "epoch": 3.7763523145022133,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1176381985238741e-05,
+      "loss": 0.1032,
+      "step": 12586
+    },
+    {
+      "epoch": 3.776652412033911,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.1161319475824673e-05,
+      "loss": 0.1038,
+      "step": 12587
+    },
+    {
+      "epoch": 3.776952509565609,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.1146256966410605e-05,
+      "loss": 0.1227,
+      "step": 12588
+    },
+    {
+      "epoch": 3.7772526070973065,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1131194456996536e-05,
+      "loss": 0.0944,
+      "step": 12589
+    },
+    {
+      "epoch": 3.7775527046290045,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.1116131947582467e-05,
+      "loss": 0.1129,
+      "step": 12590
+    },
+    {
+      "epoch": 3.777852802160702,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.1101069438168398e-05,
+      "loss": 0.1181,
+      "step": 12591
+    },
+    {
+      "epoch": 3.7781528996924,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.108600692875433e-05,
+      "loss": 0.0836,
+      "step": 12592
+    },
+    {
+      "epoch": 3.7784529972240977,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.1070944419340262e-05,
+      "loss": 0.1056,
+      "step": 12593
+    },
+    {
+      "epoch": 3.7787530947557957,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1055881909926194e-05,
+      "loss": 0.1216,
+      "step": 12594
+    },
+    {
+      "epoch": 3.7790531922874937,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1040819400512126e-05,
+      "loss": 0.0974,
+      "step": 12595
+    },
+    {
+      "epoch": 3.7793532898191913,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.1025756891098058e-05,
+      "loss": 0.1206,
+      "step": 12596
+    },
+    {
+      "epoch": 3.779653387350889,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.101069438168399e-05,
+      "loss": 0.0776,
+      "step": 12597
+    },
+    {
+      "epoch": 3.779953484882587,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.0995631872269921e-05,
+      "loss": 0.1085,
+      "step": 12598
+    },
+    {
+      "epoch": 3.780253582414285,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0980569362855853e-05,
+      "loss": 0.1018,
+      "step": 12599
+    },
+    {
+      "epoch": 3.7805536799459825,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0965506853441785e-05,
+      "loss": 0.1285,
+      "step": 12600
+    },
+    {
+      "epoch": 3.78085377747768,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0950444344027717e-05,
+      "loss": 0.0952,
+      "step": 12601
+    },
+    {
+      "epoch": 3.781153875009378,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0935381834613647e-05,
+      "loss": 0.0991,
+      "step": 12602
+    },
+    {
+      "epoch": 3.781453972541076,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0920319325199579e-05,
+      "loss": 0.1085,
+      "step": 12603
+    },
+    {
+      "epoch": 3.7817540700727736,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.090525681578551e-05,
+      "loss": 0.1058,
+      "step": 12604
+    },
+    {
+      "epoch": 3.782054167604471,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.089019430637144e-05,
+      "loss": 0.1241,
+      "step": 12605
+    },
+    {
+      "epoch": 3.7823542651361692,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0875131796957373e-05,
+      "loss": 0.1238,
+      "step": 12606
+    },
+    {
+      "epoch": 3.7826543626678673,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.0860069287543304e-05,
+      "loss": 0.1163,
+      "step": 12607
+    },
+    {
+      "epoch": 3.782954460199565,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.0845006778129236e-05,
+      "loss": 0.1204,
+      "step": 12608
+    },
+    {
+      "epoch": 3.7832545577312624,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.0829944268715168e-05,
+      "loss": 0.0887,
+      "step": 12609
+    },
+    {
+      "epoch": 3.7835546552629604,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.08148817593011e-05,
+      "loss": 0.0939,
+      "step": 12610
+    },
+    {
+      "epoch": 3.7838547527946584,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.0799819249887032e-05,
+      "loss": 0.1151,
+      "step": 12611
+    },
+    {
+      "epoch": 3.784154850326356,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0784756740472964e-05,
+      "loss": 0.0958,
+      "step": 12612
+    },
+    {
+      "epoch": 3.784454947858054,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0769694231058895e-05,
+      "loss": 0.0896,
+      "step": 12613
+    },
+    {
+      "epoch": 3.7847550453897516,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0754631721644827e-05,
+      "loss": 0.1004,
+      "step": 12614
+    },
+    {
+      "epoch": 3.7850551429214496,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0739569212230759e-05,
+      "loss": 0.0966,
+      "step": 12615
+    },
+    {
+      "epoch": 3.785355240453147,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0724506702816691e-05,
+      "loss": 0.0913,
+      "step": 12616
+    },
+    {
+      "epoch": 3.785655337984845,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0709444193402621e-05,
+      "loss": 0.0918,
+      "step": 12617
+    },
+    {
+      "epoch": 3.785955435516543,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0694381683988553e-05,
+      "loss": 0.0986,
+      "step": 12618
+    },
+    {
+      "epoch": 3.786255533048241,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0679319174574485e-05,
+      "loss": 0.0893,
+      "step": 12619
+    },
+    {
+      "epoch": 3.7865556305799384,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0664256665160417e-05,
+      "loss": 0.089,
+      "step": 12620
+    },
+    {
+      "epoch": 3.7868557281116364,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0649194155746348e-05,
+      "loss": 0.097,
+      "step": 12621
+    },
+    {
+      "epoch": 3.787155825643334,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0634131646332279e-05,
+      "loss": 0.0959,
+      "step": 12622
+    },
+    {
+      "epoch": 3.787455923175032,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.061906913691821e-05,
+      "loss": 0.1005,
+      "step": 12623
+    },
+    {
+      "epoch": 3.7877560207067296,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0604006627504142e-05,
+      "loss": 0.1237,
+      "step": 12624
+    },
+    {
+      "epoch": 3.7880561182384276,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.0588944118090074e-05,
+      "loss": 0.0916,
+      "step": 12625
+    },
+    {
+      "epoch": 3.788356215770125,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0573881608676006e-05,
+      "loss": 0.1056,
+      "step": 12626
+    },
+    {
+      "epoch": 3.788656313301823,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0558819099261938e-05,
+      "loss": 0.0903,
+      "step": 12627
+    },
+    {
+      "epoch": 3.7889564108335207,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.054375658984787e-05,
+      "loss": 0.0875,
+      "step": 12628
+    },
+    {
+      "epoch": 3.7892565083652188,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0528694080433801e-05,
+      "loss": 0.1043,
+      "step": 12629
+    },
+    {
+      "epoch": 3.7895566058969163,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.0513631571019733e-05,
+      "loss": 0.0815,
+      "step": 12630
+    },
+    {
+      "epoch": 3.7898567034286144,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.0498569061605663e-05,
+      "loss": 0.1196,
+      "step": 12631
+    },
+    {
+      "epoch": 3.790156800960312,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.0483506552191595e-05,
+      "loss": 0.0885,
+      "step": 12632
+    },
+    {
+      "epoch": 3.79045689849201,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.0468444042777527e-05,
+      "loss": 0.0888,
+      "step": 12633
+    },
+    {
+      "epoch": 3.790756996023708,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.0453381533363459e-05,
+      "loss": 0.1108,
+      "step": 12634
+    },
+    {
+      "epoch": 3.7910570935554055,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.043831902394939e-05,
+      "loss": 0.0765,
+      "step": 12635
+    },
+    {
+      "epoch": 3.791357191087103,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0423256514535323e-05,
+      "loss": 0.0797,
+      "step": 12636
+    },
+    {
+      "epoch": 3.791657288618801,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.0408194005121254e-05,
+      "loss": 0.143,
+      "step": 12637
+    },
+    {
+      "epoch": 3.791957386150499,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0393131495707186e-05,
+      "loss": 0.1056,
+      "step": 12638
+    },
+    {
+      "epoch": 3.7922574836821967,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0378068986293116e-05,
+      "loss": 0.115,
+      "step": 12639
+    },
+    {
+      "epoch": 3.7925575812138943,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0363006476879048e-05,
+      "loss": 0.1,
+      "step": 12640
+    },
+    {
+      "epoch": 3.7928576787455923,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.034794396746498e-05,
+      "loss": 0.1298,
+      "step": 12641
+    },
+    {
+      "epoch": 3.7931577762772903,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.0332881458050912e-05,
+      "loss": 0.1118,
+      "step": 12642
+    },
+    {
+      "epoch": 3.793457873808988,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0317818948636844e-05,
+      "loss": 0.0795,
+      "step": 12643
+    },
+    {
+      "epoch": 3.7937579713406855,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0302756439222774e-05,
+      "loss": 0.0957,
+      "step": 12644
+    },
+    {
+      "epoch": 3.7940580688723835,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0287693929808706e-05,
+      "loss": 0.2043,
+      "step": 12645
+    },
+    {
+      "epoch": 3.7943581664040815,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.0272631420394637e-05,
+      "loss": 0.1009,
+      "step": 12646
+    },
+    {
+      "epoch": 3.794658263935779,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.025756891098057e-05,
+      "loss": 0.1228,
+      "step": 12647
+    },
+    {
+      "epoch": 3.7949583614674767,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.0242506401566501e-05,
+      "loss": 0.0787,
+      "step": 12648
+    },
+    {
+      "epoch": 3.7952584589991747,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0227443892152433e-05,
+      "loss": 0.098,
+      "step": 12649
+    },
+    {
+      "epoch": 3.7955585565308727,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0212381382738365e-05,
+      "loss": 0.1001,
+      "step": 12650
+    },
+    {
+      "epoch": 3.7958586540625703,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0197318873324297e-05,
+      "loss": 0.1141,
+      "step": 12651
+    },
+    {
+      "epoch": 3.7961587515942683,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0182256363910228e-05,
+      "loss": 0.0898,
+      "step": 12652
+    },
+    {
+      "epoch": 3.796458849125966,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.016719385449616e-05,
+      "loss": 0.103,
+      "step": 12653
+    },
+    {
+      "epoch": 3.796758946657664,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.0152131345082092e-05,
+      "loss": 0.0922,
+      "step": 12654
+    },
+    {
+      "epoch": 3.7970590441893615,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.0137068835668024e-05,
+      "loss": 0.1347,
+      "step": 12655
+    },
+    {
+      "epoch": 3.7973591417210595,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0122006326253954e-05,
+      "loss": 0.1014,
+      "step": 12656
+    },
+    {
+      "epoch": 3.797659239252757,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0106943816839886e-05,
+      "loss": 0.1035,
+      "step": 12657
+    },
+    {
+      "epoch": 3.797959336784455,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0091881307425818e-05,
+      "loss": 0.0817,
+      "step": 12658
+    },
+    {
+      "epoch": 3.7982594343161526,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.0076818798011748e-05,
+      "loss": 0.1176,
+      "step": 12659
+    },
+    {
+      "epoch": 3.7985595318478507,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.006175628859768e-05,
+      "loss": 0.1047,
+      "step": 12660
+    },
+    {
+      "epoch": 3.7988596293795482,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0046693779183612e-05,
+      "loss": 0.1028,
+      "step": 12661
+    },
+    {
+      "epoch": 3.7991597269112463,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0031631269769543e-05,
+      "loss": 0.1324,
+      "step": 12662
+    },
+    {
+      "epoch": 3.799459824442944,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.0016568760355475e-05,
+      "loss": 0.1145,
+      "step": 12663
+    },
+    {
+      "epoch": 3.799759921974642,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0001506250941407e-05,
+      "loss": 0.1166,
+      "step": 12664
+    },
+    {
+      "epoch": 3.8000600195063394,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.986443741527339e-06,
+      "loss": 0.0771,
+      "step": 12665
+    },
+    {
+      "epoch": 3.8003601170380374,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.97138123211327e-06,
+      "loss": 0.0902,
+      "step": 12666
+    },
+    {
+      "epoch": 3.800660214569735,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.956318722699203e-06,
+      "loss": 0.0947,
+      "step": 12667
+    },
+    {
+      "epoch": 3.800960312101433,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.941256213285134e-06,
+      "loss": 0.098,
+      "step": 12668
+    },
+    {
+      "epoch": 3.8012604096331306,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.926193703871066e-06,
+      "loss": 0.1303,
+      "step": 12669
+    },
+    {
+      "epoch": 3.8015605071648286,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.911131194456998e-06,
+      "loss": 0.1124,
+      "step": 12670
+    },
+    {
+      "epoch": 3.801860604696526,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.896068685042928e-06,
+      "loss": 0.0745,
+      "step": 12671
+    },
+    {
+      "epoch": 3.802160702228224,
+      "grad_norm": 0.98046875,
+      "learning_rate": 9.88100617562886e-06,
+      "loss": 0.1171,
+      "step": 12672
+    },
+    {
+      "epoch": 3.8024607997599222,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.865943666214792e-06,
+      "loss": 0.1028,
+      "step": 12673
+    },
+    {
+      "epoch": 3.80276089729162,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.850881156800724e-06,
+      "loss": 0.1002,
+      "step": 12674
+    },
+    {
+      "epoch": 3.8030609948233174,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.835818647386656e-06,
+      "loss": 0.095,
+      "step": 12675
+    },
+    {
+      "epoch": 3.8033610923550154,
+      "grad_norm": 0.80078125,
+      "learning_rate": 9.820756137972586e-06,
+      "loss": 0.0836,
+      "step": 12676
+    },
+    {
+      "epoch": 3.8036611898867134,
+      "grad_norm": 0.875,
+      "learning_rate": 9.805693628558518e-06,
+      "loss": 0.1006,
+      "step": 12677
+    },
+    {
+      "epoch": 3.803961287418411,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.79063111914445e-06,
+      "loss": 0.0923,
+      "step": 12678
+    },
+    {
+      "epoch": 3.8042613849501086,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.775568609730381e-06,
+      "loss": 0.1135,
+      "step": 12679
+    },
+    {
+      "epoch": 3.8045614824818066,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.760506100316313e-06,
+      "loss": 0.1016,
+      "step": 12680
+    },
+    {
+      "epoch": 3.8048615800135046,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.745443590902245e-06,
+      "loss": 0.0866,
+      "step": 12681
+    },
+    {
+      "epoch": 3.805161677545202,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.730381081488177e-06,
+      "loss": 0.093,
+      "step": 12682
+    },
+    {
+      "epoch": 3.8054617750768998,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.715318572074109e-06,
+      "loss": 0.1047,
+      "step": 12683
+    },
+    {
+      "epoch": 3.8057618726085978,
+      "grad_norm": 1.03125,
+      "learning_rate": 9.70025606266004e-06,
+      "loss": 0.1278,
+      "step": 12684
+    },
+    {
+      "epoch": 3.806061970140296,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.68519355324597e-06,
+      "loss": 0.0738,
+      "step": 12685
+    },
+    {
+      "epoch": 3.8063620676719934,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.670131043831902e-06,
+      "loss": 0.0807,
+      "step": 12686
+    },
+    {
+      "epoch": 3.806662165203691,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.655068534417834e-06,
+      "loss": 0.1204,
+      "step": 12687
+    },
+    {
+      "epoch": 3.806962262735389,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.640006025003766e-06,
+      "loss": 0.1077,
+      "step": 12688
+    },
+    {
+      "epoch": 3.807262360267087,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.624943515589698e-06,
+      "loss": 0.1287,
+      "step": 12689
+    },
+    {
+      "epoch": 3.8075624577987845,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.60988100617563e-06,
+      "loss": 0.0864,
+      "step": 12690
+    },
+    {
+      "epoch": 3.8078625553304826,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.594818496761562e-06,
+      "loss": 0.0833,
+      "step": 12691
+    },
+    {
+      "epoch": 3.80816265286218,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.579755987347493e-06,
+      "loss": 0.0689,
+      "step": 12692
+    },
+    {
+      "epoch": 3.808462750393878,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.564693477933424e-06,
+      "loss": 0.121,
+      "step": 12693
+    },
+    {
+      "epoch": 3.8087628479255757,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.549630968519355e-06,
+      "loss": 0.1044,
+      "step": 12694
+    },
+    {
+      "epoch": 3.8090629454572738,
+      "grad_norm": 0.8359375,
+      "learning_rate": 9.534568459105287e-06,
+      "loss": 0.0982,
+      "step": 12695
+    },
+    {
+      "epoch": 3.8093630429889713,
+      "grad_norm": 0.8515625,
+      "learning_rate": 9.519505949691219e-06,
+      "loss": 0.0919,
+      "step": 12696
+    },
+    {
+      "epoch": 3.8096631405206693,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.504443440277151e-06,
+      "loss": 0.0968,
+      "step": 12697
+    },
+    {
+      "epoch": 3.809963238052367,
+      "grad_norm": 0.90234375,
+      "learning_rate": 9.489380930863081e-06,
+      "loss": 0.124,
+      "step": 12698
+    },
+    {
+      "epoch": 3.810263335584065,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.474318421449013e-06,
+      "loss": 0.0939,
+      "step": 12699
+    },
+    {
+      "epoch": 3.8105634331157625,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.459255912034945e-06,
+      "loss": 0.1005,
+      "step": 12700
+    },
+    {
+      "epoch": 3.8108635306474605,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.444193402620877e-06,
+      "loss": 0.0839,
+      "step": 12701
+    },
+    {
+      "epoch": 3.811163628179158,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.429130893206808e-06,
+      "loss": 0.0958,
+      "step": 12702
+    },
+    {
+      "epoch": 3.811463725710856,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.41406838379274e-06,
+      "loss": 0.1145,
+      "step": 12703
+    },
+    {
+      "epoch": 3.8117638232425537,
+      "grad_norm": 0.94921875,
+      "learning_rate": 9.399005874378672e-06,
+      "loss": 0.1121,
+      "step": 12704
+    },
+    {
+      "epoch": 3.8120639207742517,
+      "grad_norm": 0.7734375,
+      "learning_rate": 9.383943364964604e-06,
+      "loss": 0.105,
+      "step": 12705
+    },
+    {
+      "epoch": 3.8123640183059493,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.368880855550536e-06,
+      "loss": 0.1227,
+      "step": 12706
+    },
+    {
+      "epoch": 3.8126641158376473,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.353818346136468e-06,
+      "loss": 0.0907,
+      "step": 12707
+    },
+    {
+      "epoch": 3.812964213369345,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.3387558367224e-06,
+      "loss": 0.0872,
+      "step": 12708
+    },
+    {
+      "epoch": 3.813264310901043,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.323693327308331e-06,
+      "loss": 0.1235,
+      "step": 12709
+    },
+    {
+      "epoch": 3.8135644084327405,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.308630817894261e-06,
+      "loss": 0.1113,
+      "step": 12710
+    },
+    {
+      "epoch": 3.8138645059644385,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.293568308480193e-06,
+      "loss": 0.122,
+      "step": 12711
+    },
+    {
+      "epoch": 3.8141646034961365,
+      "grad_norm": 0.875,
+      "learning_rate": 9.278505799066125e-06,
+      "loss": 0.1112,
+      "step": 12712
+    },
+    {
+      "epoch": 3.814464701027834,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.263443289652055e-06,
+      "loss": 0.0899,
+      "step": 12713
+    },
+    {
+      "epoch": 3.8147647985595317,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.248380780237987e-06,
+      "loss": 0.0846,
+      "step": 12714
+    },
+    {
+      "epoch": 3.8150648960912297,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.233318270823919e-06,
+      "loss": 0.1063,
+      "step": 12715
+    },
+    {
+      "epoch": 3.8153649936229277,
+      "grad_norm": 0.84765625,
+      "learning_rate": 9.21825576140985e-06,
+      "loss": 0.0956,
+      "step": 12716
+    },
+    {
+      "epoch": 3.8156650911546253,
+      "grad_norm": 0.83984375,
+      "learning_rate": 9.203193251995783e-06,
+      "loss": 0.0929,
+      "step": 12717
+    },
+    {
+      "epoch": 3.815965188686323,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.188130742581714e-06,
+      "loss": 0.0842,
+      "step": 12718
+    },
+    {
+      "epoch": 3.816265286218021,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.173068233167646e-06,
+      "loss": 0.0846,
+      "step": 12719
+    },
+    {
+      "epoch": 3.816565383749719,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.158005723753578e-06,
+      "loss": 0.0903,
+      "step": 12720
+    },
+    {
+      "epoch": 3.8168654812814165,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.14294321433951e-06,
+      "loss": 0.073,
+      "step": 12721
+    },
+    {
+      "epoch": 3.817165578813114,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.127880704925442e-06,
+      "loss": 0.0986,
+      "step": 12722
+    },
+    {
+      "epoch": 3.817465676344812,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.112818195511374e-06,
+      "loss": 0.089,
+      "step": 12723
+    },
+    {
+      "epoch": 3.81776577387651,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.097755686097305e-06,
+      "loss": 0.1215,
+      "step": 12724
+    },
+    {
+      "epoch": 3.8180658714082076,
+      "grad_norm": 0.83203125,
+      "learning_rate": 9.082693176683237e-06,
+      "loss": 0.1184,
+      "step": 12725
+    },
+    {
+      "epoch": 3.818365968939905,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.067630667269167e-06,
+      "loss": 0.0927,
+      "step": 12726
+    },
+    {
+      "epoch": 3.8186660664716032,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.0525681578551e-06,
+      "loss": 0.1454,
+      "step": 12727
+    },
+    {
+      "epoch": 3.8189661640033012,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.037505648441031e-06,
+      "loss": 0.1213,
+      "step": 12728
+    },
+    {
+      "epoch": 3.819266261534999,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.022443139026963e-06,
+      "loss": 0.0953,
+      "step": 12729
+    },
+    {
+      "epoch": 3.819566359066697,
+      "grad_norm": 0.79296875,
+      "learning_rate": 9.007380629612893e-06,
+      "loss": 0.0971,
+      "step": 12730
+    },
+    {
+      "epoch": 3.8198664565983944,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.992318120198825e-06,
+      "loss": 0.0907,
+      "step": 12731
+    },
+    {
+      "epoch": 3.8201665541300924,
+      "grad_norm": 0.8671875,
+      "learning_rate": 8.977255610784757e-06,
+      "loss": 0.0851,
+      "step": 12732
+    },
+    {
+      "epoch": 3.82046665166179,
+      "grad_norm": 0.9375,
+      "learning_rate": 8.962193101370689e-06,
+      "loss": 0.1185,
+      "step": 12733
+    },
+    {
+      "epoch": 3.820766749193488,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.94713059195662e-06,
+      "loss": 0.1027,
+      "step": 12734
+    },
+    {
+      "epoch": 3.8210668467251856,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.932068082542552e-06,
+      "loss": 0.1032,
+      "step": 12735
+    },
+    {
+      "epoch": 3.8213669442568836,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.917005573128484e-06,
+      "loss": 0.1171,
+      "step": 12736
+    },
+    {
+      "epoch": 3.821667041788581,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.901943063714416e-06,
+      "loss": 0.093,
+      "step": 12737
+    },
+    {
+      "epoch": 3.821967139320279,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.886880554300348e-06,
+      "loss": 0.0772,
+      "step": 12738
+    },
+    {
+      "epoch": 3.822267236851977,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.871818044886278e-06,
+      "loss": 0.0973,
+      "step": 12739
+    },
+    {
+      "epoch": 3.822567334383675,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.85675553547221e-06,
+      "loss": 0.1019,
+      "step": 12740
+    },
+    {
+      "epoch": 3.8228674319153724,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.841693026058141e-06,
+      "loss": 0.0958,
+      "step": 12741
+    },
+    {
+      "epoch": 3.8231675294470704,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.826630516644073e-06,
+      "loss": 0.0951,
+      "step": 12742
+    },
+    {
+      "epoch": 3.823467626978768,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.811568007230005e-06,
+      "loss": 0.1022,
+      "step": 12743
+    },
+    {
+      "epoch": 3.823767724510466,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.796505497815937e-06,
+      "loss": 0.1082,
+      "step": 12744
+    },
+    {
+      "epoch": 3.8240678220421636,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.781442988401869e-06,
+      "loss": 0.1083,
+      "step": 12745
+    },
+    {
+      "epoch": 3.8243679195738616,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.7663804789878e-06,
+      "loss": 0.0813,
+      "step": 12746
+    },
+    {
+      "epoch": 3.824668017105559,
+      "grad_norm": 0.8359375,
+      "learning_rate": 8.75131796957373e-06,
+      "loss": 0.1001,
+      "step": 12747
+    },
+    {
+      "epoch": 3.824968114637257,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.736255460159663e-06,
+      "loss": 0.0922,
+      "step": 12748
+    },
+    {
+      "epoch": 3.8252682121689547,
+      "grad_norm": 0.765625,
+      "learning_rate": 8.721192950745594e-06,
+      "loss": 0.1007,
+      "step": 12749
+    },
+    {
+      "epoch": 3.8255683097006528,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.706130441331526e-06,
+      "loss": 0.0862,
+      "step": 12750
+    },
+    {
+      "epoch": 3.8258684072323508,
+      "grad_norm": 0.78125,
+      "learning_rate": 8.691067931917458e-06,
+      "loss": 0.0838,
+      "step": 12751
+    },
+    {
+      "epoch": 3.8261685047640484,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.67600542250339e-06,
+      "loss": 0.104,
+      "step": 12752
+    },
+    {
+      "epoch": 3.826468602295746,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.66094291308932e-06,
+      "loss": 0.0962,
+      "step": 12753
+    },
+    {
+      "epoch": 3.826768699827444,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.645880403675252e-06,
+      "loss": 0.1145,
+      "step": 12754
+    },
+    {
+      "epoch": 3.827068797359142,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.630817894261184e-06,
+      "loss": 0.1046,
+      "step": 12755
+    },
+    {
+      "epoch": 3.8273688948908395,
+      "grad_norm": 0.81640625,
+      "learning_rate": 8.615755384847116e-06,
+      "loss": 0.1058,
+      "step": 12756
+    },
+    {
+      "epoch": 3.827668992422537,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.600692875433047e-06,
+      "loss": 0.1018,
+      "step": 12757
+    },
+    {
+      "epoch": 3.827969089954235,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.58563036601898e-06,
+      "loss": 0.1269,
+      "step": 12758
+    },
+    {
+      "epoch": 3.828269187485933,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.570567856604911e-06,
+      "loss": 0.1048,
+      "step": 12759
+    },
+    {
+      "epoch": 3.8285692850176307,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.555505347190843e-06,
+      "loss": 0.1014,
+      "step": 12760
+    },
+    {
+      "epoch": 3.8288693825493283,
+      "grad_norm": 0.75,
+      "learning_rate": 8.540442837776775e-06,
+      "loss": 0.0911,
+      "step": 12761
+    },
+    {
+      "epoch": 3.8291694800810263,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.525380328362707e-06,
+      "loss": 0.0924,
+      "step": 12762
+    },
+    {
+      "epoch": 3.8294695776127243,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.510317818948638e-06,
+      "loss": 0.0945,
+      "step": 12763
+    },
+    {
+      "epoch": 3.829769675144422,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.495255309534569e-06,
+      "loss": 0.082,
+      "step": 12764
+    },
+    {
+      "epoch": 3.8300697726761195,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.4801928001205e-06,
+      "loss": 0.1036,
+      "step": 12765
+    },
+    {
+      "epoch": 3.8303698702078175,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.46513029070643e-06,
+      "loss": 0.1034,
+      "step": 12766
+    },
+    {
+      "epoch": 3.8306699677395155,
+      "grad_norm": 0.98046875,
+      "learning_rate": 8.450067781292362e-06,
+      "loss": 0.1162,
+      "step": 12767
+    },
+    {
+      "epoch": 3.830970065271213,
+      "grad_norm": 0.7734375,
+      "learning_rate": 8.435005271878294e-06,
+      "loss": 0.1007,
+      "step": 12768
+    },
+    {
+      "epoch": 3.831270162802911,
+      "grad_norm": 0.890625,
+      "learning_rate": 8.419942762464226e-06,
+      "loss": 0.1042,
+      "step": 12769
+    },
+    {
+      "epoch": 3.8315702603346087,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.404880253050158e-06,
+      "loss": 0.1084,
+      "step": 12770
+    },
+    {
+      "epoch": 3.8318703578663067,
+      "grad_norm": 0.75,
+      "learning_rate": 8.38981774363609e-06,
+      "loss": 0.0939,
+      "step": 12771
+    },
+    {
+      "epoch": 3.8321704553980043,
+      "grad_norm": 0.875,
+      "learning_rate": 8.374755234222022e-06,
+      "loss": 0.0976,
+      "step": 12772
+    },
+    {
+      "epoch": 3.8324705529297023,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.359692724807953e-06,
+      "loss": 0.0955,
+      "step": 12773
+    },
+    {
+      "epoch": 3.8327706504614,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.344630215393885e-06,
+      "loss": 0.1193,
+      "step": 12774
+    },
+    {
+      "epoch": 3.833070747993098,
+      "grad_norm": 0.84765625,
+      "learning_rate": 8.329567705979817e-06,
+      "loss": 0.1062,
+      "step": 12775
+    },
+    {
+      "epoch": 3.8333708455247955,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.314505196565749e-06,
+      "loss": 0.1223,
+      "step": 12776
+    },
+    {
+      "epoch": 3.8336709430564935,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.29944268715168e-06,
+      "loss": 0.0919,
+      "step": 12777
+    },
+    {
+      "epoch": 3.833971040588191,
+      "grad_norm": 0.859375,
+      "learning_rate": 8.284380177737613e-06,
+      "loss": 0.1031,
+      "step": 12778
+    },
+    {
+      "epoch": 3.834271138119889,
+      "grad_norm": 1.0,
+      "learning_rate": 8.269317668323544e-06,
+      "loss": 0.129,
+      "step": 12779
+    },
+    {
+      "epoch": 3.8345712356515866,
+      "grad_norm": 0.9296875,
+      "learning_rate": 8.254255158909475e-06,
+      "loss": 0.1328,
+      "step": 12780
+    },
+    {
+      "epoch": 3.8348713331832847,
+      "grad_norm": 0.8515625,
+      "learning_rate": 8.239192649495406e-06,
+      "loss": 0.0989,
+      "step": 12781
+    },
+    {
+      "epoch": 3.8351714307149822,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.224130140081338e-06,
+      "loss": 0.099,
+      "step": 12782
+    },
+    {
+      "epoch": 3.8354715282466803,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.209067630667268e-06,
+      "loss": 0.0861,
+      "step": 12783
+    },
+    {
+      "epoch": 3.835771625778378,
+      "grad_norm": 0.9140625,
+      "learning_rate": 8.1940051212532e-06,
+      "loss": 0.1115,
+      "step": 12784
+    },
+    {
+      "epoch": 3.836071723310076,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.178942611839132e-06,
+      "loss": 0.0972,
+      "step": 12785
+    },
+    {
+      "epoch": 3.8363718208417734,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.163880102425064e-06,
+      "loss": 0.1049,
+      "step": 12786
+    },
+    {
+      "epoch": 3.8366719183734714,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.148817593010996e-06,
+      "loss": 0.1051,
+      "step": 12787
+    },
+    {
+      "epoch": 3.836972015905169,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.133755083596928e-06,
+      "loss": 0.0975,
+      "step": 12788
+    },
+    {
+      "epoch": 3.837272113436867,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.11869257418286e-06,
+      "loss": 0.0864,
+      "step": 12789
+    },
+    {
+      "epoch": 3.837572210968565,
+      "grad_norm": 0.82421875,
+      "learning_rate": 8.103630064768791e-06,
+      "loss": 0.1112,
+      "step": 12790
+    },
+    {
+      "epoch": 3.8378723085002626,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.088567555354723e-06,
+      "loss": 0.1006,
+      "step": 12791
+    },
+    {
+      "epoch": 3.83817240603196,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.073505045940655e-06,
+      "loss": 0.1161,
+      "step": 12792
+    },
+    {
+      "epoch": 3.838472503563658,
+      "grad_norm": 0.7265625,
+      "learning_rate": 8.058442536526585e-06,
+      "loss": 0.0903,
+      "step": 12793
+    },
+    {
+      "epoch": 3.8387726010953562,
+      "grad_norm": 0.84375,
+      "learning_rate": 8.043380027112517e-06,
+      "loss": 0.1118,
+      "step": 12794
+    },
+    {
+      "epoch": 3.839072698627054,
+      "grad_norm": 0.75,
+      "learning_rate": 8.028317517698449e-06,
+      "loss": 0.0976,
+      "step": 12795
+    },
+    {
+      "epoch": 3.8393727961587514,
+      "grad_norm": 0.76953125,
+      "learning_rate": 8.01325500828438e-06,
+      "loss": 0.1009,
+      "step": 12796
+    },
+    {
+      "epoch": 3.8396728936904494,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.998192498870312e-06,
+      "loss": 0.096,
+      "step": 12797
+    },
+    {
+      "epoch": 3.8399729912221474,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.983129989456244e-06,
+      "loss": 0.0877,
+      "step": 12798
+    },
+    {
+      "epoch": 3.840273088753845,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.968067480042176e-06,
+      "loss": 0.0828,
+      "step": 12799
+    },
+    {
+      "epoch": 3.8405731862855426,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.953004970628108e-06,
+      "loss": 0.0879,
+      "step": 12800
+    },
+    {
+      "epoch": 3.8408732838172406,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.937942461214038e-06,
+      "loss": 0.1041,
+      "step": 12801
+    },
+    {
+      "epoch": 3.8411733813489386,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.92287995179997e-06,
+      "loss": 0.0914,
+      "step": 12802
+    },
+    {
+      "epoch": 3.841473478880636,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.907817442385902e-06,
+      "loss": 0.1171,
+      "step": 12803
+    },
+    {
+      "epoch": 3.8417735764123337,
+      "grad_norm": 0.87890625,
+      "learning_rate": 7.892754932971834e-06,
+      "loss": 0.1026,
+      "step": 12804
+    },
+    {
+      "epoch": 3.8420736739440318,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.877692423557765e-06,
+      "loss": 0.0989,
+      "step": 12805
+    },
+    {
+      "epoch": 3.84237377147573,
+      "grad_norm": 0.765625,
+      "learning_rate": 7.862629914143697e-06,
+      "loss": 0.1083,
+      "step": 12806
+    },
+    {
+      "epoch": 3.8426738690074274,
+      "grad_norm": 0.94921875,
+      "learning_rate": 7.847567404729627e-06,
+      "loss": 0.1118,
+      "step": 12807
+    },
+    {
+      "epoch": 3.8429739665391254,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.83250489531556e-06,
+      "loss": 0.1012,
+      "step": 12808
+    },
+    {
+      "epoch": 3.843274064070823,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.817442385901491e-06,
+      "loss": 0.0746,
+      "step": 12809
+    },
+    {
+      "epoch": 3.843574161602521,
+      "grad_norm": 0.859375,
+      "learning_rate": 7.802379876487423e-06,
+      "loss": 0.1148,
+      "step": 12810
+    },
+    {
+      "epoch": 3.8438742591342185,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.787317367073355e-06,
+      "loss": 0.0969,
+      "step": 12811
+    },
+    {
+      "epoch": 3.8441743566659166,
+      "grad_norm": 0.85546875,
+      "learning_rate": 7.772254857659287e-06,
+      "loss": 0.1169,
+      "step": 12812
+    },
+    {
+      "epoch": 3.844474454197614,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.757192348245218e-06,
+      "loss": 0.1508,
+      "step": 12813
+    },
+    {
+      "epoch": 3.844774551729312,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.74212983883115e-06,
+      "loss": 0.1116,
+      "step": 12814
+    },
+    {
+      "epoch": 3.8450746492610097,
+      "grad_norm": 0.984375,
+      "learning_rate": 7.727067329417082e-06,
+      "loss": 0.1289,
+      "step": 12815
+    },
+    {
+      "epoch": 3.8453747467927077,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.712004820003014e-06,
+      "loss": 0.0966,
+      "step": 12816
+    },
+    {
+      "epoch": 3.8456748443244053,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.696942310588946e-06,
+      "loss": 0.1105,
+      "step": 12817
+    },
+    {
+      "epoch": 3.8459749418561033,
+      "grad_norm": 0.79296875,
+      "learning_rate": 7.681879801174876e-06,
+      "loss": 0.0902,
+      "step": 12818
+    },
+    {
+      "epoch": 3.846275039387801,
+      "grad_norm": 0.9375,
+      "learning_rate": 7.666817291760808e-06,
+      "loss": 0.1304,
+      "step": 12819
+    },
+    {
+      "epoch": 3.846575136919499,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.651754782346738e-06,
+      "loss": 0.1142,
+      "step": 12820
+    },
+    {
+      "epoch": 3.8468752344511965,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.63669227293267e-06,
+      "loss": 0.1177,
+      "step": 12821
+    },
+    {
+      "epoch": 3.8471753319828945,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.621629763518602e-06,
+      "loss": 0.099,
+      "step": 12822
+    },
+    {
+      "epoch": 3.847475429514592,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.606567254104534e-06,
+      "loss": 0.0948,
+      "step": 12823
+    },
+    {
+      "epoch": 3.84777552704629,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.591504744690465e-06,
+      "loss": 0.0929,
+      "step": 12824
+    },
+    {
+      "epoch": 3.8480756245779877,
+      "grad_norm": 0.921875,
+      "learning_rate": 7.576442235276397e-06,
+      "loss": 0.1338,
+      "step": 12825
+    },
+    {
+      "epoch": 3.8483757221096857,
+      "grad_norm": 0.83203125,
+      "learning_rate": 7.561379725862329e-06,
+      "loss": 0.0932,
+      "step": 12826
+    },
+    {
+      "epoch": 3.8486758196413833,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.546317216448261e-06,
+      "loss": 0.1041,
+      "step": 12827
+    },
+    {
+      "epoch": 3.8489759171730813,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.5312547070341925e-06,
+      "loss": 0.0848,
+      "step": 12828
+    },
+    {
+      "epoch": 3.8492760147047793,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.516192197620124e-06,
+      "loss": 0.0874,
+      "step": 12829
+    },
+    {
+      "epoch": 3.849576112236477,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.501129688206056e-06,
+      "loss": 0.1116,
+      "step": 12830
+    },
+    {
+      "epoch": 3.8498762097681745,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.486067178791987e-06,
+      "loss": 0.0992,
+      "step": 12831
+    },
+    {
+      "epoch": 3.8501763072998725,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.471004669377919e-06,
+      "loss": 0.1009,
+      "step": 12832
+    },
+    {
+      "epoch": 3.8504764048315705,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.455942159963851e-06,
+      "loss": 0.0787,
+      "step": 12833
+    },
+    {
+      "epoch": 3.850776502363268,
+      "grad_norm": 0.7734375,
+      "learning_rate": 7.440879650549781e-06,
+      "loss": 0.1013,
+      "step": 12834
+    },
+    {
+      "epoch": 3.8510765998949656,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.425817141135713e-06,
+      "loss": 0.0885,
+      "step": 12835
+    },
+    {
+      "epoch": 3.8513766974266637,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.410754631721645e-06,
+      "loss": 0.1011,
+      "step": 12836
+    },
+    {
+      "epoch": 3.8516767949583617,
+      "grad_norm": 0.875,
+      "learning_rate": 7.3956921223075765e-06,
+      "loss": 0.0986,
+      "step": 12837
+    },
+    {
+      "epoch": 3.8519768924900593,
+      "grad_norm": 0.73046875,
+      "learning_rate": 7.380629612893508e-06,
+      "loss": 0.0906,
+      "step": 12838
+    },
+    {
+      "epoch": 3.852276990021757,
+      "grad_norm": 0.94140625,
+      "learning_rate": 7.36556710347944e-06,
+      "loss": 0.1152,
+      "step": 12839
+    },
+    {
+      "epoch": 3.852577087553455,
+      "grad_norm": 0.80078125,
+      "learning_rate": 7.350504594065372e-06,
+      "loss": 0.1051,
+      "step": 12840
+    },
+    {
+      "epoch": 3.852877185085153,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.335442084651303e-06,
+      "loss": 0.0955,
+      "step": 12841
+    },
+    {
+      "epoch": 3.8531772826168504,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.320379575237235e-06,
+      "loss": 0.1168,
+      "step": 12842
+    },
+    {
+      "epoch": 3.853477380148548,
+      "grad_norm": 0.9140625,
+      "learning_rate": 7.305317065823167e-06,
+      "loss": 0.1161,
+      "step": 12843
+    },
+    {
+      "epoch": 3.853777477680246,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.2902545564090985e-06,
+      "loss": 0.0931,
+      "step": 12844
+    },
+    {
+      "epoch": 3.854077575211944,
+      "grad_norm": 0.875,
+      "learning_rate": 7.27519204699503e-06,
+      "loss": 0.127,
+      "step": 12845
+    },
+    {
+      "epoch": 3.8543776727436416,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.260129537580962e-06,
+      "loss": 0.086,
+      "step": 12846
+    },
+    {
+      "epoch": 3.8546777702753396,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.245067028166894e-06,
+      "loss": 0.108,
+      "step": 12847
+    },
+    {
+      "epoch": 3.854977867807037,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.230004518752824e-06,
+      "loss": 0.0937,
+      "step": 12848
+    },
+    {
+      "epoch": 3.8552779653387352,
+      "grad_norm": 0.91015625,
+      "learning_rate": 7.214942009338756e-06,
+      "loss": 0.1065,
+      "step": 12849
+    },
+    {
+      "epoch": 3.855578062870433,
+      "grad_norm": 1.34375,
+      "learning_rate": 7.199879499924688e-06,
+      "loss": 0.1303,
+      "step": 12850
+    },
+    {
+      "epoch": 3.855878160402131,
+      "grad_norm": 0.75,
+      "learning_rate": 7.184816990510619e-06,
+      "loss": 0.1012,
+      "step": 12851
+    },
+    {
+      "epoch": 3.8561782579338284,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.169754481096551e-06,
+      "loss": 0.0724,
+      "step": 12852
+    },
+    {
+      "epoch": 3.8564783554655264,
+      "grad_norm": 0.75,
+      "learning_rate": 7.1546919716824825e-06,
+      "loss": 0.0982,
+      "step": 12853
+    },
+    {
+      "epoch": 3.856778452997224,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.139629462268414e-06,
+      "loss": 0.0906,
+      "step": 12854
+    },
+    {
+      "epoch": 3.857078550528922,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.124566952854346e-06,
+      "loss": 0.1089,
+      "step": 12855
+    },
+    {
+      "epoch": 3.8573786480606196,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.109504443440278e-06,
+      "loss": 0.0804,
+      "step": 12856
+    },
+    {
+      "epoch": 3.8576787455923176,
+      "grad_norm": 0.8671875,
+      "learning_rate": 7.09444193402621e-06,
+      "loss": 0.113,
+      "step": 12857
+    },
+    {
+      "epoch": 3.857978843124015,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.079379424612141e-06,
+      "loss": 0.1075,
+      "step": 12858
+    },
+    {
+      "epoch": 3.858278940655713,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.064316915198073e-06,
+      "loss": 0.0949,
+      "step": 12859
+    },
+    {
+      "epoch": 3.8585790381874108,
+      "grad_norm": 0.92578125,
+      "learning_rate": 7.0492544057840045e-06,
+      "loss": 0.137,
+      "step": 12860
+    },
+    {
+      "epoch": 3.858879135719109,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.034191896369935e-06,
+      "loss": 0.0951,
+      "step": 12861
+    },
+    {
+      "epoch": 3.8591792332508064,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.0191293869558664e-06,
+      "loss": 0.096,
+      "step": 12862
+    },
+    {
+      "epoch": 3.8594793307825044,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.004066877541798e-06,
+      "loss": 0.1093,
+      "step": 12863
+    },
+    {
+      "epoch": 3.859779428314202,
+      "grad_norm": 0.828125,
+      "learning_rate": 6.98900436812773e-06,
+      "loss": 0.0981,
+      "step": 12864
+    },
+    {
+      "epoch": 3.8600795258459,
+      "grad_norm": 0.875,
+      "learning_rate": 6.973941858713662e-06,
+      "loss": 0.1009,
+      "step": 12865
+    },
+    {
+      "epoch": 3.8603796233775975,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.958879349299594e-06,
+      "loss": 0.0972,
+      "step": 12866
+    },
+    {
+      "epoch": 3.8606797209092956,
+      "grad_norm": 0.75,
+      "learning_rate": 6.943816839885526e-06,
+      "loss": 0.1054,
+      "step": 12867
+    },
+    {
+      "epoch": 3.8609798184409936,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.928754330471457e-06,
+      "loss": 0.0903,
+      "step": 12868
+    },
+    {
+      "epoch": 3.861279915972691,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.9136918210573884e-06,
+      "loss": 0.1236,
+      "step": 12869
+    },
+    {
+      "epoch": 3.8615800135043887,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.89862931164332e-06,
+      "loss": 0.0803,
+      "step": 12870
+    },
+    {
+      "epoch": 3.8618801110360867,
+      "grad_norm": 0.875,
+      "learning_rate": 6.883566802229252e-06,
+      "loss": 0.1305,
+      "step": 12871
+    },
+    {
+      "epoch": 3.8621802085677848,
+      "grad_norm": 0.78515625,
+      "learning_rate": 6.868504292815184e-06,
+      "loss": 0.0933,
+      "step": 12872
+    },
+    {
+      "epoch": 3.8624803060994823,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.853441783401116e-06,
+      "loss": 0.0904,
+      "step": 12873
+    },
+    {
+      "epoch": 3.86278040363118,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.838379273987048e-06,
+      "loss": 0.0922,
+      "step": 12874
+    },
+    {
+      "epoch": 3.863080501162878,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.823316764572978e-06,
+      "loss": 0.1007,
+      "step": 12875
+    },
+    {
+      "epoch": 3.863380598694576,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.80825425515891e-06,
+      "loss": 0.0722,
+      "step": 12876
+    },
+    {
+      "epoch": 3.8636806962262735,
+      "grad_norm": 0.96484375,
+      "learning_rate": 6.7931917457448414e-06,
+      "loss": 0.1034,
+      "step": 12877
+    },
+    {
+      "epoch": 3.863980793757971,
+      "grad_norm": 0.8828125,
+      "learning_rate": 6.778129236330772e-06,
+      "loss": 0.0988,
+      "step": 12878
+    },
+    {
+      "epoch": 3.864280891289669,
+      "grad_norm": 0.75,
+      "learning_rate": 6.763066726916704e-06,
+      "loss": 0.095,
+      "step": 12879
+    },
+    {
+      "epoch": 3.864580988821367,
+      "grad_norm": 0.73046875,
+      "learning_rate": 6.748004217502636e-06,
+      "loss": 0.0972,
+      "step": 12880
+    },
+    {
+      "epoch": 3.8648810863530647,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.732941708088568e-06,
+      "loss": 0.1049,
+      "step": 12881
+    },
+    {
+      "epoch": 3.8651811838847623,
+      "grad_norm": 0.9296875,
+      "learning_rate": 6.7178791986745e-06,
+      "loss": 0.108,
+      "step": 12882
+    },
+    {
+      "epoch": 3.8654812814164603,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.702816689260432e-06,
+      "loss": 0.108,
+      "step": 12883
+    },
+    {
+      "epoch": 3.8657813789481583,
+      "grad_norm": 0.7421875,
+      "learning_rate": 6.6877541798463634e-06,
+      "loss": 0.1055,
+      "step": 12884
+    },
+    {
+      "epoch": 3.866081476479856,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.672691670432294e-06,
+      "loss": 0.102,
+      "step": 12885
+    },
+    {
+      "epoch": 3.866381574011554,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.657629161018226e-06,
+      "loss": 0.0741,
+      "step": 12886
+    },
+    {
+      "epoch": 3.8666816715432515,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.642566651604158e-06,
+      "loss": 0.0942,
+      "step": 12887
+    },
+    {
+      "epoch": 3.8669817690749495,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.627504142190088e-06,
+      "loss": 0.1008,
+      "step": 12888
+    },
+    {
+      "epoch": 3.867281866606647,
+      "grad_norm": 0.84375,
+      "learning_rate": 6.61244163277602e-06,
+      "loss": 0.1039,
+      "step": 12889
+    },
+    {
+      "epoch": 3.867581964138345,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.597379123361952e-06,
+      "loss": 0.1067,
+      "step": 12890
+    },
+    {
+      "epoch": 3.8678820616700427,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.582316613947884e-06,
+      "loss": 0.0932,
+      "step": 12891
+    },
+    {
+      "epoch": 3.8681821592017407,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.5672541045338156e-06,
+      "loss": 0.0843,
+      "step": 12892
+    },
+    {
+      "epoch": 3.8684822567334383,
+      "grad_norm": 1.09375,
+      "learning_rate": 6.552191595119747e-06,
+      "loss": 0.1173,
+      "step": 12893
+    },
+    {
+      "epoch": 3.8687823542651363,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.537129085705679e-06,
+      "loss": 0.1126,
+      "step": 12894
+    },
+    {
+      "epoch": 3.869082451796834,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.52206657629161e-06,
+      "loss": 0.0874,
+      "step": 12895
+    },
+    {
+      "epoch": 3.869382549328532,
+      "grad_norm": 1.03125,
+      "learning_rate": 6.507004066877542e-06,
+      "loss": 0.1087,
+      "step": 12896
+    },
+    {
+      "epoch": 3.8696826468602294,
+      "grad_norm": 0.875,
+      "learning_rate": 6.491941557463474e-06,
+      "loss": 0.103,
+      "step": 12897
+    },
+    {
+      "epoch": 3.8699827443919275,
+      "grad_norm": 0.94140625,
+      "learning_rate": 6.476879048049406e-06,
+      "loss": 0.1119,
+      "step": 12898
+    },
+    {
+      "epoch": 3.870282841923625,
+      "grad_norm": 0.83984375,
+      "learning_rate": 6.4618165386353376e-06,
+      "loss": 0.1066,
+      "step": 12899
+    },
+    {
+      "epoch": 3.870582939455323,
+      "grad_norm": 0.95703125,
+      "learning_rate": 6.446754029221269e-06,
+      "loss": 0.0999,
+      "step": 12900
+    },
+    {
+      "epoch": 3.8708830369870206,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.431691519807201e-06,
+      "loss": 0.1017,
+      "step": 12901
+    },
+    {
+      "epoch": 3.8711831345187186,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.416629010393131e-06,
+      "loss": 0.0949,
+      "step": 12902
+    },
+    {
+      "epoch": 3.8714832320504162,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.401566500979063e-06,
+      "loss": 0.1128,
+      "step": 12903
+    },
+    {
+      "epoch": 3.8717833295821142,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.386503991564994e-06,
+      "loss": 0.0921,
+      "step": 12904
+    },
+    {
+      "epoch": 3.872083427113812,
+      "grad_norm": 6.625,
+      "learning_rate": 6.371441482150926e-06,
+      "loss": 0.7373,
+      "step": 12905
+    },
+    {
+      "epoch": 3.87238352464551,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.356378972736858e-06,
+      "loss": 0.079,
+      "step": 12906
+    },
+    {
+      "epoch": 3.872683622177208,
+      "grad_norm": 0.75,
+      "learning_rate": 6.34131646332279e-06,
+      "loss": 0.0899,
+      "step": 12907
+    },
+    {
+      "epoch": 3.8729837197089054,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.3262539539087215e-06,
+      "loss": 0.1017,
+      "step": 12908
+    },
+    {
+      "epoch": 3.873283817240603,
+      "grad_norm": 0.90234375,
+      "learning_rate": 6.311191444494653e-06,
+      "loss": 0.1264,
+      "step": 12909
+    },
+    {
+      "epoch": 3.873583914772301,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.296128935080585e-06,
+      "loss": 0.0848,
+      "step": 12910
+    },
+    {
+      "epoch": 3.873884012303999,
+      "grad_norm": 0.87109375,
+      "learning_rate": 6.281066425666517e-06,
+      "loss": 0.1084,
+      "step": 12911
+    },
+    {
+      "epoch": 3.8741841098356966,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.266003916252448e-06,
+      "loss": 0.107,
+      "step": 12912
+    },
+    {
+      "epoch": 3.874484207367394,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.25094140683838e-06,
+      "loss": 0.1042,
+      "step": 12913
+    },
+    {
+      "epoch": 3.874784304899092,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.235878897424311e-06,
+      "loss": 0.0982,
+      "step": 12914
+    },
+    {
+      "epoch": 3.87508440243079,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.220816388010243e-06,
+      "loss": 0.0974,
+      "step": 12915
+    },
+    {
+      "epoch": 3.875384499962488,
+      "grad_norm": 0.74609375,
+      "learning_rate": 6.2057538785961745e-06,
+      "loss": 0.087,
+      "step": 12916
+    },
+    {
+      "epoch": 3.8756845974941854,
+      "grad_norm": 1.015625,
+      "learning_rate": 6.190691369182106e-06,
+      "loss": 0.1242,
+      "step": 12917
+    },
+    {
+      "epoch": 3.8759846950258834,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.175628859768038e-06,
+      "loss": 0.0919,
+      "step": 12918
+    },
+    {
+      "epoch": 3.8762847925575814,
+      "grad_norm": 0.796875,
+      "learning_rate": 6.160566350353969e-06,
+      "loss": 0.1061,
+      "step": 12919
+    },
+    {
+      "epoch": 3.876584890089279,
+      "grad_norm": 0.90625,
+      "learning_rate": 6.145503840939901e-06,
+      "loss": 0.1116,
+      "step": 12920
+    },
+    {
+      "epoch": 3.8768849876209766,
+      "grad_norm": 0.7578125,
+      "learning_rate": 6.130441331525833e-06,
+      "loss": 0.0946,
+      "step": 12921
+    },
+    {
+      "epoch": 3.8771850851526746,
+      "grad_norm": 0.921875,
+      "learning_rate": 6.115378822111764e-06,
+      "loss": 0.0991,
+      "step": 12922
+    },
+    {
+      "epoch": 3.8774851826843726,
+      "grad_norm": 0.9140625,
+      "learning_rate": 6.100316312697696e-06,
+      "loss": 0.1302,
+      "step": 12923
+    },
+    {
+      "epoch": 3.87778528021607,
+      "grad_norm": 0.8125,
+      "learning_rate": 6.0852538032836275e-06,
+      "loss": 0.1046,
+      "step": 12924
+    },
+    {
+      "epoch": 3.878085377747768,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.0701912938695585e-06,
+      "loss": 0.0982,
+      "step": 12925
+    },
+    {
+      "epoch": 3.8783854752794658,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.05512878445549e-06,
+      "loss": 0.0958,
+      "step": 12926
+    },
+    {
+      "epoch": 3.8786855728111638,
+      "grad_norm": 0.91796875,
+      "learning_rate": 6.040066275041422e-06,
+      "loss": 0.1044,
+      "step": 12927
+    },
+    {
+      "epoch": 3.8789856703428613,
+      "grad_norm": 0.87890625,
+      "learning_rate": 6.025003765627354e-06,
+      "loss": 0.1099,
+      "step": 12928
+    },
+    {
+      "epoch": 3.8792857678745594,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.009941256213286e-06,
+      "loss": 0.0832,
+      "step": 12929
+    },
+    {
+      "epoch": 3.879585865406257,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.994878746799217e-06,
+      "loss": 0.0978,
+      "step": 12930
+    },
+    {
+      "epoch": 3.879885962937955,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.979816237385149e-06,
+      "loss": 0.0857,
+      "step": 12931
+    },
+    {
+      "epoch": 3.8801860604696525,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.96475372797108e-06,
+      "loss": 0.1142,
+      "step": 12932
+    },
+    {
+      "epoch": 3.8804861580013505,
+      "grad_norm": 0.875,
+      "learning_rate": 5.9496912185570115e-06,
+      "loss": 0.109,
+      "step": 12933
+    },
+    {
+      "epoch": 3.880786255533048,
+      "grad_norm": 0.8359375,
+      "learning_rate": 5.934628709142943e-06,
+      "loss": 0.0995,
+      "step": 12934
+    },
+    {
+      "epoch": 3.881086353064746,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.919566199728875e-06,
+      "loss": 0.0855,
+      "step": 12935
+    },
+    {
+      "epoch": 3.8813864505964437,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.904503690314807e-06,
+      "loss": 0.0926,
+      "step": 12936
+    },
+    {
+      "epoch": 3.8816865481281417,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.889441180900739e-06,
+      "loss": 0.1022,
+      "step": 12937
+    },
+    {
+      "epoch": 3.8819866456598393,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.874378671486671e-06,
+      "loss": 0.0884,
+      "step": 12938
+    },
+    {
+      "epoch": 3.8822867431915373,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.859316162072602e-06,
+      "loss": 0.0971,
+      "step": 12939
+    },
+    {
+      "epoch": 3.882586840723235,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.844253652658533e-06,
+      "loss": 0.0971,
+      "step": 12940
+    },
+    {
+      "epoch": 3.882886938254933,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.8291911432444645e-06,
+      "loss": 0.1148,
+      "step": 12941
+    },
+    {
+      "epoch": 3.8831870357866305,
+      "grad_norm": 0.84375,
+      "learning_rate": 5.814128633830396e-06,
+      "loss": 0.1312,
+      "step": 12942
+    },
+    {
+      "epoch": 3.8834871333183285,
+      "grad_norm": 0.78515625,
+      "learning_rate": 5.799066124416328e-06,
+      "loss": 0.1017,
+      "step": 12943
+    },
+    {
+      "epoch": 3.883787230850026,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.78400361500226e-06,
+      "loss": 0.0913,
+      "step": 12944
+    },
+    {
+      "epoch": 3.884087328381724,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.768941105588192e-06,
+      "loss": 0.0815,
+      "step": 12945
+    },
+    {
+      "epoch": 3.884387425913422,
+      "grad_norm": 0.88671875,
+      "learning_rate": 5.753878596174123e-06,
+      "loss": 0.1156,
+      "step": 12946
+    },
+    {
+      "epoch": 3.8846875234451197,
+      "grad_norm": 0.75,
+      "learning_rate": 5.738816086760055e-06,
+      "loss": 0.0866,
+      "step": 12947
+    },
+    {
+      "epoch": 3.8849876209768173,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.723753577345986e-06,
+      "loss": 0.1108,
+      "step": 12948
+    },
+    {
+      "epoch": 3.8852877185085153,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.7086910679319175e-06,
+      "loss": 0.0929,
+      "step": 12949
+    },
+    {
+      "epoch": 3.8855878160402133,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.693628558517849e-06,
+      "loss": 0.1051,
+      "step": 12950
+    },
+    {
+      "epoch": 3.885887913571911,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.678566049103781e-06,
+      "loss": 0.0912,
+      "step": 12951
+    },
+    {
+      "epoch": 3.8861880111036085,
+      "grad_norm": 0.828125,
+      "learning_rate": 5.663503539689713e-06,
+      "loss": 0.0927,
+      "step": 12952
+    },
+    {
+      "epoch": 3.8864881086353065,
+      "grad_norm": 0.78125,
+      "learning_rate": 5.648441030275644e-06,
+      "loss": 0.0915,
+      "step": 12953
+    },
+    {
+      "epoch": 3.8867882061670045,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.633378520861576e-06,
+      "loss": 0.0909,
+      "step": 12954
+    },
+    {
+      "epoch": 3.887088303698702,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.618316011447508e-06,
+      "loss": 0.0755,
+      "step": 12955
+    },
+    {
+      "epoch": 3.8873884012303996,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.6032535020334395e-06,
+      "loss": 0.0848,
+      "step": 12956
+    },
+    {
+      "epoch": 3.8876884987620977,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.5881909926193705e-06,
+      "loss": 0.1209,
+      "step": 12957
+    },
+    {
+      "epoch": 3.8879885962937957,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.573128483205302e-06,
+      "loss": 0.1124,
+      "step": 12958
+    },
+    {
+      "epoch": 3.8882886938254932,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.558065973791233e-06,
+      "loss": 0.0758,
+      "step": 12959
+    },
+    {
+      "epoch": 3.888588791357191,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.543003464377165e-06,
+      "loss": 0.1144,
+      "step": 12960
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.527940954963097e-06,
+      "loss": 0.1267,
+      "step": 12961
+    },
+    {
+      "epoch": 3.889188986420587,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.512878445549029e-06,
+      "loss": 0.0994,
+      "step": 12962
+    },
+    {
+      "epoch": 3.8894890839522844,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.497815936134961e-06,
+      "loss": 0.1159,
+      "step": 12963
+    },
+    {
+      "epoch": 3.8897891814839824,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.4827534267208925e-06,
+      "loss": 0.1122,
+      "step": 12964
+    },
+    {
+      "epoch": 3.89008927901568,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.4676909173068234e-06,
+      "loss": 0.1118,
+      "step": 12965
+    },
+    {
+      "epoch": 3.890389376547378,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.452628407892755e-06,
+      "loss": 0.0979,
+      "step": 12966
+    },
+    {
+      "epoch": 3.8906894740790756,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.437565898478686e-06,
+      "loss": 0.0841,
+      "step": 12967
+    },
+    {
+      "epoch": 3.8909895716107736,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.422503389064618e-06,
+      "loss": 0.0862,
+      "step": 12968
+    },
+    {
+      "epoch": 3.891289669142471,
+      "grad_norm": 0.921875,
+      "learning_rate": 5.40744087965055e-06,
+      "loss": 0.1051,
+      "step": 12969
+    },
+    {
+      "epoch": 3.8915897666741692,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.392378370236482e-06,
+      "loss": 0.0815,
+      "step": 12970
+    },
+    {
+      "epoch": 3.891889864205867,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.377315860822414e-06,
+      "loss": 0.083,
+      "step": 12971
+    },
+    {
+      "epoch": 3.892189961737565,
+      "grad_norm": 0.953125,
+      "learning_rate": 5.3622533514083454e-06,
+      "loss": 0.129,
+      "step": 12972
+    },
+    {
+      "epoch": 3.8924900592692624,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.3471908419942764e-06,
+      "loss": 0.0806,
+      "step": 12973
+    },
+    {
+      "epoch": 3.8927901568009604,
+      "grad_norm": 0.89453125,
+      "learning_rate": 5.332128332580208e-06,
+      "loss": 0.114,
+      "step": 12974
+    },
+    {
+      "epoch": 3.893090254332658,
+      "grad_norm": 0.85546875,
+      "learning_rate": 5.317065823166139e-06,
+      "loss": 0.1004,
+      "step": 12975
+    },
+    {
+      "epoch": 3.893390351864356,
+      "grad_norm": 0.75390625,
+      "learning_rate": 5.302003313752071e-06,
+      "loss": 0.0987,
+      "step": 12976
+    },
+    {
+      "epoch": 3.8936904493960536,
+      "grad_norm": 0.90234375,
+      "learning_rate": 5.286940804338003e-06,
+      "loss": 0.1152,
+      "step": 12977
+    },
+    {
+      "epoch": 3.8939905469277516,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.271878294923935e-06,
+      "loss": 0.1084,
+      "step": 12978
+    },
+    {
+      "epoch": 3.894290644459449,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.256815785509867e-06,
+      "loss": 0.0952,
+      "step": 12979
+    },
+    {
+      "epoch": 3.894590741991147,
+      "grad_norm": 0.8671875,
+      "learning_rate": 5.241753276095798e-06,
+      "loss": 0.1213,
+      "step": 12980
+    },
+    {
+      "epoch": 3.8948908395228448,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.226690766681729e-06,
+      "loss": 0.0837,
+      "step": 12981
+    },
+    {
+      "epoch": 3.895190937054543,
+      "grad_norm": 0.75,
+      "learning_rate": 5.211628257267661e-06,
+      "loss": 0.0768,
+      "step": 12982
+    },
+    {
+      "epoch": 3.8954910345862404,
+      "grad_norm": 0.9765625,
+      "learning_rate": 5.196565747853593e-06,
+      "loss": 0.1217,
+      "step": 12983
+    },
+    {
+      "epoch": 3.8957911321179384,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.181503238439524e-06,
+      "loss": 0.1046,
+      "step": 12984
+    },
+    {
+      "epoch": 3.8960912296496364,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.166440729025456e-06,
+      "loss": 0.0891,
+      "step": 12985
+    },
+    {
+      "epoch": 3.896391327181334,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.151378219611387e-06,
+      "loss": 0.0975,
+      "step": 12986
+    },
+    {
+      "epoch": 3.8966914247130315,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.136315710197319e-06,
+      "loss": 0.1016,
+      "step": 12987
+    },
+    {
+      "epoch": 3.8969915222447296,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.1212532007832506e-06,
+      "loss": 0.089,
+      "step": 12988
+    },
+    {
+      "epoch": 3.8972916197764276,
+      "grad_norm": 0.97265625,
+      "learning_rate": 5.106190691369182e-06,
+      "loss": 0.1183,
+      "step": 12989
+    },
+    {
+      "epoch": 3.897591717308125,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.091128181955114e-06,
+      "loss": 0.104,
+      "step": 12990
+    },
+    {
+      "epoch": 3.8978918148398227,
+      "grad_norm": 0.8203125,
+      "learning_rate": 5.076065672541046e-06,
+      "loss": 0.1059,
+      "step": 12991
+    },
+    {
+      "epoch": 3.8981919123715207,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5.061003163126977e-06,
+      "loss": 0.1053,
+      "step": 12992
+    },
+    {
+      "epoch": 3.8984920099032188,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.045940653712909e-06,
+      "loss": 0.1283,
+      "step": 12993
+    },
+    {
+      "epoch": 3.8987921074349163,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.03087814429884e-06,
+      "loss": 0.088,
+      "step": 12994
+    },
+    {
+      "epoch": 3.899092204966614,
+      "grad_norm": 1.125,
+      "learning_rate": 5.015815634884772e-06,
+      "loss": 0.1766,
+      "step": 12995
+    },
+    {
+      "epoch": 3.899392302498312,
+      "grad_norm": 0.87109375,
+      "learning_rate": 5.0007531254707036e-06,
+      "loss": 0.1097,
+      "step": 12996
+    },
+    {
+      "epoch": 3.89969240003001,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.985690616056635e-06,
+      "loss": 0.108,
+      "step": 12997
+    },
+    {
+      "epoch": 3.8999924975617075,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.970628106642567e-06,
+      "loss": 0.1082,
+      "step": 12998
+    },
+    {
+      "epoch": 3.900292595093405,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.955565597228499e-06,
+      "loss": 0.0841,
+      "step": 12999
+    },
+    {
+      "epoch": 3.900592692625103,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.94050308781443e-06,
+      "loss": 0.1073,
+      "step": 13000
+    },
+    {
+      "epoch": 3.900892790156801,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.925440578400362e-06,
+      "loss": 0.1004,
+      "step": 13001
+    },
+    {
+      "epoch": 3.9011928876884987,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.910378068986293e-06,
+      "loss": 0.1125,
+      "step": 13002
+    },
+    {
+      "epoch": 3.9014929852201967,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.895315559572225e-06,
+      "loss": 0.1159,
+      "step": 13003
+    },
+    {
+      "epoch": 3.9017930827518943,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.8802530501581566e-06,
+      "loss": 0.0965,
+      "step": 13004
+    },
+    {
+      "epoch": 3.9020931802835923,
+      "grad_norm": 0.875,
+      "learning_rate": 4.865190540744088e-06,
+      "loss": 0.1194,
+      "step": 13005
+    },
+    {
+      "epoch": 3.90239327781529,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.85012803133002e-06,
+      "loss": 0.1195,
+      "step": 13006
+    },
+    {
+      "epoch": 3.902693375346988,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.835065521915951e-06,
+      "loss": 0.0984,
+      "step": 13007
+    },
+    {
+      "epoch": 3.9029934728786855,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.820003012501883e-06,
+      "loss": 0.1003,
+      "step": 13008
+    },
+    {
+      "epoch": 3.9032935704103835,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.804940503087815e-06,
+      "loss": 0.0905,
+      "step": 13009
+    },
+    {
+      "epoch": 3.903593667942081,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.789877993673747e-06,
+      "loss": 0.0923,
+      "step": 13010
+    },
+    {
+      "epoch": 3.903893765473779,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.774815484259678e-06,
+      "loss": 0.0973,
+      "step": 13011
+    },
+    {
+      "epoch": 3.9041938630054767,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.7597529748456095e-06,
+      "loss": 0.1211,
+      "step": 13012
+    },
+    {
+      "epoch": 3.9044939605371747,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.7446904654315405e-06,
+      "loss": 0.101,
+      "step": 13013
+    },
+    {
+      "epoch": 3.9047940580688723,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.729627956017472e-06,
+      "loss": 0.1171,
+      "step": 13014
+    },
+    {
+      "epoch": 3.9050941556005703,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.714565446603404e-06,
+      "loss": 0.1086,
+      "step": 13015
+    },
+    {
+      "epoch": 3.905394253132268,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.699502937189336e-06,
+      "loss": 0.0966,
+      "step": 13016
+    },
+    {
+      "epoch": 3.905694350663966,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.684440427775268e-06,
+      "loss": 0.1108,
+      "step": 13017
+    },
+    {
+      "epoch": 3.9059944481956634,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.6693779183612e-06,
+      "loss": 0.098,
+      "step": 13018
+    },
+    {
+      "epoch": 3.9062945457273615,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.654315408947131e-06,
+      "loss": 0.0863,
+      "step": 13019
+    },
+    {
+      "epoch": 3.906594643259059,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.6392528995330625e-06,
+      "loss": 0.1158,
+      "step": 13020
+    },
+    {
+      "epoch": 3.906894740790757,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.6241903901189935e-06,
+      "loss": 0.1142,
+      "step": 13021
+    },
+    {
+      "epoch": 3.9071948383224546,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.609127880704925e-06,
+      "loss": 0.1238,
+      "step": 13022
+    },
+    {
+      "epoch": 3.9074949358541526,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.594065371290857e-06,
+      "loss": 0.1013,
+      "step": 13023
+    },
+    {
+      "epoch": 3.9077950333858507,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.579002861876789e-06,
+      "loss": 0.0818,
+      "step": 13024
+    },
+    {
+      "epoch": 3.9080951309175482,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.563940352462721e-06,
+      "loss": 0.0805,
+      "step": 13025
+    },
+    {
+      "epoch": 3.908395228449246,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.548877843048653e-06,
+      "loss": 0.077,
+      "step": 13026
+    },
+    {
+      "epoch": 3.908695325980944,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.533815333634584e-06,
+      "loss": 0.0833,
+      "step": 13027
+    },
+    {
+      "epoch": 3.908995423512642,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.5187528242205155e-06,
+      "loss": 0.1019,
+      "step": 13028
+    },
+    {
+      "epoch": 3.9092955210443394,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.5036903148064465e-06,
+      "loss": 0.0875,
+      "step": 13029
+    },
+    {
+      "epoch": 3.909595618576037,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.488627805392378e-06,
+      "loss": 0.0847,
+      "step": 13030
+    },
+    {
+      "epoch": 3.909895716107735,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.47356529597831e-06,
+      "loss": 0.0953,
+      "step": 13031
+    },
+    {
+      "epoch": 3.910195813639433,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.458502786564242e-06,
+      "loss": 0.0931,
+      "step": 13032
+    },
+    {
+      "epoch": 3.9104959111711306,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.443440277150174e-06,
+      "loss": 0.1044,
+      "step": 13033
+    },
+    {
+      "epoch": 3.910796008702828,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.428377767736105e-06,
+      "loss": 0.0934,
+      "step": 13034
+    },
+    {
+      "epoch": 3.911096106234526,
+      "grad_norm": 0.875,
+      "learning_rate": 4.413315258322037e-06,
+      "loss": 0.097,
+      "step": 13035
+    },
+    {
+      "epoch": 3.911396203766224,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.3982527489079685e-06,
+      "loss": 0.0974,
+      "step": 13036
+    },
+    {
+      "epoch": 3.911696301297922,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.3831902394939e-06,
+      "loss": 0.1009,
+      "step": 13037
+    },
+    {
+      "epoch": 3.9119963988296194,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.368127730079831e-06,
+      "loss": 0.0897,
+      "step": 13038
+    },
+    {
+      "epoch": 3.9122964963613174,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.353065220665763e-06,
+      "loss": 0.1032,
+      "step": 13039
+    },
+    {
+      "epoch": 3.9125965938930154,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.338002711251695e-06,
+      "loss": 0.097,
+      "step": 13040
+    },
+    {
+      "epoch": 3.912896691424713,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.322940201837626e-06,
+      "loss": 0.0955,
+      "step": 13041
+    },
+    {
+      "epoch": 3.913196788956411,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.307877692423558e-06,
+      "loss": 0.0946,
+      "step": 13042
+    },
+    {
+      "epoch": 3.9134968864881086,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.29281518300949e-06,
+      "loss": 0.0795,
+      "step": 13043
+    },
+    {
+      "epoch": 3.9137969840198066,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.2777526735954215e-06,
+      "loss": 0.0974,
+      "step": 13044
+    },
+    {
+      "epoch": 3.914097081551504,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.262690164181353e-06,
+      "loss": 0.1088,
+      "step": 13045
+    },
+    {
+      "epoch": 3.914397179083202,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.247627654767284e-06,
+      "loss": 0.0938,
+      "step": 13046
+    },
+    {
+      "epoch": 3.9146972766148997,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.232565145353215e-06,
+      "loss": 0.0895,
+      "step": 13047
+    },
+    {
+      "epoch": 3.9149973741465978,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.217502635939147e-06,
+      "loss": 0.0924,
+      "step": 13048
+    },
+    {
+      "epoch": 3.9152974716782953,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.202440126525079e-06,
+      "loss": 0.1282,
+      "step": 13049
+    },
+    {
+      "epoch": 3.9155975692099934,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.187377617111011e-06,
+      "loss": 0.1039,
+      "step": 13050
+    },
+    {
+      "epoch": 3.915897666741691,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.172315107696943e-06,
+      "loss": 0.1055,
+      "step": 13051
+    },
+    {
+      "epoch": 3.916197764273389,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.1572525982828745e-06,
+      "loss": 0.0865,
+      "step": 13052
+    },
+    {
+      "epoch": 3.9164978618050865,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.142190088868806e-06,
+      "loss": 0.0922,
+      "step": 13053
+    },
+    {
+      "epoch": 3.9167979593367845,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.127127579454737e-06,
+      "loss": 0.1442,
+      "step": 13054
+    },
+    {
+      "epoch": 3.917098056868482,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.112065070040669e-06,
+      "loss": 0.1005,
+      "step": 13055
+    },
+    {
+      "epoch": 3.91739815440018,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.0970025606266e-06,
+      "loss": 0.104,
+      "step": 13056
+    },
+    {
+      "epoch": 3.9176982519318777,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.081940051212532e-06,
+      "loss": 0.1083,
+      "step": 13057
+    },
+    {
+      "epoch": 3.9179983494635757,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.066877541798464e-06,
+      "loss": 0.0936,
+      "step": 13058
+    },
+    {
+      "epoch": 3.9182984469952733,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.051815032384396e-06,
+      "loss": 0.098,
+      "step": 13059
+    },
+    {
+      "epoch": 3.9185985445269713,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.0367525229703275e-06,
+      "loss": 0.107,
+      "step": 13060
+    },
+    {
+      "epoch": 3.918898642058669,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.0216900135562585e-06,
+      "loss": 0.0832,
+      "step": 13061
+    },
+    {
+      "epoch": 3.919198739590367,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.00662750414219e-06,
+      "loss": 0.1002,
+      "step": 13062
+    },
+    {
+      "epoch": 3.919498837122065,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.991564994728122e-06,
+      "loss": 0.0948,
+      "step": 13063
+    },
+    {
+      "epoch": 3.9197989346537625,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.976502485314054e-06,
+      "loss": 0.0969,
+      "step": 13064
+    },
+    {
+      "epoch": 3.92009903218546,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.961439975899985e-06,
+      "loss": 0.0987,
+      "step": 13065
+    },
+    {
+      "epoch": 3.920399129717158,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.946377466485917e-06,
+      "loss": 0.1004,
+      "step": 13066
+    },
+    {
+      "epoch": 3.920699227248856,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.931314957071849e-06,
+      "loss": 0.0988,
+      "step": 13067
+    },
+    {
+      "epoch": 3.9209993247805537,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.91625244765778e-06,
+      "loss": 0.1302,
+      "step": 13068
+    },
+    {
+      "epoch": 3.9212994223122513,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.9011899382437114e-06,
+      "loss": 0.0902,
+      "step": 13069
+    },
+    {
+      "epoch": 3.9215995198439493,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.886127428829643e-06,
+      "loss": 0.0906,
+      "step": 13070
+    },
+    {
+      "epoch": 3.9218996173756473,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.871064919415575e-06,
+      "loss": 0.106,
+      "step": 13071
+    },
+    {
+      "epoch": 3.922199714907345,
+      "grad_norm": 0.875,
+      "learning_rate": 3.856002410001507e-06,
+      "loss": 0.1147,
+      "step": 13072
+    },
+    {
+      "epoch": 3.9224998124390424,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.840939900587438e-06,
+      "loss": 0.1026,
+      "step": 13073
+    },
+    {
+      "epoch": 3.9227999099707405,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.825877391173369e-06,
+      "loss": 0.1095,
+      "step": 13074
+    },
+    {
+      "epoch": 3.9231000075024385,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.810814881759301e-06,
+      "loss": 0.1119,
+      "step": 13075
+    },
+    {
+      "epoch": 3.923400105034136,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.7957523723452326e-06,
+      "loss": 0.1053,
+      "step": 13076
+    },
+    {
+      "epoch": 3.9237002025658336,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.7806898629311644e-06,
+      "loss": 0.0961,
+      "step": 13077
+    },
+    {
+      "epoch": 3.9240003000975316,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.7656273535170963e-06,
+      "loss": 0.0899,
+      "step": 13078
+    },
+    {
+      "epoch": 3.9243003976292297,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.750564844103028e-06,
+      "loss": 0.1128,
+      "step": 13079
+    },
+    {
+      "epoch": 3.9246004951609272,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.7355023346889595e-06,
+      "loss": 0.1002,
+      "step": 13080
+    },
+    {
+      "epoch": 3.9249005926926253,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.7204398252748905e-06,
+      "loss": 0.0827,
+      "step": 13081
+    },
+    {
+      "epoch": 3.925200690224323,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.7053773158608223e-06,
+      "loss": 0.0937,
+      "step": 13082
+    },
+    {
+      "epoch": 3.925500787756021,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.690314806446754e-06,
+      "loss": 0.0862,
+      "step": 13083
+    },
+    {
+      "epoch": 3.9258008852877184,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.675252297032686e-06,
+      "loss": 0.1045,
+      "step": 13084
+    },
+    {
+      "epoch": 3.9261009828194164,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.6601897876186174e-06,
+      "loss": 0.1315,
+      "step": 13085
+    },
+    {
+      "epoch": 3.926401080351114,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.6451272782045492e-06,
+      "loss": 0.0846,
+      "step": 13086
+    },
+    {
+      "epoch": 3.926701177882812,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.630064768790481e-06,
+      "loss": 0.1157,
+      "step": 13087
+    },
+    {
+      "epoch": 3.9270012754145096,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.615002259376412e-06,
+      "loss": 0.1342,
+      "step": 13088
+    },
+    {
+      "epoch": 3.9273013729462076,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.599939749962344e-06,
+      "loss": 0.0851,
+      "step": 13089
+    },
+    {
+      "epoch": 3.927601470477905,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.5848772405482753e-06,
+      "loss": 0.0868,
+      "step": 13090
+    },
+    {
+      "epoch": 3.927901568009603,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.569814731134207e-06,
+      "loss": 0.1125,
+      "step": 13091
+    },
+    {
+      "epoch": 3.928201665541301,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.554752221720139e-06,
+      "loss": 0.1004,
+      "step": 13092
+    },
+    {
+      "epoch": 3.928501763072999,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.5396897123060704e-06,
+      "loss": 0.1136,
+      "step": 13093
+    },
+    {
+      "epoch": 3.9288018606046964,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.5246272028920022e-06,
+      "loss": 0.1177,
+      "step": 13094
+    },
+    {
+      "epoch": 3.9291019581363944,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.5095646934779332e-06,
+      "loss": 0.1122,
+      "step": 13095
+    },
+    {
+      "epoch": 3.929402055668092,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.494502184063865e-06,
+      "loss": 0.1037,
+      "step": 13096
+    },
+    {
+      "epoch": 3.92970215319979,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.479439674649797e-06,
+      "loss": 0.1172,
+      "step": 13097
+    },
+    {
+      "epoch": 3.9300022507314876,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.4643771652357283e-06,
+      "loss": 0.1256,
+      "step": 13098
+    },
+    {
+      "epoch": 3.9303023482631856,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.44931465582166e-06,
+      "loss": 0.1174,
+      "step": 13099
+    },
+    {
+      "epoch": 3.930602445794883,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.434252146407592e-06,
+      "loss": 0.0972,
+      "step": 13100
+    },
+    {
+      "epoch": 3.930902543326581,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.419189636993524e-06,
+      "loss": 0.0853,
+      "step": 13101
+    },
+    {
+      "epoch": 3.931202640858279,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.404127127579455e-06,
+      "loss": 0.0826,
+      "step": 13102
+    },
+    {
+      "epoch": 3.9315027383899768,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.389064618165386e-06,
+      "loss": 0.1078,
+      "step": 13103
+    },
+    {
+      "epoch": 3.9318028359216743,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.374002108751318e-06,
+      "loss": 0.0921,
+      "step": 13104
+    },
+    {
+      "epoch": 3.9321029334533724,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.35893959933725e-06,
+      "loss": 0.0942,
+      "step": 13105
+    },
+    {
+      "epoch": 3.9324030309850704,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.3438770899231817e-06,
+      "loss": 0.0864,
+      "step": 13106
+    },
+    {
+      "epoch": 3.932703128516768,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.328814580509113e-06,
+      "loss": 0.0924,
+      "step": 13107
+    },
+    {
+      "epoch": 3.9330032260484655,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.313752071095044e-06,
+      "loss": 0.1136,
+      "step": 13108
+    },
+    {
+      "epoch": 3.9333033235801635,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.298689561680976e-06,
+      "loss": 0.1019,
+      "step": 13109
+    },
+    {
+      "epoch": 3.9336034211118616,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.2836270522669078e-06,
+      "loss": 0.11,
+      "step": 13110
+    },
+    {
+      "epoch": 3.933903518643559,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.2685645428528396e-06,
+      "loss": 0.0978,
+      "step": 13111
+    },
+    {
+      "epoch": 3.9342036161752567,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.253502033438771e-06,
+      "loss": 0.1017,
+      "step": 13112
+    },
+    {
+      "epoch": 3.9345037137069547,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.238439524024703e-06,
+      "loss": 0.0944,
+      "step": 13113
+    },
+    {
+      "epoch": 3.9348038112386527,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.2233770146106347e-06,
+      "loss": 0.0781,
+      "step": 13114
+    },
+    {
+      "epoch": 3.9351039087703503,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.2083145051965657e-06,
+      "loss": 0.089,
+      "step": 13115
+    },
+    {
+      "epoch": 3.935404006302048,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.193251995782497e-06,
+      "loss": 0.1047,
+      "step": 13116
+    },
+    {
+      "epoch": 3.935704103833746,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.178189486368429e-06,
+      "loss": 0.2125,
+      "step": 13117
+    },
+    {
+      "epoch": 3.936004201365444,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.1631269769543608e-06,
+      "loss": 0.1066,
+      "step": 13118
+    },
+    {
+      "epoch": 3.9363042988971415,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.1480644675402926e-06,
+      "loss": 0.0835,
+      "step": 13119
+    },
+    {
+      "epoch": 3.9366043964288395,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.133001958126224e-06,
+      "loss": 0.0995,
+      "step": 13120
+    },
+    {
+      "epoch": 3.936904493960537,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.1179394487121554e-06,
+      "loss": 0.0946,
+      "step": 13121
+    },
+    {
+      "epoch": 3.937204591492235,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.1028769392980873e-06,
+      "loss": 0.0836,
+      "step": 13122
+    },
+    {
+      "epoch": 3.9375046890239327,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.087814429884019e-06,
+      "loss": 0.095,
+      "step": 13123
+    },
+    {
+      "epoch": 3.9378047865556307,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.0727519204699505e-06,
+      "loss": 0.1054,
+      "step": 13124
+    },
+    {
+      "epoch": 3.9381048840873283,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.057689411055882e-06,
+      "loss": 0.1156,
+      "step": 13125
+    },
+    {
+      "epoch": 3.9384049816190263,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.0426269016418138e-06,
+      "loss": 0.1338,
+      "step": 13126
+    },
+    {
+      "epoch": 3.938705079150724,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.027564392227745e-06,
+      "loss": 0.1197,
+      "step": 13127
+    },
+    {
+      "epoch": 3.939005176682422,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.012501882813677e-06,
+      "loss": 0.0869,
+      "step": 13128
+    },
+    {
+      "epoch": 3.9393052742141195,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.9974393733996084e-06,
+      "loss": 0.1115,
+      "step": 13129
+    },
+    {
+      "epoch": 3.9396053717458175,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.98237686398554e-06,
+      "loss": 0.0829,
+      "step": 13130
+    },
+    {
+      "epoch": 3.939905469277515,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.9673143545714717e-06,
+      "loss": 0.113,
+      "step": 13131
+    },
+    {
+      "epoch": 3.940205566809213,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.9522518451574035e-06,
+      "loss": 0.107,
+      "step": 13132
+    },
+    {
+      "epoch": 3.9405056643409107,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.9371893357433353e-06,
+      "loss": 0.1167,
+      "step": 13133
+    },
+    {
+      "epoch": 3.9408057618726087,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.9221268263292663e-06,
+      "loss": 0.1106,
+      "step": 13134
+    },
+    {
+      "epoch": 3.9411058594043062,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.907064316915198e-06,
+      "loss": 0.1078,
+      "step": 13135
+    },
+    {
+      "epoch": 3.9414059569360043,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.89200180750113e-06,
+      "loss": 0.1019,
+      "step": 13136
+    },
+    {
+      "epoch": 3.9417060544677023,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.8769392980870614e-06,
+      "loss": 0.0955,
+      "step": 13137
+    },
+    {
+      "epoch": 3.9420061519994,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.861876788672993e-06,
+      "loss": 0.117,
+      "step": 13138
+    },
+    {
+      "epoch": 3.9423062495310974,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.8468142792589247e-06,
+      "loss": 0.0966,
+      "step": 13139
+    },
+    {
+      "epoch": 3.9426063470627954,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.8317517698448565e-06,
+      "loss": 0.1112,
+      "step": 13140
+    },
+    {
+      "epoch": 3.9429064445944935,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.816689260430788e-06,
+      "loss": 0.0833,
+      "step": 13141
+    },
+    {
+      "epoch": 3.943206542126191,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.8016267510167197e-06,
+      "loss": 0.1075,
+      "step": 13142
+    },
+    {
+      "epoch": 3.9435066396578886,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.786564241602651e-06,
+      "loss": 0.1145,
+      "step": 13143
+    },
+    {
+      "epoch": 3.9438067371895866,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.7715017321885826e-06,
+      "loss": 0.1142,
+      "step": 13144
+    },
+    {
+      "epoch": 3.9441068347212846,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.7564392227745144e-06,
+      "loss": 0.1167,
+      "step": 13145
+    },
+    {
+      "epoch": 3.9444069322529822,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.7413767133604462e-06,
+      "loss": 0.1097,
+      "step": 13146
+    },
+    {
+      "epoch": 3.94470702978468,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.7263142039463776e-06,
+      "loss": 0.0879,
+      "step": 13147
+    },
+    {
+      "epoch": 3.945007127316378,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.711251694532309e-06,
+      "loss": 0.1171,
+      "step": 13148
+    },
+    {
+      "epoch": 3.945307224848076,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.696189185118241e-06,
+      "loss": 0.0977,
+      "step": 13149
+    },
+    {
+      "epoch": 3.9456073223797734,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.6811266757041727e-06,
+      "loss": 0.1106,
+      "step": 13150
+    },
+    {
+      "epoch": 3.945907419911471,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.666064166290104e-06,
+      "loss": 0.0957,
+      "step": 13151
+    },
+    {
+      "epoch": 3.946207517443169,
+      "grad_norm": 0.875,
+      "learning_rate": 2.6510016568760355e-06,
+      "loss": 0.0982,
+      "step": 13152
+    },
+    {
+      "epoch": 3.946507614974867,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6359391474619674e-06,
+      "loss": 0.085,
+      "step": 13153
+    },
+    {
+      "epoch": 3.9468077125065646,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.620876638047899e-06,
+      "loss": 0.1116,
+      "step": 13154
+    },
+    {
+      "epoch": 3.947107810038262,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.6058141286338306e-06,
+      "loss": 0.1127,
+      "step": 13155
+    },
+    {
+      "epoch": 3.94740790756996,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.590751619219762e-06,
+      "loss": 0.0672,
+      "step": 13156
+    },
+    {
+      "epoch": 3.947708005101658,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.5756891098056935e-06,
+      "loss": 0.1175,
+      "step": 13157
+    },
+    {
+      "epoch": 3.9480081026333558,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.5606266003916253e-06,
+      "loss": 0.1057,
+      "step": 13158
+    },
+    {
+      "epoch": 3.948308200165054,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.545564090977557e-06,
+      "loss": 0.0798,
+      "step": 13159
+    },
+    {
+      "epoch": 3.9486082976967514,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.5305015815634885e-06,
+      "loss": 0.0807,
+      "step": 13160
+    },
+    {
+      "epoch": 3.9489083952284494,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.51543907214942e-06,
+      "loss": 0.1157,
+      "step": 13161
+    },
+    {
+      "epoch": 3.949208492760147,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.5003765627353518e-06,
+      "loss": 0.1172,
+      "step": 13162
+    },
+    {
+      "epoch": 3.949508590291845,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.4853140533212836e-06,
+      "loss": 0.1015,
+      "step": 13163
+    },
+    {
+      "epoch": 3.9498086878235426,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.470251543907215e-06,
+      "loss": 0.0971,
+      "step": 13164
+    },
+    {
+      "epoch": 3.9501087853552406,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.4551890344931464e-06,
+      "loss": 0.1153,
+      "step": 13165
+    },
+    {
+      "epoch": 3.950408882886938,
+      "grad_norm": 0.875,
+      "learning_rate": 2.4401265250790783e-06,
+      "loss": 0.0988,
+      "step": 13166
+    },
+    {
+      "epoch": 3.950708980418636,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.42506401566501e-06,
+      "loss": 0.1134,
+      "step": 13167
+    },
+    {
+      "epoch": 3.9510090779503337,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4100015062509415e-06,
+      "loss": 0.0805,
+      "step": 13168
+    },
+    {
+      "epoch": 3.9513091754820318,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.3949389968368734e-06,
+      "loss": 0.1132,
+      "step": 13169
+    },
+    {
+      "epoch": 3.9516092730137293,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.3798764874228048e-06,
+      "loss": 0.1262,
+      "step": 13170
+    },
+    {
+      "epoch": 3.9519093705454273,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.364813978008736e-06,
+      "loss": 0.0952,
+      "step": 13171
+    },
+    {
+      "epoch": 3.952209468077125,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.349751468594668e-06,
+      "loss": 0.0829,
+      "step": 13172
+    },
+    {
+      "epoch": 3.952509565608823,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.3346889591806e-06,
+      "loss": 0.0951,
+      "step": 13173
+    },
+    {
+      "epoch": 3.9528096631405205,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.3196264497665313e-06,
+      "loss": 0.1022,
+      "step": 13174
+    },
+    {
+      "epoch": 3.9531097606722185,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.3045639403524627e-06,
+      "loss": 0.0927,
+      "step": 13175
+    },
+    {
+      "epoch": 3.9534098582039165,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.2895014309383945e-06,
+      "loss": 0.1287,
+      "step": 13176
+    },
+    {
+      "epoch": 3.953709955735614,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.2744389215243263e-06,
+      "loss": 0.1149,
+      "step": 13177
+    },
+    {
+      "epoch": 3.9540100532673117,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.2593764121102578e-06,
+      "loss": 0.1251,
+      "step": 13178
+    },
+    {
+      "epoch": 3.9543101507990097,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.244313902696189e-06,
+      "loss": 0.0923,
+      "step": 13179
+    },
+    {
+      "epoch": 3.9546102483307077,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.229251393282121e-06,
+      "loss": 0.2125,
+      "step": 13180
+    },
+    {
+      "epoch": 3.9549103458624053,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.2141888838680524e-06,
+      "loss": 0.0885,
+      "step": 13181
+    },
+    {
+      "epoch": 3.955210443394103,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.1991263744539843e-06,
+      "loss": 0.1203,
+      "step": 13182
+    },
+    {
+      "epoch": 3.955510540925801,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.1840638650399157e-06,
+      "loss": 0.0873,
+      "step": 13183
+    },
+    {
+      "epoch": 3.955810638457499,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.1690013556258475e-06,
+      "loss": 0.098,
+      "step": 13184
+    },
+    {
+      "epoch": 3.9561107359891965,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.153938846211779e-06,
+      "loss": 0.0873,
+      "step": 13185
+    },
+    {
+      "epoch": 3.956410833520894,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.1388763367977107e-06,
+      "loss": 0.0963,
+      "step": 13186
+    },
+    {
+      "epoch": 3.956710931052592,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.123813827383642e-06,
+      "loss": 0.1425,
+      "step": 13187
+    },
+    {
+      "epoch": 3.95701102858429,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.1087513179695736e-06,
+      "loss": 0.1024,
+      "step": 13188
+    },
+    {
+      "epoch": 3.9573111261159877,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.0936888085555054e-06,
+      "loss": 0.0956,
+      "step": 13189
+    },
+    {
+      "epoch": 3.9576112236476853,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.0786262991414372e-06,
+      "loss": 0.1116,
+      "step": 13190
+    },
+    {
+      "epoch": 3.9579113211793833,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.0635637897273686e-06,
+      "loss": 0.1011,
+      "step": 13191
+    },
+    {
+      "epoch": 3.9582114187110813,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.0485012803133e-06,
+      "loss": 0.1453,
+      "step": 13192
+    },
+    {
+      "epoch": 3.958511516242779,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.033438770899232e-06,
+      "loss": 0.096,
+      "step": 13193
+    },
+    {
+      "epoch": 3.9588116137744764,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.0183762614851637e-06,
+      "loss": 0.092,
+      "step": 13194
+    },
+    {
+      "epoch": 3.9591117113061745,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.003313752071095e-06,
+      "loss": 0.1047,
+      "step": 13195
+    },
+    {
+      "epoch": 3.9594118088378725,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.988251242657027e-06,
+      "loss": 0.1128,
+      "step": 13196
+    },
+    {
+      "epoch": 3.95971190636957,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9731887332429584e-06,
+      "loss": 0.1274,
+      "step": 13197
+    },
+    {
+      "epoch": 3.960012003901268,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.95812622382889e-06,
+      "loss": 0.0854,
+      "step": 13198
+    },
+    {
+      "epoch": 3.9603121014329656,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9430637144148216e-06,
+      "loss": 0.0789,
+      "step": 13199
+    },
+    {
+      "epoch": 3.9606121989646637,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9280012050007535e-06,
+      "loss": 0.1004,
+      "step": 13200
+    },
+    {
+      "epoch": 3.9609122964963612,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9129386955866845e-06,
+      "loss": 0.1079,
+      "step": 13201
+    },
+    {
+      "epoch": 3.9612123940280592,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8978761861726163e-06,
+      "loss": 0.0949,
+      "step": 13202
+    },
+    {
+      "epoch": 3.961512491559757,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8828136767585481e-06,
+      "loss": 0.0817,
+      "step": 13203
+    },
+    {
+      "epoch": 3.961812589091455,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.8677511673444798e-06,
+      "loss": 0.1212,
+      "step": 13204
+    },
+    {
+      "epoch": 3.9621126866231524,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8526886579304112e-06,
+      "loss": 0.0926,
+      "step": 13205
+    },
+    {
+      "epoch": 3.9624127841548504,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.837626148516343e-06,
+      "loss": 0.0929,
+      "step": 13206
+    },
+    {
+      "epoch": 3.962712881686548,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8225636391022746e-06,
+      "loss": 0.0924,
+      "step": 13207
+    },
+    {
+      "epoch": 3.963012979218246,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.807501129688206e-06,
+      "loss": 0.0943,
+      "step": 13208
+    },
+    {
+      "epoch": 3.9633130767499436,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7924386202741377e-06,
+      "loss": 0.0751,
+      "step": 13209
+    },
+    {
+      "epoch": 3.9636131742816416,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.7773761108600695e-06,
+      "loss": 0.1096,
+      "step": 13210
+    },
+    {
+      "epoch": 3.963913271813339,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.7623136014460011e-06,
+      "loss": 0.1132,
+      "step": 13211
+    },
+    {
+      "epoch": 3.964213369345037,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7472510920319325e-06,
+      "loss": 0.0989,
+      "step": 13212
+    },
+    {
+      "epoch": 3.964513466876735,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.7321885826178642e-06,
+      "loss": 0.1222,
+      "step": 13213
+    },
+    {
+      "epoch": 3.964813564408433,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.717126073203796e-06,
+      "loss": 0.105,
+      "step": 13214
+    },
+    {
+      "epoch": 3.965113661940131,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.7020635637897274e-06,
+      "loss": 0.1452,
+      "step": 13215
+    },
+    {
+      "epoch": 3.9654137594718284,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.687001054375659e-06,
+      "loss": 0.1091,
+      "step": 13216
+    },
+    {
+      "epoch": 3.965713857003526,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.6719385449615909e-06,
+      "loss": 0.1337,
+      "step": 13217
+    },
+    {
+      "epoch": 3.966013954535224,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.656876035547522e-06,
+      "loss": 0.1052,
+      "step": 13218
+    },
+    {
+      "epoch": 3.966314052066922,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.6418135261334539e-06,
+      "loss": 0.0823,
+      "step": 13219
+    },
+    {
+      "epoch": 3.9666141495986196,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.6267510167193855e-06,
+      "loss": 0.126,
+      "step": 13220
+    },
+    {
+      "epoch": 3.966914247130317,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6116885073053174e-06,
+      "loss": 0.1103,
+      "step": 13221
+    },
+    {
+      "epoch": 3.967214344662015,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5966259978912486e-06,
+      "loss": 0.1018,
+      "step": 13222
+    },
+    {
+      "epoch": 3.967514442193713,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5815634884771804e-06,
+      "loss": 0.0914,
+      "step": 13223
+    },
+    {
+      "epoch": 3.9678145397254108,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.566500979063112e-06,
+      "loss": 0.1089,
+      "step": 13224
+    },
+    {
+      "epoch": 3.9681146372571083,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.5514384696490436e-06,
+      "loss": 0.1181,
+      "step": 13225
+    },
+    {
+      "epoch": 3.9684147347888064,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.5363759602349753e-06,
+      "loss": 0.0897,
+      "step": 13226
+    },
+    {
+      "epoch": 3.9687148323205044,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5213134508209069e-06,
+      "loss": 0.0921,
+      "step": 13227
+    },
+    {
+      "epoch": 3.969014929852202,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.5062509414068385e-06,
+      "loss": 0.0933,
+      "step": 13228
+    },
+    {
+      "epoch": 3.9693150273838995,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.49118843199277e-06,
+      "loss": 0.0936,
+      "step": 13229
+    },
+    {
+      "epoch": 3.9696151249155975,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4761259225787018e-06,
+      "loss": 0.0773,
+      "step": 13230
+    },
+    {
+      "epoch": 3.9699152224472956,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.4610634131646332e-06,
+      "loss": 0.1033,
+      "step": 13231
+    },
+    {
+      "epoch": 3.970215319978993,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.446000903750565e-06,
+      "loss": 0.1045,
+      "step": 13232
+    },
+    {
+      "epoch": 3.9705154175106907,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4309383943364964e-06,
+      "loss": 0.0976,
+      "step": 13233
+    },
+    {
+      "epoch": 3.9708155150423887,
+      "grad_norm": 0.875,
+      "learning_rate": 1.4158758849224282e-06,
+      "loss": 0.0991,
+      "step": 13234
+    },
+    {
+      "epoch": 3.9711156125740867,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4008133755083599e-06,
+      "loss": 0.1008,
+      "step": 13235
+    },
+    {
+      "epoch": 3.9714157101057843,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.3857508660942913e-06,
+      "loss": 0.0965,
+      "step": 13236
+    },
+    {
+      "epoch": 3.9717158076374823,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3706883566802231e-06,
+      "loss": 0.0931,
+      "step": 13237
+    },
+    {
+      "epoch": 3.97201590516918,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3556258472661545e-06,
+      "loss": 0.09,
+      "step": 13238
+    },
+    {
+      "epoch": 3.972316002700878,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.3405633378520864e-06,
+      "loss": 0.1105,
+      "step": 13239
+    },
+    {
+      "epoch": 3.9726161002325755,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3255008284380178e-06,
+      "loss": 0.097,
+      "step": 13240
+    },
+    {
+      "epoch": 3.9729161977642735,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3104383190239494e-06,
+      "loss": 0.1188,
+      "step": 13241
+    },
+    {
+      "epoch": 3.973216295295971,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.295375809609881e-06,
+      "loss": 0.1347,
+      "step": 13242
+    },
+    {
+      "epoch": 3.973516392827669,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.2803133001958126e-06,
+      "loss": 0.0928,
+      "step": 13243
+    },
+    {
+      "epoch": 3.9738164903593667,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2652507907817443e-06,
+      "loss": 0.102,
+      "step": 13244
+    },
+    {
+      "epoch": 3.9741165878910647,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2501882813676759e-06,
+      "loss": 0.0943,
+      "step": 13245
+    },
+    {
+      "epoch": 3.9744166854227623,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2351257719536075e-06,
+      "loss": 0.1,
+      "step": 13246
+    },
+    {
+      "epoch": 3.9747167829544603,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2200632625395391e-06,
+      "loss": 0.1118,
+      "step": 13247
+    },
+    {
+      "epoch": 3.975016880486158,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2050007531254708e-06,
+      "loss": 0.0923,
+      "step": 13248
+    },
+    {
+      "epoch": 3.975316978017856,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.1899382437114024e-06,
+      "loss": 0.1195,
+      "step": 13249
+    },
+    {
+      "epoch": 3.9756170755495535,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.174875734297334e-06,
+      "loss": 0.0912,
+      "step": 13250
+    },
+    {
+      "epoch": 3.9759171730812515,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.1598132248832656e-06,
+      "loss": 0.1,
+      "step": 13251
+    },
+    {
+      "epoch": 3.976217270612949,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.1447507154691973e-06,
+      "loss": 0.1495,
+      "step": 13252
+    },
+    {
+      "epoch": 3.976517368144647,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.1296882060551289e-06,
+      "loss": 0.1041,
+      "step": 13253
+    },
+    {
+      "epoch": 3.976817465676345,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1146256966410605e-06,
+      "loss": 0.105,
+      "step": 13254
+    },
+    {
+      "epoch": 3.9771175632080427,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0995631872269921e-06,
+      "loss": 0.0916,
+      "step": 13255
+    },
+    {
+      "epoch": 3.9774176607397402,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.0845006778129237e-06,
+      "loss": 0.0989,
+      "step": 13256
+    },
+    {
+      "epoch": 3.9777177582714383,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.0694381683988554e-06,
+      "loss": 0.0822,
+      "step": 13257
+    },
+    {
+      "epoch": 3.9780178558031363,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0543756589847868e-06,
+      "loss": 0.0857,
+      "step": 13258
+    },
+    {
+      "epoch": 3.978317953334834,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0393131495707186e-06,
+      "loss": 0.1019,
+      "step": 13259
+    },
+    {
+      "epoch": 3.9786180508665314,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.02425064015665e-06,
+      "loss": 0.1043,
+      "step": 13260
+    },
+    {
+      "epoch": 3.9789181483982294,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.0091881307425819e-06,
+      "loss": 0.107,
+      "step": 13261
+    },
+    {
+      "epoch": 3.9792182459299275,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.941256213285135e-07,
+      "loss": 0.1018,
+      "step": 13262
+    },
+    {
+      "epoch": 3.979518343461625,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.79063111914445e-07,
+      "loss": 0.1061,
+      "step": 13263
+    },
+    {
+      "epoch": 3.9798184409933226,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.640006025003767e-07,
+      "loss": 0.0898,
+      "step": 13264
+    },
+    {
+      "epoch": 3.9801185385250206,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.489380930863081e-07,
+      "loss": 0.1085,
+      "step": 13265
+    },
+    {
+      "epoch": 3.9804186360567186,
+      "grad_norm": 0.9453125,
+      "learning_rate": 9.338755836722399e-07,
+      "loss": 0.129,
+      "step": 13266
+    },
+    {
+      "epoch": 3.980718733588416,
+      "grad_norm": 0.92578125,
+      "learning_rate": 9.188130742581715e-07,
+      "loss": 0.1144,
+      "step": 13267
+    },
+    {
+      "epoch": 3.981018831120114,
+      "grad_norm": 0.76171875,
+      "learning_rate": 9.03750564844103e-07,
+      "loss": 0.0826,
+      "step": 13268
+    },
+    {
+      "epoch": 3.981318928651812,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.886880554300347e-07,
+      "loss": 0.1251,
+      "step": 13269
+    },
+    {
+      "epoch": 3.98161902618351,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.736255460159663e-07,
+      "loss": 0.1277,
+      "step": 13270
+    },
+    {
+      "epoch": 3.9819191237152074,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.58563036601898e-07,
+      "loss": 0.1158,
+      "step": 13271
+    },
+    {
+      "epoch": 3.982219221246905,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.435005271878295e-07,
+      "loss": 0.0859,
+      "step": 13272
+    },
+    {
+      "epoch": 3.982519318778603,
+      "grad_norm": 0.85546875,
+      "learning_rate": 8.28438017773761e-07,
+      "loss": 0.1042,
+      "step": 13273
+    },
+    {
+      "epoch": 3.982819416310301,
+      "grad_norm": 0.93359375,
+      "learning_rate": 8.133755083596928e-07,
+      "loss": 0.1185,
+      "step": 13274
+    },
+    {
+      "epoch": 3.9831195138419986,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.983129989456243e-07,
+      "loss": 0.108,
+      "step": 13275
+    },
+    {
+      "epoch": 3.9834196113736966,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.83250489531556e-07,
+      "loss": 0.1046,
+      "step": 13276
+    },
+    {
+      "epoch": 3.983719708905394,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.681879801174876e-07,
+      "loss": 0.1054,
+      "step": 13277
+    },
+    {
+      "epoch": 3.984019806437092,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.531254707034193e-07,
+      "loss": 0.109,
+      "step": 13278
+    },
+    {
+      "epoch": 3.9843199039687898,
+      "grad_norm": 0.71875,
+      "learning_rate": 7.380629612893509e-07,
+      "loss": 0.103,
+      "step": 13279
+    },
+    {
+      "epoch": 3.984620001500488,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.230004518752825e-07,
+      "loss": 0.1329,
+      "step": 13280
+    },
+    {
+      "epoch": 3.9849200990321854,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.079379424612141e-07,
+      "loss": 0.105,
+      "step": 13281
+    },
+    {
+      "epoch": 3.9852201965638834,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.928754330471456e-07,
+      "loss": 0.0925,
+      "step": 13282
+    },
+    {
+      "epoch": 3.985520294095581,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.778129236330773e-07,
+      "loss": 0.0869,
+      "step": 13283
+    },
+    {
+      "epoch": 3.985820391627279,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.627504142190089e-07,
+      "loss": 0.0878,
+      "step": 13284
+    },
+    {
+      "epoch": 3.9861204891589765,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.476879048049405e-07,
+      "loss": 0.0935,
+      "step": 13285
+    },
+    {
+      "epoch": 3.9864205866906746,
+      "grad_norm": 0.9765625,
+      "learning_rate": 6.326253953908721e-07,
+      "loss": 0.1128,
+      "step": 13286
+    },
+    {
+      "epoch": 3.986720684222372,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6.175628859768038e-07,
+      "loss": 0.1032,
+      "step": 13287
+    },
+    {
+      "epoch": 3.98702078175407,
+      "grad_norm": 0.75390625,
+      "learning_rate": 6.025003765627354e-07,
+      "loss": 0.0919,
+      "step": 13288
+    },
+    {
+      "epoch": 3.9873208792857677,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.87437867148667e-07,
+      "loss": 0.1132,
+      "step": 13289
+    },
+    {
+      "epoch": 3.9876209768174657,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.723753577345986e-07,
+      "loss": 0.0922,
+      "step": 13290
+    },
+    {
+      "epoch": 3.9879210743491633,
+      "grad_norm": 0.796875,
+      "learning_rate": 5.573128483205303e-07,
+      "loss": 0.0975,
+      "step": 13291
+    },
+    {
+      "epoch": 3.9882211718808613,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.422503389064619e-07,
+      "loss": 0.0983,
+      "step": 13292
+    },
+    {
+      "epoch": 3.9885212694125594,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.271878294923934e-07,
+      "loss": 0.1034,
+      "step": 13293
+    },
+    {
+      "epoch": 3.988821366944257,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.12125320078325e-07,
+      "loss": 0.112,
+      "step": 13294
+    },
+    {
+      "epoch": 3.9891214644759545,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.970628106642567e-07,
+      "loss": 0.0852,
+      "step": 13295
+    },
+    {
+      "epoch": 3.9894215620076525,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.820003012501884e-07,
+      "loss": 0.1017,
+      "step": 13296
+    },
+    {
+      "epoch": 3.9897216595393505,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.6693779183611994e-07,
+      "loss": 0.099,
+      "step": 13297
+    },
+    {
+      "epoch": 3.990021757071048,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.518752824220515e-07,
+      "loss": 0.1105,
+      "step": 13298
+    },
+    {
+      "epoch": 3.9903218546027457,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.3681277300798313e-07,
+      "loss": 0.0961,
+      "step": 13299
+    },
+    {
+      "epoch": 3.9906219521344437,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.2175026359391476e-07,
+      "loss": 0.0825,
+      "step": 13300
+    },
+    {
+      "epoch": 3.9909220496661417,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.066877541798464e-07,
+      "loss": 0.0988,
+      "step": 13301
+    },
+    {
+      "epoch": 3.9912221471978393,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.91625244765778e-07,
+      "loss": 0.0876,
+      "step": 13302
+    },
+    {
+      "epoch": 3.991522244729537,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.765627353517096e-07,
+      "loss": 0.1054,
+      "step": 13303
+    },
+    {
+      "epoch": 3.991822342261235,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.6150022593764125e-07,
+      "loss": 0.0819,
+      "step": 13304
+    },
+    {
+      "epoch": 3.992122439792933,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.464377165235728e-07,
+      "loss": 0.0958,
+      "step": 13305
+    },
+    {
+      "epoch": 3.9924225373246305,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.3137520710950444e-07,
+      "loss": 0.116,
+      "step": 13306
+    },
+    {
+      "epoch": 3.992722634856328,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.1631269769543607e-07,
+      "loss": 0.097,
+      "step": 13307
+    },
+    {
+      "epoch": 3.993022732388026,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.012501882813677e-07,
+      "loss": 0.0877,
+      "step": 13308
+    },
+    {
+      "epoch": 3.993322829919724,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.861876788672993e-07,
+      "loss": 0.1023,
+      "step": 13309
+    },
+    {
+      "epoch": 3.9936229274514217,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.7112516945323094e-07,
+      "loss": 0.1083,
+      "step": 13310
+    },
+    {
+      "epoch": 3.9939230249831192,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.560626600391625e-07,
+      "loss": 0.1015,
+      "step": 13311
+    },
+    {
+      "epoch": 3.9942231225148173,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.410001506250942e-07,
+      "loss": 0.1145,
+      "step": 13312
+    },
+    {
+      "epoch": 3.9945232200465153,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.2593764121102575e-07,
+      "loss": 0.1139,
+      "step": 13313
+    },
+    {
+      "epoch": 3.994823317578213,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.1087513179695738e-07,
+      "loss": 0.1165,
+      "step": 13314
+    },
+    {
+      "epoch": 3.995123415109911,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.95812622382889e-07,
+      "loss": 0.1249,
+      "step": 13315
+    },
+    {
+      "epoch": 3.9954235126416084,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8075011296882062e-07,
+      "loss": 0.0842,
+      "step": 13316
+    },
+    {
+      "epoch": 3.9957236101733065,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.6568760355475222e-07,
+      "loss": 0.1042,
+      "step": 13317
+    },
+    {
+      "epoch": 3.996023707705004,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5062509414068385e-07,
+      "loss": 0.0782,
+      "step": 13318
+    },
+    {
+      "epoch": 3.996323805236702,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3556258472661547e-07,
+      "loss": 0.0877,
+      "step": 13319
+    },
+    {
+      "epoch": 3.9966239027683996,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.205000753125471e-07,
+      "loss": 0.0813,
+      "step": 13320
+    },
+    {
+      "epoch": 3.9969240003000976,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.0543756589847869e-07,
+      "loss": 0.1088,
+      "step": 13321
+    },
+    {
+      "epoch": 3.997224097831795,
+      "grad_norm": 0.9140625,
+      "learning_rate": 9.037505648441031e-08,
+      "loss": 0.1012,
+      "step": 13322
+    },
+    {
+      "epoch": 3.9975241953634932,
+      "grad_norm": 0.7578125,
+      "learning_rate": 7.531254707034192e-08,
+      "loss": 0.086,
+      "step": 13323
+    },
+    {
+      "epoch": 3.997824292895191,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.025003765627355e-08,
+      "loss": 0.0933,
+      "step": 13324
+    },
+    {
+      "epoch": 3.998124390426889,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.5187528242205156e-08,
+      "loss": 0.0959,
+      "step": 13325
+    },
+    {
+      "epoch": 3.9984244879585864,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.012501882813677e-08,
+      "loss": 0.1045,
+      "step": 13326
+    },
+    {
+      "epoch": 3.9987245854902844,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5062509414068386e-08,
+      "loss": 0.1021,
+      "step": 13327
+    },
+    {
+      "epoch": 3.999024683021982,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0,
+      "loss": 0.0964,
+      "step": 13328
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 13328,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.7840312548082586e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}