diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,129073 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.7495898583146907,
+  "eval_steps": 500,
+  "global_step": 18436,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.499163826921794,
+      "learning_rate": 1.9880715705765407e-09,
+      "loss": 0.7246,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.844882231508665,
+      "learning_rate": 3.976143141153081e-09,
+      "loss": 0.6927,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.869892765553143,
+      "learning_rate": 5.9642147117296215e-09,
+      "loss": 0.7552,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.385384547798115,
+      "learning_rate": 7.952286282306163e-09,
+      "loss": 0.7389,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.218001303560865,
+      "learning_rate": 9.940357852882704e-09,
+      "loss": 0.7572,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.35391251439786,
+      "learning_rate": 1.1928429423459243e-08,
+      "loss": 0.7227,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.310519781810637,
+      "learning_rate": 1.3916500994035786e-08,
+      "loss": 0.6953,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.6447604220992,
+      "learning_rate": 1.5904572564612325e-08,
+      "loss": 0.7064,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.480838387500022,
+      "learning_rate": 1.7892644135188866e-08,
+      "loss": 0.7038,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.876805425738677,
+      "learning_rate": 1.9880715705765407e-08,
+      "loss": 0.7513,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.660078722475282,
+      "learning_rate": 2.186878727634195e-08,
+      "loss": 0.7044,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.21963344106092,
+      "learning_rate": 2.3856858846918486e-08,
+      "loss": 0.7383,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.019857314615871,
+      "learning_rate": 2.584493041749503e-08,
+      "loss": 0.7292,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.496130485277952,
+      "learning_rate": 2.783300198807157e-08,
+      "loss": 0.7233,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.09147161519466,
+      "learning_rate": 2.9821073558648106e-08,
+      "loss": 0.7233,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.387203937810977,
+      "learning_rate": 3.180914512922465e-08,
+      "loss": 0.737,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.288880135026632,
+      "learning_rate": 3.3797216699801195e-08,
+      "loss": 0.6986,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.269914776207411,
+      "learning_rate": 3.578528827037773e-08,
+      "loss": 0.7129,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.059142453553292,
+      "learning_rate": 3.777335984095427e-08,
+      "loss": 0.7259,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.776974026045124,
+      "learning_rate": 3.9761431411530815e-08,
+      "loss": 0.7305,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.135453576211118,
+      "learning_rate": 4.174950298210735e-08,
+      "loss": 0.6771,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.404769665161314,
+      "learning_rate": 4.37375745526839e-08,
+      "loss": 0.722,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.725641784414934,
+      "learning_rate": 4.5725646123260435e-08,
+      "loss": 0.6986,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.805350818812755,
+      "learning_rate": 4.771371769383697e-08,
+      "loss": 0.722,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.579340790709406,
+      "learning_rate": 4.970178926441352e-08,
+      "loss": 0.7578,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.353250087231306,
+      "learning_rate": 5.168986083499006e-08,
+      "loss": 0.7467,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.737559850019615,
+      "learning_rate": 5.36779324055666e-08,
+      "loss": 0.6979,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.51774692766562,
+      "learning_rate": 5.566600397614314e-08,
+      "loss": 0.7663,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.340207312680363,
+      "learning_rate": 5.765407554671968e-08,
+      "loss": 0.7441,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.611359067464182,
+      "learning_rate": 5.964214711729621e-08,
+      "loss": 0.7305,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.911174163776828,
+      "learning_rate": 6.163021868787277e-08,
+      "loss": 0.7201,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.443041480710598,
+      "learning_rate": 6.36182902584493e-08,
+      "loss": 0.7331,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.750800834579348,
+      "learning_rate": 6.560636182902585e-08,
+      "loss": 0.7454,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.389633183706188,
+      "learning_rate": 6.759443339960239e-08,
+      "loss": 0.7155,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.614846715291455,
+      "learning_rate": 6.958250497017892e-08,
+      "loss": 0.694,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.303020098843145,
+      "learning_rate": 7.157057654075547e-08,
+      "loss": 0.7129,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.976729238305676,
+      "learning_rate": 7.3558648111332e-08,
+      "loss": 0.7142,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.281141553521149,
+      "learning_rate": 7.554671968190854e-08,
+      "loss": 0.6921,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.97590389010619,
+      "learning_rate": 7.753479125248509e-08,
+      "loss": 0.696,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.404236684870334,
+      "learning_rate": 7.952286282306163e-08,
+      "loss": 0.694,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.67511430769177,
+      "learning_rate": 8.151093439363816e-08,
+      "loss": 0.7435,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.447412412274621,
+      "learning_rate": 8.34990059642147e-08,
+      "loss": 0.7064,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.79141707379374,
+      "learning_rate": 8.548707753479125e-08,
+      "loss": 0.7168,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.826111898578436,
+      "learning_rate": 8.74751491053678e-08,
+      "loss": 0.7292,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.118190427804656,
+      "learning_rate": 8.946322067594434e-08,
+      "loss": 0.7122,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.489061860847059,
+      "learning_rate": 9.145129224652087e-08,
+      "loss": 0.6999,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.417653960941234,
+      "learning_rate": 9.343936381709741e-08,
+      "loss": 0.7064,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.234906080918769,
+      "learning_rate": 9.542743538767394e-08,
+      "loss": 0.7318,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.724395544227617,
+      "learning_rate": 9.741550695825049e-08,
+      "loss": 0.709,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.291154194949023,
+      "learning_rate": 9.940357852882703e-08,
+      "loss": 0.7324,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.472890011552881,
+      "learning_rate": 1.0139165009940358e-07,
+      "loss": 0.7201,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.989606563656245,
+      "learning_rate": 1.0337972166998012e-07,
+      "loss": 0.7246,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.761329531813221,
+      "learning_rate": 1.0536779324055665e-07,
+      "loss": 0.7031,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.826023943398433,
+      "learning_rate": 1.073558648111332e-07,
+      "loss": 0.6934,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.198638910120003,
+      "learning_rate": 1.0934393638170973e-07,
+      "loss": 0.7253,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.560391281868366,
+      "learning_rate": 1.1133200795228629e-07,
+      "loss": 0.7103,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.02576284572131,
+      "learning_rate": 1.1332007952286282e-07,
+      "loss": 0.6901,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.007662960941557,
+      "learning_rate": 1.1530815109343936e-07,
+      "loss": 0.7057,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.265760313566703,
+      "learning_rate": 1.1729622266401589e-07,
+      "loss": 0.6803,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.566401304297036,
+      "learning_rate": 1.1928429423459242e-07,
+      "loss": 0.7064,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9881631788582945,
+      "learning_rate": 1.2127236580516898e-07,
+      "loss": 0.6986,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.540851415471602,
+      "learning_rate": 1.2326043737574554e-07,
+      "loss": 0.7266,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.859211696523065,
+      "learning_rate": 1.2524850894632204e-07,
+      "loss": 0.7012,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.238447945620525,
+      "learning_rate": 1.272365805168986e-07,
+      "loss": 0.6934,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.331280405705904,
+      "learning_rate": 1.2922465208747516e-07,
+      "loss": 0.6999,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.806504875394937,
+      "learning_rate": 1.312127236580517e-07,
+      "loss": 0.7018,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.690019869078491,
+      "learning_rate": 1.3320079522862822e-07,
+      "loss": 0.7109,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.859994989916984,
+      "learning_rate": 1.3518886679920478e-07,
+      "loss": 0.7214,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.361773269896933,
+      "learning_rate": 1.371769383697813e-07,
+      "loss": 0.7057,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.291579054779309,
+      "learning_rate": 1.3916500994035784e-07,
+      "loss": 0.7018,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.534546399712637,
+      "learning_rate": 1.4115308151093437e-07,
+      "loss": 0.7116,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.802269215676397,
+      "learning_rate": 1.4314115308151093e-07,
+      "loss": 0.6829,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.129126282797703,
+      "learning_rate": 1.4512922465208746e-07,
+      "loss": 0.6992,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.058949419865149,
+      "learning_rate": 1.47117296222664e-07,
+      "loss": 0.7201,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.45112455235193,
+      "learning_rate": 1.4910536779324055e-07,
+      "loss": 0.7214,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.567052664136712,
+      "learning_rate": 1.5109343936381708e-07,
+      "loss": 0.7025,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.281020543727984,
+      "learning_rate": 1.5308151093439364e-07,
+      "loss": 0.6914,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.666454420555736,
+      "learning_rate": 1.5506958250497017e-07,
+      "loss": 0.6979,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.382466428751778,
+      "learning_rate": 1.5705765407554673e-07,
+      "loss": 0.7038,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.091048282639214,
+      "learning_rate": 1.5904572564612326e-07,
+      "loss": 0.7135,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.150491817630403,
+      "learning_rate": 1.610337972166998e-07,
+      "loss": 0.7096,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.64761922867085,
+      "learning_rate": 1.6302186878727632e-07,
+      "loss": 0.6934,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.816630042662357,
+      "learning_rate": 1.6500994035785288e-07,
+      "loss": 0.7051,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.643628404497148,
+      "learning_rate": 1.669980119284294e-07,
+      "loss": 0.6901,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.097322687284572,
+      "learning_rate": 1.6898608349900594e-07,
+      "loss": 0.6901,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.096391842239505,
+      "learning_rate": 1.709741550695825e-07,
+      "loss": 0.7012,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.35751914390632,
+      "learning_rate": 1.7296222664015903e-07,
+      "loss": 0.7038,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.967649705514523,
+      "learning_rate": 1.749502982107356e-07,
+      "loss": 0.7174,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.215951849720813,
+      "learning_rate": 1.7693836978131212e-07,
+      "loss": 0.7083,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.627669209565727,
+      "learning_rate": 1.7892644135188868e-07,
+      "loss": 0.6895,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.945488033487356,
+      "learning_rate": 1.809145129224652e-07,
+      "loss": 0.6882,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.859445740889933,
+      "learning_rate": 1.8290258449304174e-07,
+      "loss": 0.6986,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.626364205274266,
+      "learning_rate": 1.848906560636183e-07,
+      "loss": 0.6999,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.642856374584046,
+      "learning_rate": 1.8687872763419483e-07,
+      "loss": 0.6908,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.940388225305853,
+      "learning_rate": 1.8886679920477136e-07,
+      "loss": 0.7155,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 14.77869579158993,
+      "learning_rate": 1.908548707753479e-07,
+      "loss": 0.6986,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.495662336550748,
+      "learning_rate": 1.9284294234592445e-07,
+      "loss": 0.7109,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.969066596698855,
+      "learning_rate": 1.9483101391650098e-07,
+      "loss": 0.7103,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.151406641235983,
+      "learning_rate": 1.968190854870775e-07,
+      "loss": 0.6966,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.054713987899888,
+      "learning_rate": 1.9880715705765407e-07,
+      "loss": 0.7161,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.431771945316525,
+      "learning_rate": 2.0079522862823062e-07,
+      "loss": 0.7077,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.932734251906004,
+      "learning_rate": 2.0278330019880716e-07,
+      "loss": 0.7031,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.314518051799478,
+      "learning_rate": 2.047713717693837e-07,
+      "loss": 0.6966,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.439046498761027,
+      "learning_rate": 2.0675944333996024e-07,
+      "loss": 0.7142,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.900652834713968,
+      "learning_rate": 2.0874751491053678e-07,
+      "loss": 0.696,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.161657127109603,
+      "learning_rate": 2.107355864811133e-07,
+      "loss": 0.6999,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.950955066276629,
+      "learning_rate": 2.1272365805168984e-07,
+      "loss": 0.6953,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.067386347840333,
+      "learning_rate": 2.147117296222664e-07,
+      "loss": 0.6725,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.543152464772712,
+      "learning_rate": 2.1669980119284293e-07,
+      "loss": 0.6849,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.604300046963612,
+      "learning_rate": 2.1868787276341946e-07,
+      "loss": 0.6973,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.78293640711407,
+      "learning_rate": 2.20675944333996e-07,
+      "loss": 0.6992,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.804083128828628,
+      "learning_rate": 2.2266401590457257e-07,
+      "loss": 0.7077,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.354349578729526,
+      "learning_rate": 2.246520874751491e-07,
+      "loss": 0.6908,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.33395134404084,
+      "learning_rate": 2.2664015904572564e-07,
+      "loss": 0.7103,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.319865071781924,
+      "learning_rate": 2.286282306163022e-07,
+      "loss": 0.709,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.226714659637416,
+      "learning_rate": 2.3061630218687872e-07,
+      "loss": 0.7025,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.380441824199682,
+      "learning_rate": 2.3260437375745526e-07,
+      "loss": 0.6888,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.710454499920294,
+      "learning_rate": 2.3459244532803179e-07,
+      "loss": 0.6973,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.613694452575448,
+      "learning_rate": 2.3658051689860834e-07,
+      "loss": 0.6895,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.234806556900114,
+      "learning_rate": 2.3856858846918485e-07,
+      "loss": 0.7031,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.809416112059246,
+      "learning_rate": 2.405566600397614e-07,
+      "loss": 0.6973,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.702103063908027,
+      "learning_rate": 2.4254473161033796e-07,
+      "loss": 0.6934,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.56965141757914,
+      "learning_rate": 2.445328031809145e-07,
+      "loss": 0.7077,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.110497199174853,
+      "learning_rate": 2.465208747514911e-07,
+      "loss": 0.6934,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.988481809796136,
+      "learning_rate": 2.485089463220676e-07,
+      "loss": 0.7012,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.2477238778832875,
+      "learning_rate": 2.504970178926441e-07,
+      "loss": 0.6927,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.723928197530425,
+      "learning_rate": 2.5248508946322065e-07,
+      "loss": 0.7005,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.613792941341233,
+      "learning_rate": 2.544731610337972e-07,
+      "loss": 0.6895,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.624417831100566,
+      "learning_rate": 2.5646123260437376e-07,
+      "loss": 0.694,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.313441311786733,
+      "learning_rate": 2.584493041749503e-07,
+      "loss": 0.7005,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.1267435033805935,
+      "learning_rate": 2.604373757455268e-07,
+      "loss": 0.6829,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.403280107465562,
+      "learning_rate": 2.624254473161034e-07,
+      "loss": 0.6947,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.340734408965702,
+      "learning_rate": 2.644135188866799e-07,
+      "loss": 0.7096,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.588529061202205,
+      "learning_rate": 2.6640159045725644e-07,
+      "loss": 0.694,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.383856981558846,
+      "learning_rate": 2.68389662027833e-07,
+      "loss": 0.6836,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.93930587340735,
+      "learning_rate": 2.7037773359840956e-07,
+      "loss": 0.6979,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.353156560726323,
+      "learning_rate": 2.7236580516898606e-07,
+      "loss": 0.6882,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.407364884013392,
+      "learning_rate": 2.743538767395626e-07,
+      "loss": 0.6836,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.067924299107743,
+      "learning_rate": 2.763419483101391e-07,
+      "loss": 0.7064,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.248695636900457,
+      "learning_rate": 2.783300198807157e-07,
+      "loss": 0.6908,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.017377545311174,
+      "learning_rate": 2.8031809145129224e-07,
+      "loss": 0.7122,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.357652234362792,
+      "learning_rate": 2.8230616302186875e-07,
+      "loss": 0.6947,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.57942372839956,
+      "learning_rate": 2.8429423459244536e-07,
+      "loss": 0.7083,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.922456186225475,
+      "learning_rate": 2.8628230616302186e-07,
+      "loss": 0.7194,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.473115631800136,
+      "learning_rate": 2.882703777335984e-07,
+      "loss": 0.7148,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.331664942924284,
+      "learning_rate": 2.902584493041749e-07,
+      "loss": 0.7109,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.673539635049096,
+      "learning_rate": 2.922465208747515e-07,
+      "loss": 0.7038,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.04107337352101,
+      "learning_rate": 2.94234592445328e-07,
+      "loss": 0.6901,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.255638228404129,
+      "learning_rate": 2.9622266401590454e-07,
+      "loss": 0.6973,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.758485345818503,
+      "learning_rate": 2.982107355864811e-07,
+      "loss": 0.7148,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.832774466364654,
+      "learning_rate": 3.0019880715705766e-07,
+      "loss": 0.7018,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.936029466771296,
+      "learning_rate": 3.0218687872763416e-07,
+      "loss": 0.7051,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.11735546998571,
+      "learning_rate": 3.041749502982107e-07,
+      "loss": 0.7129,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.934445035623654,
+      "learning_rate": 3.061630218687873e-07,
+      "loss": 0.6927,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.860628341537415,
+      "learning_rate": 3.081510934393638e-07,
+      "loss": 0.709,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.776979428644395,
+      "learning_rate": 3.1013916500994034e-07,
+      "loss": 0.6973,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.060839235696866,
+      "learning_rate": 3.121272365805169e-07,
+      "loss": 0.6875,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.016636156484555,
+      "learning_rate": 3.1411530815109346e-07,
+      "loss": 0.7005,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.664366811602976,
+      "learning_rate": 3.1610337972166996e-07,
+      "loss": 0.6953,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.97073369438387,
+      "learning_rate": 3.180914512922465e-07,
+      "loss": 0.6986,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.626382143390192,
+      "learning_rate": 3.20079522862823e-07,
+      "loss": 0.7038,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.069823255980422,
+      "learning_rate": 3.220675944333996e-07,
+      "loss": 0.6973,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.293485849087814,
+      "learning_rate": 3.240556660039761e-07,
+      "loss": 0.7051,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.194341501675366,
+      "learning_rate": 3.2604373757455264e-07,
+      "loss": 0.709,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.760642254352529,
+      "learning_rate": 3.2803180914512925e-07,
+      "loss": 0.6908,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.718925011211521,
+      "learning_rate": 3.3001988071570576e-07,
+      "loss": 0.6888,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.51965340046931,
+      "learning_rate": 3.320079522862823e-07,
+      "loss": 0.709,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.370216470811773,
+      "learning_rate": 3.339960238568588e-07,
+      "loss": 0.6999,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.641459829903065,
+      "learning_rate": 3.359840954274354e-07,
+      "loss": 0.7005,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.227145148657694,
+      "learning_rate": 3.379721669980119e-07,
+      "loss": 0.6986,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.030945147939798,
+      "learning_rate": 3.3996023856858844e-07,
+      "loss": 0.7083,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.754347731302461,
+      "learning_rate": 3.41948310139165e-07,
+      "loss": 0.6973,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.306275307575275,
+      "learning_rate": 3.4393638170974155e-07,
+      "loss": 0.7148,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.938887954973122,
+      "learning_rate": 3.4592445328031806e-07,
+      "loss": 0.694,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.033173801266212,
+      "learning_rate": 3.479125248508946e-07,
+      "loss": 0.6921,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.15575759084492,
+      "learning_rate": 3.499005964214712e-07,
+      "loss": 0.7038,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.081121940232887,
+      "learning_rate": 3.518886679920477e-07,
+      "loss": 0.7012,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.144905665693727,
+      "learning_rate": 3.5387673956262424e-07,
+      "loss": 0.6895,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.840880249842803,
+      "learning_rate": 3.558648111332008e-07,
+      "loss": 0.7038,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.656604381349382,
+      "learning_rate": 3.5785288270377735e-07,
+      "loss": 0.7057,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.173394848571524,
+      "learning_rate": 3.5984095427435386e-07,
+      "loss": 0.6979,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.665419354119585,
+      "learning_rate": 3.618290258449304e-07,
+      "loss": 0.6953,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.25288637593796,
+      "learning_rate": 3.638170974155069e-07,
+      "loss": 0.6855,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 19.802545830999897,
+      "learning_rate": 3.658051689860835e-07,
+      "loss": 0.7044,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.949341678898739,
+      "learning_rate": 3.6779324055666e-07,
+      "loss": 0.6927,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.233998150714905,
+      "learning_rate": 3.697813121272366e-07,
+      "loss": 0.6979,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.981150297023035,
+      "learning_rate": 3.717693836978131e-07,
+      "loss": 0.6927,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.258410302466865,
+      "learning_rate": 3.7375745526838965e-07,
+      "loss": 0.6986,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.97096519852186,
+      "learning_rate": 3.757455268389662e-07,
+      "loss": 0.6986,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.662015794529262,
+      "learning_rate": 3.777335984095427e-07,
+      "loss": 0.6966,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.101682221907611,
+      "learning_rate": 3.797216699801193e-07,
+      "loss": 0.7005,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.063709314217139,
+      "learning_rate": 3.817097415506958e-07,
+      "loss": 0.7018,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.267474946684223,
+      "learning_rate": 3.8369781312127234e-07,
+      "loss": 0.6973,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.724586513910879,
+      "learning_rate": 3.856858846918489e-07,
+      "loss": 0.7064,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.577509707164589,
+      "learning_rate": 3.8767395626242545e-07,
+      "loss": 0.7005,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.262078697888994,
+      "learning_rate": 3.8966202783300196e-07,
+      "loss": 0.6979,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.432761043712745,
+      "learning_rate": 3.916500994035785e-07,
+      "loss": 0.6921,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.170122449542886,
+      "learning_rate": 3.93638170974155e-07,
+      "loss": 0.6901,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.92221347737594,
+      "learning_rate": 3.956262425447316e-07,
+      "loss": 0.696,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.47046784375388,
+      "learning_rate": 3.9761431411530813e-07,
+      "loss": 0.6888,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.16719653926047,
+      "learning_rate": 3.996023856858847e-07,
+      "loss": 0.6784,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.225514652092729,
+      "learning_rate": 4.0159045725646125e-07,
+      "loss": 0.6901,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.837374911018087,
+      "learning_rate": 4.0357852882703775e-07,
+      "loss": 0.6751,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.297190274257355,
+      "learning_rate": 4.055666003976143e-07,
+      "loss": 0.6849,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.255966378500446,
+      "learning_rate": 4.075546719681908e-07,
+      "loss": 0.7025,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.611746354866194,
+      "learning_rate": 4.095427435387674e-07,
+      "loss": 0.6953,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.829940443601568,
+      "learning_rate": 4.115308151093439e-07,
+      "loss": 0.696,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.840135382636408,
+      "learning_rate": 4.135188866799205e-07,
+      "loss": 0.707,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.499216410916286,
+      "learning_rate": 4.15506958250497e-07,
+      "loss": 0.6986,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.801073041781994,
+      "learning_rate": 4.1749502982107355e-07,
+      "loss": 0.7103,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.988704484186014,
+      "learning_rate": 4.194831013916501e-07,
+      "loss": 0.707,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.907977393176701,
+      "learning_rate": 4.214711729622266e-07,
+      "loss": 0.6829,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.375759500490327,
+      "learning_rate": 4.2345924453280317e-07,
+      "loss": 0.7031,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.946511169671699,
+      "learning_rate": 4.254473161033797e-07,
+      "loss": 0.7044,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.998716399542566,
+      "learning_rate": 4.274353876739563e-07,
+      "loss": 0.6947,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.959329216082167,
+      "learning_rate": 4.294234592445328e-07,
+      "loss": 0.6979,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.042285846305406,
+      "learning_rate": 4.3141153081510935e-07,
+      "loss": 0.7161,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.662373378121598,
+      "learning_rate": 4.3339960238568585e-07,
+      "loss": 0.6973,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.55518016185341,
+      "learning_rate": 4.353876739562624e-07,
+      "loss": 0.6875,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.234713264660966,
+      "learning_rate": 4.373757455268389e-07,
+      "loss": 0.7142,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 16.953434684821925,
+      "learning_rate": 4.3936381709741547e-07,
+      "loss": 0.7122,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.956100526957465,
+      "learning_rate": 4.41351888667992e-07,
+      "loss": 0.7018,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.7557209006064145,
+      "learning_rate": 4.433399602385686e-07,
+      "loss": 0.6953,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.816441376599446,
+      "learning_rate": 4.4532803180914515e-07,
+      "loss": 0.6992,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.419234313521548,
+      "learning_rate": 4.4731610337972165e-07,
+      "loss": 0.7031,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.944268927972523,
+      "learning_rate": 4.493041749502982e-07,
+      "loss": 0.6914,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.803096837220538,
+      "learning_rate": 4.512922465208747e-07,
+      "loss": 0.6999,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.815172575766862,
+      "learning_rate": 4.5328031809145127e-07,
+      "loss": 0.7038,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.891398203553509,
+      "learning_rate": 4.552683896620278e-07,
+      "loss": 0.696,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.826592139031577,
+      "learning_rate": 4.572564612326044e-07,
+      "loss": 0.6966,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.151296064025711,
+      "learning_rate": 4.592445328031809e-07,
+      "loss": 0.6992,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.599015253621187,
+      "learning_rate": 4.6123260437375745e-07,
+      "loss": 0.6979,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.293974973231238,
+      "learning_rate": 4.6322067594433395e-07,
+      "loss": 0.6921,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.225342863274202,
+      "learning_rate": 4.652087475149105e-07,
+      "loss": 0.6914,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.056641346363579,
+      "learning_rate": 4.6719681908548707e-07,
+      "loss": 0.7012,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.536569729443992,
+      "learning_rate": 4.6918489065606357e-07,
+      "loss": 0.6999,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.35370236589335,
+      "learning_rate": 4.711729622266402e-07,
+      "loss": 0.6979,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.52220772969608,
+      "learning_rate": 4.731610337972167e-07,
+      "loss": 0.6914,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.076343622547835,
+      "learning_rate": 4.7514910536779325e-07,
+      "loss": 0.6914,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.554060849744815,
+      "learning_rate": 4.771371769383697e-07,
+      "loss": 0.6979,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.480350807203205,
+      "learning_rate": 4.791252485089463e-07,
+      "loss": 0.6927,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.342242299633007,
+      "learning_rate": 4.811133200795228e-07,
+      "loss": 0.7012,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.664015945809206,
+      "learning_rate": 4.831013916500994e-07,
+      "loss": 0.6882,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.958828991977214,
+      "learning_rate": 4.850894632206759e-07,
+      "loss": 0.7012,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.77870046224619,
+      "learning_rate": 4.870775347912524e-07,
+      "loss": 0.6947,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 18.493526176559502,
+      "learning_rate": 4.89065606361829e-07,
+      "loss": 0.7051,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.502814758093153,
+      "learning_rate": 4.910536779324055e-07,
+      "loss": 0.6777,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.431062177489339,
+      "learning_rate": 4.930417495029822e-07,
+      "loss": 0.7214,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.080436747086564,
+      "learning_rate": 4.950298210735587e-07,
+      "loss": 0.696,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.34637214528923,
+      "learning_rate": 4.970178926441352e-07,
+      "loss": 0.6849,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.074846397954222,
+      "learning_rate": 4.990059642147117e-07,
+      "loss": 0.7025,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.661703632519153,
+      "learning_rate": 5.009940357852882e-07,
+      "loss": 0.6908,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.772694233410913,
+      "learning_rate": 5.029821073558648e-07,
+      "loss": 0.707,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.5479659506665975,
+      "learning_rate": 5.049701789264413e-07,
+      "loss": 0.6855,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.003122075923246,
+      "learning_rate": 5.069582504970179e-07,
+      "loss": 0.694,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.28882510341652,
+      "learning_rate": 5.089463220675944e-07,
+      "loss": 0.6901,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.135345888142215,
+      "learning_rate": 5.109343936381709e-07,
+      "loss": 0.7025,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.775941597901639,
+      "learning_rate": 5.129224652087475e-07,
+      "loss": 0.696,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.338576180857166,
+      "learning_rate": 5.14910536779324e-07,
+      "loss": 0.7038,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.42640118414473,
+      "learning_rate": 5.168986083499006e-07,
+      "loss": 0.7018,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.873366258012261,
+      "learning_rate": 5.18886679920477e-07,
+      "loss": 0.6979,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.039274197150805,
+      "learning_rate": 5.208747514910536e-07,
+      "loss": 0.7025,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.518861953464397,
+      "learning_rate": 5.228628230616303e-07,
+      "loss": 0.6855,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.259611086979236,
+      "learning_rate": 5.248508946322068e-07,
+      "loss": 0.7096,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.92115540095605,
+      "learning_rate": 5.268389662027833e-07,
+      "loss": 0.6803,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.794876894489468,
+      "learning_rate": 5.288270377733598e-07,
+      "loss": 0.6947,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.474805137418248,
+      "learning_rate": 5.308151093439364e-07,
+      "loss": 0.6914,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.691991274029654,
+      "learning_rate": 5.328031809145129e-07,
+      "loss": 0.7018,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.681309089813006,
+      "learning_rate": 5.347912524850894e-07,
+      "loss": 0.6914,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.488518547907686,
+      "learning_rate": 5.36779324055666e-07,
+      "loss": 0.7116,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.375410364517245,
+      "learning_rate": 5.387673956262425e-07,
+      "loss": 0.6914,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.790093212642846,
+      "learning_rate": 5.407554671968191e-07,
+      "loss": 0.7005,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.5063460069214,
+      "learning_rate": 5.427435387673956e-07,
+      "loss": 0.681,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.451475537748225,
+      "learning_rate": 5.447316103379721e-07,
+      "loss": 0.7025,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.859379857762386,
+      "learning_rate": 5.467196819085486e-07,
+      "loss": 0.7077,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.711859151555616,
+      "learning_rate": 5.487077534791252e-07,
+      "loss": 0.694,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.964127175962684,
+      "learning_rate": 5.506958250497019e-07,
+      "loss": 0.7064,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.224217021485792,
+      "learning_rate": 5.526838966202783e-07,
+      "loss": 0.6934,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.560728741945537,
+      "learning_rate": 5.546719681908549e-07,
+      "loss": 0.6836,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.353656930563279,
+      "learning_rate": 5.566600397614314e-07,
+      "loss": 0.6947,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.661472247020613,
+      "learning_rate": 5.58648111332008e-07,
+      "loss": 0.7083,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.149483876298925,
+      "learning_rate": 5.606361829025845e-07,
+      "loss": 0.6921,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.139296603661725,
+      "learning_rate": 5.62624254473161e-07,
+      "loss": 0.6842,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.739375060295059,
+      "learning_rate": 5.646123260437375e-07,
+      "loss": 0.6901,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.895838255805702,
+      "learning_rate": 5.666003976143141e-07,
+      "loss": 0.6934,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.899371361015414,
+      "learning_rate": 5.685884691848907e-07,
+      "loss": 0.6921,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.847583962669692,
+      "learning_rate": 5.705765407554671e-07,
+      "loss": 0.6829,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.451751245363089,
+      "learning_rate": 5.725646123260437e-07,
+      "loss": 0.6914,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.221409177869833,
+      "learning_rate": 5.745526838966202e-07,
+      "loss": 0.7064,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.123138139895522,
+      "learning_rate": 5.765407554671968e-07,
+      "loss": 0.7077,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.3840821201685936,
+      "learning_rate": 5.785288270377732e-07,
+      "loss": 0.7012,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.41883436126114,
+      "learning_rate": 5.805168986083498e-07,
+      "loss": 0.7207,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.93971694207146,
+      "learning_rate": 5.825049701789265e-07,
+      "loss": 0.7051,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.307009090413998,
+      "learning_rate": 5.84493041749503e-07,
+      "loss": 0.7005,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.045378058309787,
+      "learning_rate": 5.864811133200796e-07,
+      "loss": 0.7012,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.290509368371086,
+      "learning_rate": 5.88469184890656e-07,
+      "loss": 0.7005,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.937779129744793,
+      "learning_rate": 5.904572564612326e-07,
+      "loss": 0.6927,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.502500035460147,
+      "learning_rate": 5.924453280318091e-07,
+      "loss": 0.6914,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.403420813579434,
+      "learning_rate": 5.944333996023857e-07,
+      "loss": 0.7018,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.23040202536099,
+      "learning_rate": 5.964214711729622e-07,
+      "loss": 0.6927,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.527493376871403,
+      "learning_rate": 5.984095427435387e-07,
+      "loss": 0.6836,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.040441324881188,
+      "learning_rate": 6.003976143141153e-07,
+      "loss": 0.6914,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.559476574674447,
+      "learning_rate": 6.023856858846918e-07,
+      "loss": 0.6888,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.497114163905213,
+      "learning_rate": 6.043737574552683e-07,
+      "loss": 0.6953,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.788947607189833,
+      "learning_rate": 6.063618290258448e-07,
+      "loss": 0.6803,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.970019694739763,
+      "learning_rate": 6.083499005964214e-07,
+      "loss": 0.6797,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.031853208461742,
+      "learning_rate": 6.10337972166998e-07,
+      "loss": 0.7012,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.668592005887024,
+      "learning_rate": 6.123260437375746e-07,
+      "loss": 0.7194,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.242629683158277,
+      "learning_rate": 6.143141153081511e-07,
+      "loss": 0.681,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.958835029968174,
+      "learning_rate": 6.163021868787276e-07,
+      "loss": 0.6999,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.15745856648311,
+      "learning_rate": 6.182902584493042e-07,
+      "loss": 0.7012,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.7483854137757,
+      "learning_rate": 6.202783300198807e-07,
+      "loss": 0.7064,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.3213437620023,
+      "learning_rate": 6.222664015904572e-07,
+      "loss": 0.6797,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.89129801222546,
+      "learning_rate": 6.242544731610338e-07,
+      "loss": 0.7161,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.487854901671309,
+      "learning_rate": 6.262425447316103e-07,
+      "loss": 0.681,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.631634579184208,
+      "learning_rate": 6.282306163021869e-07,
+      "loss": 0.6829,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.479589784559503,
+      "learning_rate": 6.302186878727634e-07,
+      "loss": 0.7103,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.88680818714399,
+      "learning_rate": 6.322067594433399e-07,
+      "loss": 0.7331,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.16012988872503,
+      "learning_rate": 6.341948310139164e-07,
+      "loss": 0.6986,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.506304447319892,
+      "learning_rate": 6.36182902584493e-07,
+      "loss": 0.6953,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.117948308429597,
+      "learning_rate": 6.381709741550696e-07,
+      "loss": 0.6882,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.869821176900725,
+      "learning_rate": 6.40159045725646e-07,
+      "loss": 0.6862,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.270755309687003,
+      "learning_rate": 6.421471172962227e-07,
+      "loss": 0.6973,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 23.256454017634205,
+      "learning_rate": 6.441351888667992e-07,
+      "loss": 0.7109,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.2785209851603,
+      "learning_rate": 6.461232604373758e-07,
+      "loss": 0.6882,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.652188222633303,
+      "learning_rate": 6.481113320079522e-07,
+      "loss": 0.7168,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.2935449198507,
+      "learning_rate": 6.500994035785288e-07,
+      "loss": 0.6999,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.584956941540586,
+      "learning_rate": 6.520874751491053e-07,
+      "loss": 0.7031,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.32122696980023,
+      "learning_rate": 6.540755467196819e-07,
+      "loss": 0.7109,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.228668761961366,
+      "learning_rate": 6.560636182902585e-07,
+      "loss": 0.6855,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.269165490326464,
+      "learning_rate": 6.580516898608349e-07,
+      "loss": 0.6979,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.469514141986535,
+      "learning_rate": 6.600397614314115e-07,
+      "loss": 0.6908,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.648791090569956,
+      "learning_rate": 6.62027833001988e-07,
+      "loss": 0.7129,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.435671443202395,
+      "learning_rate": 6.640159045725646e-07,
+      "loss": 0.7077,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.02730451898599,
+      "learning_rate": 6.66003976143141e-07,
+      "loss": 0.7057,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.411932022858572,
+      "learning_rate": 6.679920477137176e-07,
+      "loss": 0.6966,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.456042587914371,
+      "learning_rate": 6.699801192842943e-07,
+      "loss": 0.7012,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.09701742298638,
+      "learning_rate": 6.719681908548708e-07,
+      "loss": 0.7279,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.255417844602752,
+      "learning_rate": 6.739562624254473e-07,
+      "loss": 0.6992,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.715787721672236,
+      "learning_rate": 6.759443339960238e-07,
+      "loss": 0.7005,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.663202418344564,
+      "learning_rate": 6.779324055666004e-07,
+      "loss": 0.679,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.518507768649536,
+      "learning_rate": 6.799204771371769e-07,
+      "loss": 0.7018,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.246755644395869,
+      "learning_rate": 6.819085487077535e-07,
+      "loss": 0.7148,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.423848041665416,
+      "learning_rate": 6.8389662027833e-07,
+      "loss": 0.7031,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.81795434462883,
+      "learning_rate": 6.858846918489065e-07,
+      "loss": 0.694,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.803204587071182,
+      "learning_rate": 6.878727634194831e-07,
+      "loss": 0.7174,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.044685463189259,
+      "learning_rate": 6.898608349900596e-07,
+      "loss": 0.707,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 19.31576570191332,
+      "learning_rate": 6.918489065606361e-07,
+      "loss": 0.7357,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5244057852237125,
+      "learning_rate": 6.938369781312126e-07,
+      "loss": 0.6842,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.672189949041798,
+      "learning_rate": 6.958250497017892e-07,
+      "loss": 0.7018,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.708731563152467,
+      "learning_rate": 6.978131212723658e-07,
+      "loss": 0.6888,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.346615596017632,
+      "learning_rate": 6.998011928429423e-07,
+      "loss": 0.6882,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.217379093401352,
+      "learning_rate": 7.017892644135189e-07,
+      "loss": 0.7044,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.22298505042355,
+      "learning_rate": 7.037773359840954e-07,
+      "loss": 0.7096,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.965405242738737,
+      "learning_rate": 7.05765407554672e-07,
+      "loss": 0.6862,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.07423428462319,
+      "learning_rate": 7.077534791252485e-07,
+      "loss": 0.6934,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.538680568175445,
+      "learning_rate": 7.09741550695825e-07,
+      "loss": 0.6934,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.447196378650604,
+      "learning_rate": 7.117296222664016e-07,
+      "loss": 0.6882,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.595457251445302,
+      "learning_rate": 7.137176938369781e-07,
+      "loss": 0.7116,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.270679139037185,
+      "learning_rate": 7.157057654075547e-07,
+      "loss": 0.6992,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.95894194849116,
+      "learning_rate": 7.176938369781311e-07,
+      "loss": 0.6934,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.580946630106774,
+      "learning_rate": 7.196819085487077e-07,
+      "loss": 0.6979,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.934452903396781,
+      "learning_rate": 7.216699801192842e-07,
+      "loss": 0.6895,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.242218230831424,
+      "learning_rate": 7.236580516898608e-07,
+      "loss": 0.6882,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.582871328306762,
+      "learning_rate": 7.256461232604374e-07,
+      "loss": 0.7057,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.420252218903687,
+      "learning_rate": 7.276341948310138e-07,
+      "loss": 0.6992,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.218207045599389,
+      "learning_rate": 7.296222664015904e-07,
+      "loss": 0.6914,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.992630320930392,
+      "learning_rate": 7.31610337972167e-07,
+      "loss": 0.6888,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.260562920680358,
+      "learning_rate": 7.335984095427436e-07,
+      "loss": 0.7064,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.62249577164887,
+      "learning_rate": 7.3558648111332e-07,
+      "loss": 0.6927,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.023191653089489,
+      "learning_rate": 7.375745526838966e-07,
+      "loss": 0.6966,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.715700792629364,
+      "learning_rate": 7.395626242544732e-07,
+      "loss": 0.6947,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.709381222170816,
+      "learning_rate": 7.415506958250497e-07,
+      "loss": 0.7096,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.9668223725696095,
+      "learning_rate": 7.435387673956262e-07,
+      "loss": 0.6914,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.429722063057382,
+      "learning_rate": 7.455268389662027e-07,
+      "loss": 0.6992,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.765781287024563,
+      "learning_rate": 7.475149105367793e-07,
+      "loss": 0.6992,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.237920315094957,
+      "learning_rate": 7.495029821073558e-07,
+      "loss": 0.6868,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.753510666237331,
+      "learning_rate": 7.514910536779324e-07,
+      "loss": 0.6953,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.819031307744193,
+      "learning_rate": 7.534791252485088e-07,
+      "loss": 0.6979,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.359119021154847,
+      "learning_rate": 7.554671968190854e-07,
+      "loss": 0.6882,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.975239810305357,
+      "learning_rate": 7.57455268389662e-07,
+      "loss": 0.6986,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.391374484385123,
+      "learning_rate": 7.594433399602385e-07,
+      "loss": 0.6882,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.18481705077234,
+      "learning_rate": 7.614314115308151e-07,
+      "loss": 0.6973,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7108158854706517,
+      "learning_rate": 7.634194831013916e-07,
+      "loss": 0.6947,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.969229283378244,
+      "learning_rate": 7.654075546719682e-07,
+      "loss": 0.6966,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.498605957072399,
+      "learning_rate": 7.673956262425447e-07,
+      "loss": 0.6888,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.167664576577339,
+      "learning_rate": 7.693836978131213e-07,
+      "loss": 0.6875,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.648576794861765,
+      "learning_rate": 7.713717693836978e-07,
+      "loss": 0.6992,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.69657815357591,
+      "learning_rate": 7.733598409542743e-07,
+      "loss": 0.6953,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.759916539951178,
+      "learning_rate": 7.753479125248509e-07,
+      "loss": 0.7018,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.002824055520336,
+      "learning_rate": 7.773359840954274e-07,
+      "loss": 0.6921,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.27988438047741,
+      "learning_rate": 7.793240556660039e-07,
+      "loss": 0.6882,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.899708947821971,
+      "learning_rate": 7.813121272365804e-07,
+      "loss": 0.6914,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.310142559440736,
+      "learning_rate": 7.83300198807157e-07,
+      "loss": 0.6868,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.348290669805046,
+      "learning_rate": 7.852882703777336e-07,
+      "loss": 0.6914,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.074712235233003,
+      "learning_rate": 7.8727634194831e-07,
+      "loss": 0.6973,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.755669635860118,
+      "learning_rate": 7.892644135188866e-07,
+      "loss": 0.6764,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 22.060253342539383,
+      "learning_rate": 7.912524850894632e-07,
+      "loss": 0.7233,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.622080041868642,
+      "learning_rate": 7.932405566600398e-07,
+      "loss": 0.6908,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.01335992502021,
+      "learning_rate": 7.952286282306163e-07,
+      "loss": 0.6706,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.430039412991032,
+      "learning_rate": 7.972166998011928e-07,
+      "loss": 0.7155,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.5076534878656505,
+      "learning_rate": 7.992047713717694e-07,
+      "loss": 0.6999,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.408574469276335,
+      "learning_rate": 8.011928429423459e-07,
+      "loss": 0.6823,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.422820808627556,
+      "learning_rate": 8.031809145129225e-07,
+      "loss": 0.6921,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.449839818787293,
+      "learning_rate": 8.051689860834989e-07,
+      "loss": 0.7116,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.761614260195895,
+      "learning_rate": 8.071570576540755e-07,
+      "loss": 0.7096,
+      "step": 406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.6536118745375745,
+      "learning_rate": 8.09145129224652e-07,
+      "loss": 0.7012,
+      "step": 407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.80124304011683,
+      "learning_rate": 8.111332007952286e-07,
+      "loss": 0.6934,
+      "step": 408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.076430830711598,
+      "learning_rate": 8.131212723658051e-07,
+      "loss": 0.6921,
+      "step": 409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5804951230610906,
+      "learning_rate": 8.151093439363816e-07,
+      "loss": 0.7018,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.663045435912005,
+      "learning_rate": 8.170974155069582e-07,
+      "loss": 0.6986,
+      "step": 411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.617753884722543,
+      "learning_rate": 8.190854870775347e-07,
+      "loss": 0.6927,
+      "step": 412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.263784393923575,
+      "learning_rate": 8.210735586481114e-07,
+      "loss": 0.694,
+      "step": 413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.087364771630666,
+      "learning_rate": 8.230616302186878e-07,
+      "loss": 0.6868,
+      "step": 414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.845926581637131,
+      "learning_rate": 8.250497017892644e-07,
+      "loss": 0.6868,
+      "step": 415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.35737299848163,
+      "learning_rate": 8.27037773359841e-07,
+      "loss": 0.6953,
+      "step": 416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.9204887768672,
+      "learning_rate": 8.290258449304175e-07,
+      "loss": 0.6888,
+      "step": 417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.01726074190478,
+      "learning_rate": 8.31013916500994e-07,
+      "loss": 0.6986,
+      "step": 418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.910697122066416,
+      "learning_rate": 8.330019880715705e-07,
+      "loss": 0.7103,
+      "step": 419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.00278808441115,
+      "learning_rate": 8.349900596421471e-07,
+      "loss": 0.6927,
+      "step": 420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.85072521664762,
+      "learning_rate": 8.369781312127236e-07,
+      "loss": 0.7044,
+      "step": 421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.291031958074681,
+      "learning_rate": 8.389662027833002e-07,
+      "loss": 0.6829,
+      "step": 422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.345945126420645,
+      "learning_rate": 8.409542743538766e-07,
+      "loss": 0.7057,
+      "step": 423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.496387549038152,
+      "learning_rate": 8.429423459244532e-07,
+      "loss": 0.6784,
+      "step": 424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.203790711474915,
+      "learning_rate": 8.449304174950298e-07,
+      "loss": 0.7103,
+      "step": 425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.173600079443536,
+      "learning_rate": 8.469184890656063e-07,
+      "loss": 0.7064,
+      "step": 426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.3966878786481844,
+      "learning_rate": 8.489065606361828e-07,
+      "loss": 0.6986,
+      "step": 427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.41491011466304,
+      "learning_rate": 8.508946322067594e-07,
+      "loss": 0.7031,
+      "step": 428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 25.221646466088906,
+      "learning_rate": 8.52882703777336e-07,
+      "loss": 0.7031,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.440835669617919,
+      "learning_rate": 8.548707753479126e-07,
+      "loss": 0.6882,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.410414979566638,
+      "learning_rate": 8.56858846918489e-07,
+      "loss": 0.6829,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6541470471282707,
+      "learning_rate": 8.588469184890656e-07,
+      "loss": 0.694,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.825399853341736,
+      "learning_rate": 8.608349900596421e-07,
+      "loss": 0.7012,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.241159712497785,
+      "learning_rate": 8.628230616302187e-07,
+      "loss": 0.6895,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.64648569340244,
+      "learning_rate": 8.648111332007952e-07,
+      "loss": 0.7109,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.618834453677141,
+      "learning_rate": 8.667992047713717e-07,
+      "loss": 0.6882,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.961434646159333,
+      "learning_rate": 8.687872763419482e-07,
+      "loss": 0.6882,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.412230465469577,
+      "learning_rate": 8.707753479125248e-07,
+      "loss": 0.694,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.271707813629812,
+      "learning_rate": 8.727634194831014e-07,
+      "loss": 0.7025,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9123434932165413,
+      "learning_rate": 8.747514910536778e-07,
+      "loss": 0.694,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.727255073248538,
+      "learning_rate": 8.767395626242544e-07,
+      "loss": 0.6829,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.663170765197002,
+      "learning_rate": 8.787276341948309e-07,
+      "loss": 0.6986,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.58580329212734,
+      "learning_rate": 8.807157057654076e-07,
+      "loss": 0.6986,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.382640569657567,
+      "learning_rate": 8.82703777335984e-07,
+      "loss": 0.7005,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.998459850477437,
+      "learning_rate": 8.846918489065606e-07,
+      "loss": 0.6966,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.196969531762864,
+      "learning_rate": 8.866799204771372e-07,
+      "loss": 0.694,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.110700316840523,
+      "learning_rate": 8.886679920477137e-07,
+      "loss": 0.6908,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.872377673505329,
+      "learning_rate": 8.906560636182903e-07,
+      "loss": 0.6979,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.792518513803613,
+      "learning_rate": 8.926441351888667e-07,
+      "loss": 0.7038,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.936113753412021,
+      "learning_rate": 8.946322067594433e-07,
+      "loss": 0.696,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.402359219039022,
+      "learning_rate": 8.966202783300198e-07,
+      "loss": 0.7018,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.250859017357887,
+      "learning_rate": 8.986083499005964e-07,
+      "loss": 0.6908,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.647497302263608,
+      "learning_rate": 9.005964214711729e-07,
+      "loss": 0.6986,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.239188208441256,
+      "learning_rate": 9.025844930417494e-07,
+      "loss": 0.7129,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.01944299677375,
+      "learning_rate": 9.04572564612326e-07,
+      "loss": 0.7025,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.47478599367114,
+      "learning_rate": 9.065606361829025e-07,
+      "loss": 0.6973,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.07670611041669,
+      "learning_rate": 9.085487077534792e-07,
+      "loss": 0.6966,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.938007873172984,
+      "learning_rate": 9.105367793240556e-07,
+      "loss": 0.6953,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.398989247042089,
+      "learning_rate": 9.125248508946322e-07,
+      "loss": 0.6882,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.234039483570864,
+      "learning_rate": 9.145129224652088e-07,
+      "loss": 0.6953,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 19.256819380324774,
+      "learning_rate": 9.165009940357853e-07,
+      "loss": 0.7077,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.085103493094819,
+      "learning_rate": 9.184890656063618e-07,
+      "loss": 0.6992,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.179958248418888,
+      "learning_rate": 9.204771371769383e-07,
+      "loss": 0.6934,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.99224820934023,
+      "learning_rate": 9.224652087475149e-07,
+      "loss": 0.6927,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.601916318311844,
+      "learning_rate": 9.244532803180914e-07,
+      "loss": 0.6868,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.122625753962767,
+      "learning_rate": 9.264413518886679e-07,
+      "loss": 0.6797,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.061398905431677,
+      "learning_rate": 9.284294234592445e-07,
+      "loss": 0.7057,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.285879006891568,
+      "learning_rate": 9.30417495029821e-07,
+      "loss": 0.7038,
+      "step": 468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.9173601426388425,
+      "learning_rate": 9.324055666003976e-07,
+      "loss": 0.707,
+      "step": 469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.356312970302222,
+      "learning_rate": 9.343936381709741e-07,
+      "loss": 0.6986,
+      "step": 470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.220522688834469,
+      "learning_rate": 9.363817097415506e-07,
+      "loss": 0.6947,
+      "step": 471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.275996947554272,
+      "learning_rate": 9.383697813121271e-07,
+      "loss": 0.6849,
+      "step": 472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.3495858663739595,
+      "learning_rate": 9.403578528827038e-07,
+      "loss": 0.6888,
+      "step": 473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.369505746521481,
+      "learning_rate": 9.423459244532804e-07,
+      "loss": 0.7057,
+      "step": 474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.391852754326754,
+      "learning_rate": 9.443339960238568e-07,
+      "loss": 0.6966,
+      "step": 475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.950069493750081,
+      "learning_rate": 9.463220675944334e-07,
+      "loss": 0.6934,
+      "step": 476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.024658626942301,
+      "learning_rate": 9.483101391650099e-07,
+      "loss": 0.6953,
+      "step": 477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.691549091054149,
+      "learning_rate": 9.502982107355865e-07,
+      "loss": 0.6829,
+      "step": 478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.775700605840227,
+      "learning_rate": 9.522862823061629e-07,
+      "loss": 0.7096,
+      "step": 479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0041927243758595,
+      "learning_rate": 9.542743538767394e-07,
+      "loss": 0.6862,
+      "step": 480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.533423179187292,
+      "learning_rate": 9.56262425447316e-07,
+      "loss": 0.7018,
+      "step": 481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 17.884333810444446,
+      "learning_rate": 9.582504970178926e-07,
+      "loss": 0.7142,
+      "step": 482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.359430917658416,
+      "learning_rate": 9.602385685884692e-07,
+      "loss": 0.6901,
+      "step": 483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.146621502821807,
+      "learning_rate": 9.622266401590456e-07,
+      "loss": 0.7005,
+      "step": 484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.080225126192796,
+      "learning_rate": 9.642147117296222e-07,
+      "loss": 0.6934,
+      "step": 485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.267687070628451,
+      "learning_rate": 9.662027833001988e-07,
+      "loss": 0.6999,
+      "step": 486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.032027219328823,
+      "learning_rate": 9.681908548707752e-07,
+      "loss": 0.6901,
+      "step": 487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.583521019047392,
+      "learning_rate": 9.701789264413519e-07,
+      "loss": 0.6829,
+      "step": 488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.84528949586536,
+      "learning_rate": 9.721669980119285e-07,
+      "loss": 0.6934,
+      "step": 489
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9676123568870856,
+      "learning_rate": 9.741550695825049e-07,
+      "loss": 0.6914,
+      "step": 490
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.0339132379314995,
+      "learning_rate": 9.761431411530815e-07,
+      "loss": 0.7018,
+      "step": 491
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.324983497481007,
+      "learning_rate": 9.78131212723658e-07,
+      "loss": 0.7031,
+      "step": 492
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.743191483733425,
+      "learning_rate": 9.801192842942345e-07,
+      "loss": 0.6934,
+      "step": 493
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.35293849240906,
+      "learning_rate": 9.82107355864811e-07,
+      "loss": 0.6927,
+      "step": 494
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.315914290259441,
+      "learning_rate": 9.840954274353877e-07,
+      "loss": 0.6751,
+      "step": 495
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.522622491525206,
+      "learning_rate": 9.860834990059643e-07,
+      "loss": 0.6979,
+      "step": 496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.682024879932865,
+      "learning_rate": 9.880715705765407e-07,
+      "loss": 0.7025,
+      "step": 497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.530071851446565,
+      "learning_rate": 9.900596421471173e-07,
+      "loss": 0.6979,
+      "step": 498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.31542231209298,
+      "learning_rate": 9.920477137176937e-07,
+      "loss": 0.7038,
+      "step": 499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.090412387018624,
+      "learning_rate": 9.940357852882703e-07,
+      "loss": 0.6979,
+      "step": 500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.262156457250189,
+      "learning_rate": 9.960238568588467e-07,
+      "loss": 0.6953,
+      "step": 501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.183620395435941,
+      "learning_rate": 9.980119284294233e-07,
+      "loss": 0.6947,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.163453231005009,
+      "learning_rate": 1e-06,
+      "loss": 0.6842,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.418602916514263,
+      "learning_rate": 1.0019880715705764e-06,
+      "loss": 0.6979,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 20.072188656649313,
+      "learning_rate": 1.0039761431411532e-06,
+      "loss": 0.7129,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.619144111725088,
+      "learning_rate": 1.0059642147117296e-06,
+      "loss": 0.6888,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.226641974353168,
+      "learning_rate": 1.0079522862823062e-06,
+      "loss": 0.6914,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.677779879467535,
+      "learning_rate": 1.0099403578528826e-06,
+      "loss": 0.6986,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.02624753693883,
+      "learning_rate": 1.0119284294234592e-06,
+      "loss": 0.7122,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7104013240754985,
+      "learning_rate": 1.0139165009940358e-06,
+      "loss": 0.6849,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.260916557391647,
+      "learning_rate": 1.0159045725646122e-06,
+      "loss": 0.7012,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.901448748100435,
+      "learning_rate": 1.0178926441351888e-06,
+      "loss": 0.7005,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 15.43048172735375,
+      "learning_rate": 1.0198807157057654e-06,
+      "loss": 0.709,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.304640476317108,
+      "learning_rate": 1.0218687872763418e-06,
+      "loss": 0.724,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.8421785817269685,
+      "learning_rate": 1.0238568588469184e-06,
+      "loss": 0.6973,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.06734693129196,
+      "learning_rate": 1.025844930417495e-06,
+      "loss": 0.6966,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9452917410078827,
+      "learning_rate": 1.0278330019880714e-06,
+      "loss": 0.7005,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.430703518510104,
+      "learning_rate": 1.029821073558648e-06,
+      "loss": 0.7083,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.255136346579855,
+      "learning_rate": 1.0318091451292247e-06,
+      "loss": 0.707,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.970970427974498,
+      "learning_rate": 1.0337972166998013e-06,
+      "loss": 0.694,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 19.13403930888799,
+      "learning_rate": 1.0357852882703777e-06,
+      "loss": 0.7331,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.859881512946767,
+      "learning_rate": 1.037773359840954e-06,
+      "loss": 0.7103,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.879063950265113,
+      "learning_rate": 1.039761431411531e-06,
+      "loss": 0.6927,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.6663259397210695,
+      "learning_rate": 1.0417495029821073e-06,
+      "loss": 0.6882,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9339650397260515,
+      "learning_rate": 1.0437375745526837e-06,
+      "loss": 0.6979,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.1883454948212,
+      "learning_rate": 1.0457256461232605e-06,
+      "loss": 0.7012,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.077547783119257,
+      "learning_rate": 1.047713717693837e-06,
+      "loss": 0.6953,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.547127559078849,
+      "learning_rate": 1.0497017892644135e-06,
+      "loss": 0.6979,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.28098370004434,
+      "learning_rate": 1.05168986083499e-06,
+      "loss": 0.7005,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.516615459205955,
+      "learning_rate": 1.0536779324055665e-06,
+      "loss": 0.7096,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.497336664777578,
+      "learning_rate": 1.0556660039761431e-06,
+      "loss": 0.7038,
+      "step": 531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.322412516111854,
+      "learning_rate": 1.0576540755467195e-06,
+      "loss": 0.6966,
+      "step": 532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.006924577776056,
+      "learning_rate": 1.0596421471172964e-06,
+      "loss": 0.6901,
+      "step": 533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.304927135750833,
+      "learning_rate": 1.0616302186878728e-06,
+      "loss": 0.7025,
+      "step": 534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.094680724742609,
+      "learning_rate": 1.0636182902584492e-06,
+      "loss": 0.6979,
+      "step": 535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.163511060546819,
+      "learning_rate": 1.0656063618290258e-06,
+      "loss": 0.6979,
+      "step": 536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.444300291700802,
+      "learning_rate": 1.0675944333996024e-06,
+      "loss": 0.6908,
+      "step": 537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.181437040736761,
+      "learning_rate": 1.0695825049701788e-06,
+      "loss": 0.6895,
+      "step": 538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.391461439272584,
+      "learning_rate": 1.0715705765407554e-06,
+      "loss": 0.7031,
+      "step": 539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.12804606707508,
+      "learning_rate": 1.073558648111332e-06,
+      "loss": 0.7031,
+      "step": 540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.478678982006108,
+      "learning_rate": 1.0755467196819086e-06,
+      "loss": 0.6947,
+      "step": 541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6733800384955217,
+      "learning_rate": 1.077534791252485e-06,
+      "loss": 0.6803,
+      "step": 542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.894014155780146,
+      "learning_rate": 1.0795228628230614e-06,
+      "loss": 0.6992,
+      "step": 543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.988728839969133,
+      "learning_rate": 1.0815109343936382e-06,
+      "loss": 0.696,
+      "step": 544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.632688240609994,
+      "learning_rate": 1.0834990059642146e-06,
+      "loss": 0.6836,
+      "step": 545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.636605857739047,
+      "learning_rate": 1.0854870775347912e-06,
+      "loss": 0.7005,
+      "step": 546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.832613026014412,
+      "learning_rate": 1.0874751491053679e-06,
+      "loss": 0.7031,
+      "step": 547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.36679552263619,
+      "learning_rate": 1.0894632206759443e-06,
+      "loss": 0.6966,
+      "step": 548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.247422152422811,
+      "learning_rate": 1.0914512922465209e-06,
+      "loss": 0.6823,
+      "step": 549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.65899564673693,
+      "learning_rate": 1.0934393638170973e-06,
+      "loss": 0.7038,
+      "step": 550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.834442825999414,
+      "learning_rate": 1.095427435387674e-06,
+      "loss": 0.6934,
+      "step": 551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.1319823971564915,
+      "learning_rate": 1.0974155069582505e-06,
+      "loss": 0.6953,
+      "step": 552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.43470451247899,
+      "learning_rate": 1.0994035785288269e-06,
+      "loss": 0.7064,
+      "step": 553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.433089481322773,
+      "learning_rate": 1.1013916500994037e-06,
+      "loss": 0.7109,
+      "step": 554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.434571010222418,
+      "learning_rate": 1.10337972166998e-06,
+      "loss": 0.6868,
+      "step": 555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.29523521974586,
+      "learning_rate": 1.1053677932405565e-06,
+      "loss": 0.7135,
+      "step": 556
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.319717067813744,
+      "learning_rate": 1.1073558648111331e-06,
+      "loss": 0.6947,
+      "step": 557
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.115030841778958,
+      "learning_rate": 1.1093439363817097e-06,
+      "loss": 0.6992,
+      "step": 558
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.382198520422729,
+      "learning_rate": 1.1113320079522863e-06,
+      "loss": 0.696,
+      "step": 559
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.330662841633563,
+      "learning_rate": 1.1133200795228627e-06,
+      "loss": 0.6992,
+      "step": 560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.414789693749533,
+      "learning_rate": 1.1153081510934391e-06,
+      "loss": 0.6901,
+      "step": 561
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7086701740571186,
+      "learning_rate": 1.117296222664016e-06,
+      "loss": 0.6868,
+      "step": 562
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.97436901558898,
+      "learning_rate": 1.1192842942345924e-06,
+      "loss": 0.694,
+      "step": 563
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.606245042088812,
+      "learning_rate": 1.121272365805169e-06,
+      "loss": 0.7012,
+      "step": 564
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.252655827873364,
+      "learning_rate": 1.1232604373757456e-06,
+      "loss": 0.6875,
+      "step": 565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.511350212359119,
+      "learning_rate": 1.125248508946322e-06,
+      "loss": 0.7064,
+      "step": 566
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.9828458133611,
+      "learning_rate": 1.1272365805168986e-06,
+      "loss": 0.7083,
+      "step": 567
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.6594079055623485,
+      "learning_rate": 1.129224652087475e-06,
+      "loss": 0.6986,
+      "step": 568
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.686213666272609,
+      "learning_rate": 1.1312127236580516e-06,
+      "loss": 0.6992,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.080922020949418,
+      "learning_rate": 1.1332007952286282e-06,
+      "loss": 0.6914,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.643354733008756,
+      "learning_rate": 1.1351888667992046e-06,
+      "loss": 0.6855,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.363525015418019,
+      "learning_rate": 1.1371769383697814e-06,
+      "loss": 0.6895,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.229756977715118,
+      "learning_rate": 1.1391650099403578e-06,
+      "loss": 0.6953,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.780942098933473,
+      "learning_rate": 1.1411530815109342e-06,
+      "loss": 0.6934,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4199084015895833,
+      "learning_rate": 1.1431411530815108e-06,
+      "loss": 0.6882,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.2390547986045055,
+      "learning_rate": 1.1451292246520874e-06,
+      "loss": 0.6875,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.11115449556599,
+      "learning_rate": 1.147117296222664e-06,
+      "loss": 0.7031,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.316263138657083,
+      "learning_rate": 1.1491053677932405e-06,
+      "loss": 0.6966,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.197562937472902,
+      "learning_rate": 1.151093439363817e-06,
+      "loss": 0.6953,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.496578413211144,
+      "learning_rate": 1.1530815109343937e-06,
+      "loss": 0.696,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.223130693210218,
+      "learning_rate": 1.15506958250497e-06,
+      "loss": 0.6921,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.211105917304277,
+      "learning_rate": 1.1570576540755465e-06,
+      "loss": 0.6901,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.98675792202501,
+      "learning_rate": 1.1590457256461233e-06,
+      "loss": 0.7038,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.324975606750778,
+      "learning_rate": 1.1610337972166997e-06,
+      "loss": 0.6875,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.823120912532936,
+      "learning_rate": 1.1630218687872763e-06,
+      "loss": 0.6921,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.758654226409222,
+      "learning_rate": 1.165009940357853e-06,
+      "loss": 0.6868,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.81769607925928,
+      "learning_rate": 1.1669980119284293e-06,
+      "loss": 0.7012,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.097128201123073,
+      "learning_rate": 1.168986083499006e-06,
+      "loss": 0.6868,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.067156859268307,
+      "learning_rate": 1.1709741550695823e-06,
+      "loss": 0.6868,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.208893985365875,
+      "learning_rate": 1.1729622266401591e-06,
+      "loss": 0.6966,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9518001746524343,
+      "learning_rate": 1.1749502982107355e-06,
+      "loss": 0.7012,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.509828392003515,
+      "learning_rate": 1.176938369781312e-06,
+      "loss": 0.6947,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.599005545430293,
+      "learning_rate": 1.1789264413518888e-06,
+      "loss": 0.6914,
+      "step": 593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.685852035882882,
+      "learning_rate": 1.1809145129224652e-06,
+      "loss": 0.6973,
+      "step": 594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.609262652777283,
+      "learning_rate": 1.1829025844930416e-06,
+      "loss": 0.6914,
+      "step": 595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.897498417845508,
+      "learning_rate": 1.1848906560636182e-06,
+      "loss": 0.6862,
+      "step": 596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8035611272751213,
+      "learning_rate": 1.1868787276341948e-06,
+      "loss": 0.6947,
+      "step": 597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.500835531422666,
+      "learning_rate": 1.1888667992047714e-06,
+      "loss": 0.6934,
+      "step": 598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.710218068328336,
+      "learning_rate": 1.1908548707753478e-06,
+      "loss": 0.7005,
+      "step": 599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.14820936720081,
+      "learning_rate": 1.1928429423459244e-06,
+      "loss": 0.7057,
+      "step": 600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.442736038552533,
+      "learning_rate": 1.194831013916501e-06,
+      "loss": 0.6868,
+      "step": 601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.036574265775883,
+      "learning_rate": 1.1968190854870774e-06,
+      "loss": 0.6966,
+      "step": 602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.271070586493611,
+      "learning_rate": 1.198807157057654e-06,
+      "loss": 0.6921,
+      "step": 603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.252597700257553,
+      "learning_rate": 1.2007952286282306e-06,
+      "loss": 0.6777,
+      "step": 604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.323094064669315,
+      "learning_rate": 1.202783300198807e-06,
+      "loss": 0.6888,
+      "step": 605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.6766081651647715,
+      "learning_rate": 1.2047713717693836e-06,
+      "loss": 0.7135,
+      "step": 606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.174125467267188,
+      "learning_rate": 1.2067594433399603e-06,
+      "loss": 0.6901,
+      "step": 607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.643546738305966,
+      "learning_rate": 1.2087475149105367e-06,
+      "loss": 0.6999,
+      "step": 608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.72836797884339,
+      "learning_rate": 1.2107355864811133e-06,
+      "loss": 0.6882,
+      "step": 609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.597207955075841,
+      "learning_rate": 1.2127236580516897e-06,
+      "loss": 0.707,
+      "step": 610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.280970898747162,
+      "learning_rate": 1.2147117296222665e-06,
+      "loss": 0.709,
+      "step": 611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.068098018157594,
+      "learning_rate": 1.2166998011928429e-06,
+      "loss": 0.7122,
+      "step": 612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8420286689733354,
+      "learning_rate": 1.2186878727634193e-06,
+      "loss": 0.6816,
+      "step": 613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.688499458491182,
+      "learning_rate": 1.220675944333996e-06,
+      "loss": 0.6986,
+      "step": 614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.298814618960796,
+      "learning_rate": 1.2226640159045725e-06,
+      "loss": 0.679,
+      "step": 615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.694461714325365,
+      "learning_rate": 1.2246520874751491e-06,
+      "loss": 0.696,
+      "step": 616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.704332122573015,
+      "learning_rate": 1.2266401590457255e-06,
+      "loss": 0.6862,
+      "step": 617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0239484795059814,
+      "learning_rate": 1.2286282306163021e-06,
+      "loss": 0.696,
+      "step": 618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.196575200588363,
+      "learning_rate": 1.2306163021868787e-06,
+      "loss": 0.7168,
+      "step": 619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.506721977824297,
+      "learning_rate": 1.2326043737574551e-06,
+      "loss": 0.6986,
+      "step": 620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.727090407606426,
+      "learning_rate": 1.234592445328032e-06,
+      "loss": 0.7018,
+      "step": 621
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.121996299907073,
+      "learning_rate": 1.2365805168986084e-06,
+      "loss": 0.7044,
+      "step": 622
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.736573134637975,
+      "learning_rate": 1.2385685884691848e-06,
+      "loss": 0.7064,
+      "step": 623
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.653981756090145,
+      "learning_rate": 1.2405566600397614e-06,
+      "loss": 0.7031,
+      "step": 624
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.375611345406801,
+      "learning_rate": 1.242544731610338e-06,
+      "loss": 0.6895,
+      "step": 625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6994060349320694,
+      "learning_rate": 1.2445328031809144e-06,
+      "loss": 0.6914,
+      "step": 626
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.095606249831906,
+      "learning_rate": 1.246520874751491e-06,
+      "loss": 0.7044,
+      "step": 627
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.505045665206723,
+      "learning_rate": 1.2485089463220676e-06,
+      "loss": 0.6953,
+      "step": 628
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.382530463728617,
+      "learning_rate": 1.2504970178926442e-06,
+      "loss": 0.7038,
+      "step": 629
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.35557147372081,
+      "learning_rate": 1.2524850894632206e-06,
+      "loss": 0.6875,
+      "step": 630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.857828621907085,
+      "learning_rate": 1.254473161033797e-06,
+      "loss": 0.6855,
+      "step": 631
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.186696923835534,
+      "learning_rate": 1.2564612326043738e-06,
+      "loss": 0.6992,
+      "step": 632
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.261246141549304,
+      "learning_rate": 1.2584493041749502e-06,
+      "loss": 0.7077,
+      "step": 633
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.832393800920693,
+      "learning_rate": 1.2604373757455268e-06,
+      "loss": 0.709,
+      "step": 634
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.02198232140092,
+      "learning_rate": 1.2624254473161034e-06,
+      "loss": 0.7096,
+      "step": 635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.15939200119468,
+      "learning_rate": 1.2644135188866798e-06,
+      "loss": 0.6901,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7424973055227113,
+      "learning_rate": 1.2664015904572565e-06,
+      "loss": 0.6947,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.630347838780393,
+      "learning_rate": 1.2683896620278328e-06,
+      "loss": 0.6888,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4869373451962167,
+      "learning_rate": 1.2703777335984095e-06,
+      "loss": 0.696,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.625763686153396,
+      "learning_rate": 1.272365805168986e-06,
+      "loss": 0.7012,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.7216312323956116,
+      "learning_rate": 1.2743538767395625e-06,
+      "loss": 0.6979,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.9474608779915785,
+      "learning_rate": 1.2763419483101393e-06,
+      "loss": 0.6934,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.24545569342793,
+      "learning_rate": 1.2783300198807157e-06,
+      "loss": 0.6979,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.746720068903385,
+      "learning_rate": 1.280318091451292e-06,
+      "loss": 0.6875,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4487058475976426,
+      "learning_rate": 1.2823061630218687e-06,
+      "loss": 0.6927,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.794807657098796,
+      "learning_rate": 1.2842942345924453e-06,
+      "loss": 0.6842,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.54106958413656,
+      "learning_rate": 1.286282306163022e-06,
+      "loss": 0.709,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.058476441368475,
+      "learning_rate": 1.2882703777335983e-06,
+      "loss": 0.6901,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.849345462002358,
+      "learning_rate": 1.2902584493041747e-06,
+      "loss": 0.6999,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.417058739590987,
+      "learning_rate": 1.2922465208747515e-06,
+      "loss": 0.6999,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.072228597645178,
+      "learning_rate": 1.294234592445328e-06,
+      "loss": 0.6842,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.185718715847009,
+      "learning_rate": 1.2962226640159043e-06,
+      "loss": 0.696,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.0774946461725925,
+      "learning_rate": 1.2982107355864812e-06,
+      "loss": 0.6882,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.588609709820584,
+      "learning_rate": 1.3001988071570576e-06,
+      "loss": 0.7077,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.2333722844678,
+      "learning_rate": 1.3021868787276342e-06,
+      "loss": 0.6868,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8894288895334497,
+      "learning_rate": 1.3041749502982106e-06,
+      "loss": 0.6934,
+      "step": 656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.06667322926377,
+      "learning_rate": 1.3061630218687872e-06,
+      "loss": 0.6875,
+      "step": 657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.0186944117810866,
+      "learning_rate": 1.3081510934393638e-06,
+      "loss": 0.6966,
+      "step": 658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.784197743621213,
+      "learning_rate": 1.3101391650099402e-06,
+      "loss": 0.694,
+      "step": 659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.900353426606522,
+      "learning_rate": 1.312127236580517e-06,
+      "loss": 0.7057,
+      "step": 660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0124914382529964,
+      "learning_rate": 1.3141153081510934e-06,
+      "loss": 0.6855,
+      "step": 661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8834033368839913,
+      "learning_rate": 1.3161033797216698e-06,
+      "loss": 0.6901,
+      "step": 662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.567263050881383,
+      "learning_rate": 1.3180914512922466e-06,
+      "loss": 0.7038,
+      "step": 663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.242099647498014,
+      "learning_rate": 1.320079522862823e-06,
+      "loss": 0.7012,
+      "step": 664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9723518901399488,
+      "learning_rate": 1.3220675944333994e-06,
+      "loss": 0.6882,
+      "step": 665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.461923132465631,
+      "learning_rate": 1.324055666003976e-06,
+      "loss": 0.694,
+      "step": 666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.58953646806992,
+      "learning_rate": 1.3260437375745526e-06,
+      "loss": 0.6908,
+      "step": 667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.47409607666015,
+      "learning_rate": 1.3280318091451293e-06,
+      "loss": 0.7129,
+      "step": 668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.213808292299258,
+      "learning_rate": 1.3300198807157057e-06,
+      "loss": 0.6849,
+      "step": 669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4717154935488086,
+      "learning_rate": 1.332007952286282e-06,
+      "loss": 0.6921,
+      "step": 670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.070376911700755,
+      "learning_rate": 1.3339960238568589e-06,
+      "loss": 0.7077,
+      "step": 671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.420341158899666,
+      "learning_rate": 1.3359840954274353e-06,
+      "loss": 0.7031,
+      "step": 672
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.613933634344162,
+      "learning_rate": 1.3379721669980119e-06,
+      "loss": 0.7031,
+      "step": 673
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5405311902519383,
+      "learning_rate": 1.3399602385685885e-06,
+      "loss": 0.6927,
+      "step": 674
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.613747829709012,
+      "learning_rate": 1.341948310139165e-06,
+      "loss": 0.6953,
+      "step": 675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9741554156520436,
+      "learning_rate": 1.3439363817097415e-06,
+      "loss": 0.6895,
+      "step": 676
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.267987670972875,
+      "learning_rate": 1.345924453280318e-06,
+      "loss": 0.694,
+      "step": 677
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.124753952560912,
+      "learning_rate": 1.3479125248508945e-06,
+      "loss": 0.6921,
+      "step": 678
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5428385266758267,
+      "learning_rate": 1.3499005964214711e-06,
+      "loss": 0.6986,
+      "step": 679
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.004663899475316,
+      "learning_rate": 1.3518886679920475e-06,
+      "loss": 0.696,
+      "step": 680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.92167149725243,
+      "learning_rate": 1.3538767395626244e-06,
+      "loss": 0.709,
+      "step": 681
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.67542665245066,
+      "learning_rate": 1.3558648111332007e-06,
+      "loss": 0.6953,
+      "step": 682
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.674141716070041,
+      "learning_rate": 1.3578528827037771e-06,
+      "loss": 0.694,
+      "step": 683
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8650857575912183,
+      "learning_rate": 1.3598409542743538e-06,
+      "loss": 0.6934,
+      "step": 684
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.575851291054732,
+      "learning_rate": 1.3618290258449304e-06,
+      "loss": 0.696,
+      "step": 685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.381687272808106,
+      "learning_rate": 1.363817097415507e-06,
+      "loss": 0.7012,
+      "step": 686
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.274017349782966,
+      "learning_rate": 1.3658051689860834e-06,
+      "loss": 0.7031,
+      "step": 687
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.28274353590572,
+      "learning_rate": 1.36779324055666e-06,
+      "loss": 0.7044,
+      "step": 688
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.126059850363927,
+      "learning_rate": 1.3697813121272366e-06,
+      "loss": 0.7109,
+      "step": 689
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7320886731214613,
+      "learning_rate": 1.371769383697813e-06,
+      "loss": 0.6921,
+      "step": 690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.970846101608013,
+      "learning_rate": 1.3737574552683896e-06,
+      "loss": 0.6947,
+      "step": 691
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.976906344046878,
+      "learning_rate": 1.3757455268389662e-06,
+      "loss": 0.6986,
+      "step": 692
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.318895732265828,
+      "learning_rate": 1.3777335984095426e-06,
+      "loss": 0.6947,
+      "step": 693
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.016659482680183,
+      "learning_rate": 1.3797216699801192e-06,
+      "loss": 0.6914,
+      "step": 694
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.06260670235136,
+      "learning_rate": 1.3817097415506958e-06,
+      "loss": 0.6842,
+      "step": 695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.509494425910525,
+      "learning_rate": 1.3836978131212722e-06,
+      "loss": 0.6921,
+      "step": 696
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.904654335595832,
+      "learning_rate": 1.3856858846918488e-06,
+      "loss": 0.6973,
+      "step": 697
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.15615686953745,
+      "learning_rate": 1.3876739562624252e-06,
+      "loss": 0.7018,
+      "step": 698
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.611280419490965,
+      "learning_rate": 1.389662027833002e-06,
+      "loss": 0.7051,
+      "step": 699
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.482498495024733,
+      "learning_rate": 1.3916500994035785e-06,
+      "loss": 0.6986,
+      "step": 700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.406271857962913,
+      "learning_rate": 1.3936381709741549e-06,
+      "loss": 0.6868,
+      "step": 701
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1281539354218375,
+      "learning_rate": 1.3956262425447317e-06,
+      "loss": 0.6927,
+      "step": 702
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1731932316236207,
+      "learning_rate": 1.397614314115308e-06,
+      "loss": 0.6901,
+      "step": 703
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.518857666658953,
+      "learning_rate": 1.3996023856858847e-06,
+      "loss": 0.7018,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.517245983851073,
+      "learning_rate": 1.401590457256461e-06,
+      "loss": 0.6888,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0719156233168006,
+      "learning_rate": 1.4035785288270377e-06,
+      "loss": 0.6953,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.0395656791189225,
+      "learning_rate": 1.4055666003976143e-06,
+      "loss": 0.6934,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.06152076356384,
+      "learning_rate": 1.4075546719681907e-06,
+      "loss": 0.7038,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.418956003729204,
+      "learning_rate": 1.4095427435387673e-06,
+      "loss": 0.6868,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.079939706110253,
+      "learning_rate": 1.411530815109344e-06,
+      "loss": 0.6849,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.737939421479942,
+      "learning_rate": 1.4135188866799203e-06,
+      "loss": 0.6901,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.920454004241152,
+      "learning_rate": 1.415506958250497e-06,
+      "loss": 0.707,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5733712631304413,
+      "learning_rate": 1.4174950298210736e-06,
+      "loss": 0.7005,
+      "step": 713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.633089096184563,
+      "learning_rate": 1.41948310139165e-06,
+      "loss": 0.7012,
+      "step": 714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.96071105235213,
+      "learning_rate": 1.4214711729622266e-06,
+      "loss": 0.6914,
+      "step": 715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.886094691804784,
+      "learning_rate": 1.4234592445328032e-06,
+      "loss": 0.7044,
+      "step": 716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.209582020112334,
+      "learning_rate": 1.4254473161033798e-06,
+      "loss": 0.7077,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.73087379011374,
+      "learning_rate": 1.4274353876739562e-06,
+      "loss": 0.6921,
+      "step": 718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.795742744308835,
+      "learning_rate": 1.4294234592445326e-06,
+      "loss": 0.7174,
+      "step": 719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.978140228777336,
+      "learning_rate": 1.4314115308151094e-06,
+      "loss": 0.6921,
+      "step": 720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.170140702662188,
+      "learning_rate": 1.4333996023856858e-06,
+      "loss": 0.6934,
+      "step": 721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.927656232446777,
+      "learning_rate": 1.4353876739562622e-06,
+      "loss": 0.7025,
+      "step": 722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.243543086265904,
+      "learning_rate": 1.437375745526839e-06,
+      "loss": 0.7012,
+      "step": 723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.802369803367568,
+      "learning_rate": 1.4393638170974154e-06,
+      "loss": 0.6868,
+      "step": 724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.29732395639185,
+      "learning_rate": 1.441351888667992e-06,
+      "loss": 0.6947,
+      "step": 725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.255932921480646,
+      "learning_rate": 1.4433399602385684e-06,
+      "loss": 0.7057,
+      "step": 726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.188879582778989,
+      "learning_rate": 1.445328031809145e-06,
+      "loss": 0.6901,
+      "step": 727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9189520420944173,
+      "learning_rate": 1.4473161033797217e-06,
+      "loss": 0.6973,
+      "step": 728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8099476699365287,
+      "learning_rate": 1.449304174950298e-06,
+      "loss": 0.6882,
+      "step": 729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.38115792724125,
+      "learning_rate": 1.4512922465208749e-06,
+      "loss": 0.707,
+      "step": 730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.85124453896501,
+      "learning_rate": 1.4532803180914513e-06,
+      "loss": 0.6934,
+      "step": 731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.03414713169438,
+      "learning_rate": 1.4552683896620277e-06,
+      "loss": 0.7005,
+      "step": 732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.0744709454407895,
+      "learning_rate": 1.4572564612326043e-06,
+      "loss": 0.7077,
+      "step": 733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7442063877091454,
+      "learning_rate": 1.459244532803181e-06,
+      "loss": 0.6966,
+      "step": 734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.571013589241142,
+      "learning_rate": 1.4612326043737573e-06,
+      "loss": 0.7025,
+      "step": 735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8455058392229227,
+      "learning_rate": 1.463220675944334e-06,
+      "loss": 0.6842,
+      "step": 736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.8481905873117634,
+      "learning_rate": 1.4652087475149105e-06,
+      "loss": 0.6934,
+      "step": 737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.86089570029459,
+      "learning_rate": 1.4671968190854871e-06,
+      "loss": 0.7038,
+      "step": 738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.011539143721105,
+      "learning_rate": 1.4691848906560635e-06,
+      "loss": 0.7005,
+      "step": 739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.5612696023568065,
+      "learning_rate": 1.47117296222664e-06,
+      "loss": 0.7005,
+      "step": 740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3346354350585243,
+      "learning_rate": 1.4731610337972167e-06,
+      "loss": 0.6992,
+      "step": 741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9369823820493375,
+      "learning_rate": 1.4751491053677931e-06,
+      "loss": 0.6895,
+      "step": 742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.186277853291797,
+      "learning_rate": 1.4771371769383698e-06,
+      "loss": 0.694,
+      "step": 743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5293727387739136,
+      "learning_rate": 1.4791252485089464e-06,
+      "loss": 0.6901,
+      "step": 744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4279978995442795,
+      "learning_rate": 1.4811133200795228e-06,
+      "loss": 0.6901,
+      "step": 745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.676656842267215,
+      "learning_rate": 1.4831013916500994e-06,
+      "loss": 0.6953,
+      "step": 746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2076244796453994,
+      "learning_rate": 1.4850894632206758e-06,
+      "loss": 0.6966,
+      "step": 747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.807778496006594,
+      "learning_rate": 1.4870775347912524e-06,
+      "loss": 0.6953,
+      "step": 748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.854190312274918,
+      "learning_rate": 1.489065606361829e-06,
+      "loss": 0.6868,
+      "step": 749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2288436256685733,
+      "learning_rate": 1.4910536779324054e-06,
+      "loss": 0.6855,
+      "step": 750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.546913073582353,
+      "learning_rate": 1.4930417495029822e-06,
+      "loss": 0.6979,
+      "step": 751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.476842305239768,
+      "learning_rate": 1.4950298210735586e-06,
+      "loss": 0.6816,
+      "step": 752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.860560601971525,
+      "learning_rate": 1.497017892644135e-06,
+      "loss": 0.6927,
+      "step": 753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.09201614461111,
+      "learning_rate": 1.4990059642147116e-06,
+      "loss": 0.6862,
+      "step": 754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.9646642375804175,
+      "learning_rate": 1.5009940357852882e-06,
+      "loss": 0.6875,
+      "step": 755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.303105821147822,
+      "learning_rate": 1.5029821073558648e-06,
+      "loss": 0.7083,
+      "step": 756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.57978633517658,
+      "learning_rate": 1.5049701789264412e-06,
+      "loss": 0.6862,
+      "step": 757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.113271185174702,
+      "learning_rate": 1.5069582504970176e-06,
+      "loss": 0.7025,
+      "step": 758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.531494791935649,
+      "learning_rate": 1.5089463220675945e-06,
+      "loss": 0.7005,
+      "step": 759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.279278305900604,
+      "learning_rate": 1.5109343936381709e-06,
+      "loss": 0.6842,
+      "step": 760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.239824252608611,
+      "learning_rate": 1.5129224652087475e-06,
+      "loss": 0.6999,
+      "step": 761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.133398688881819,
+      "learning_rate": 1.514910536779324e-06,
+      "loss": 0.696,
+      "step": 762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.092101666798966,
+      "learning_rate": 1.5168986083499005e-06,
+      "loss": 0.681,
+      "step": 763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6324887664400154,
+      "learning_rate": 1.518886679920477e-06,
+      "loss": 0.6927,
+      "step": 764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0698263954982004,
+      "learning_rate": 1.5208747514910535e-06,
+      "loss": 0.6908,
+      "step": 765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.44348397070806,
+      "learning_rate": 1.5228628230616301e-06,
+      "loss": 0.7012,
+      "step": 766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.001789144022712,
+      "learning_rate": 1.5248508946322067e-06,
+      "loss": 0.6953,
+      "step": 767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.205625044195502,
+      "learning_rate": 1.5268389662027831e-06,
+      "loss": 0.694,
+      "step": 768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.257676359377128,
+      "learning_rate": 1.52882703777336e-06,
+      "loss": 0.6934,
+      "step": 769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.066924888978332,
+      "learning_rate": 1.5308151093439363e-06,
+      "loss": 0.6992,
+      "step": 770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.626236107943773,
+      "learning_rate": 1.5328031809145127e-06,
+      "loss": 0.7031,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1571463453615696,
+      "learning_rate": 1.5347912524850893e-06,
+      "loss": 0.694,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.325181065740585,
+      "learning_rate": 1.536779324055666e-06,
+      "loss": 0.6862,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.748178990754408,
+      "learning_rate": 1.5387673956262426e-06,
+      "loss": 0.6979,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.957090298163642,
+      "learning_rate": 1.540755467196819e-06,
+      "loss": 0.694,
+      "step": 775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.260243368872378,
+      "learning_rate": 1.5427435387673956e-06,
+      "loss": 0.7018,
+      "step": 776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.121314697846971,
+      "learning_rate": 1.5447316103379722e-06,
+      "loss": 0.6999,
+      "step": 777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.621044847287063,
+      "learning_rate": 1.5467196819085486e-06,
+      "loss": 0.6868,
+      "step": 778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.633354935330933,
+      "learning_rate": 1.548707753479125e-06,
+      "loss": 0.6979,
+      "step": 779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.258156757916286,
+      "learning_rate": 1.5506958250497018e-06,
+      "loss": 0.7018,
+      "step": 780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.718622393824843,
+      "learning_rate": 1.5526838966202782e-06,
+      "loss": 0.6986,
+      "step": 781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.883887156184672,
+      "learning_rate": 1.5546719681908548e-06,
+      "loss": 0.7051,
+      "step": 782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.091463540000469,
+      "learning_rate": 1.5566600397614314e-06,
+      "loss": 0.6914,
+      "step": 783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.98827483212532,
+      "learning_rate": 1.5586481113320078e-06,
+      "loss": 0.6947,
+      "step": 784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8967075696144398,
+      "learning_rate": 1.5606361829025844e-06,
+      "loss": 0.6862,
+      "step": 785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.874866217761622,
+      "learning_rate": 1.5626242544731608e-06,
+      "loss": 0.6888,
+      "step": 786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.790886456253155,
+      "learning_rate": 1.5646123260437377e-06,
+      "loss": 0.6979,
+      "step": 787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.311910489759373,
+      "learning_rate": 1.566600397614314e-06,
+      "loss": 0.6836,
+      "step": 788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.851523073087906,
+      "learning_rate": 1.5685884691848905e-06,
+      "loss": 0.6719,
+      "step": 789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.248698390718802,
+      "learning_rate": 1.5705765407554673e-06,
+      "loss": 0.7018,
+      "step": 790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5911043318535123,
+      "learning_rate": 1.5725646123260437e-06,
+      "loss": 0.6979,
+      "step": 791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9172251752473777,
+      "learning_rate": 1.57455268389662e-06,
+      "loss": 0.6947,
+      "step": 792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.658620566041053,
+      "learning_rate": 1.5765407554671967e-06,
+      "loss": 0.7155,
+      "step": 793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.51799950136088,
+      "learning_rate": 1.5785288270377733e-06,
+      "loss": 0.7077,
+      "step": 794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.1275633695255,
+      "learning_rate": 1.58051689860835e-06,
+      "loss": 0.7031,
+      "step": 795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0903253384747473,
+      "learning_rate": 1.5825049701789263e-06,
+      "loss": 0.6901,
+      "step": 796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1670889986735915,
+      "learning_rate": 1.584493041749503e-06,
+      "loss": 0.696,
+      "step": 797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.243664722168212,
+      "learning_rate": 1.5864811133200795e-06,
+      "loss": 0.6966,
+      "step": 798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8149891677560364,
+      "learning_rate": 1.588469184890656e-06,
+      "loss": 0.6868,
+      "step": 799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.361473306979857,
+      "learning_rate": 1.5904572564612325e-06,
+      "loss": 0.7109,
+      "step": 800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.7540764958430755,
+      "learning_rate": 1.5924453280318091e-06,
+      "loss": 0.7064,
+      "step": 801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.040970476980475,
+      "learning_rate": 1.5944333996023855e-06,
+      "loss": 0.6842,
+      "step": 802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.697334296091878,
+      "learning_rate": 1.5964214711729622e-06,
+      "loss": 0.6914,
+      "step": 803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.822933270695639,
+      "learning_rate": 1.5984095427435388e-06,
+      "loss": 0.6953,
+      "step": 804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.12098867089369,
+      "learning_rate": 1.6003976143141152e-06,
+      "loss": 0.6849,
+      "step": 805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.04958074298262,
+      "learning_rate": 1.6023856858846918e-06,
+      "loss": 0.694,
+      "step": 806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.512802393091038,
+      "learning_rate": 1.6043737574552682e-06,
+      "loss": 0.6973,
+      "step": 807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.244398525200364,
+      "learning_rate": 1.606361829025845e-06,
+      "loss": 0.7057,
+      "step": 808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1680227943799606,
+      "learning_rate": 1.6083499005964214e-06,
+      "loss": 0.6908,
+      "step": 809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.571008011877753,
+      "learning_rate": 1.6103379721669978e-06,
+      "loss": 0.6979,
+      "step": 810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.487897550527417,
+      "learning_rate": 1.6123260437375746e-06,
+      "loss": 0.696,
+      "step": 811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.687701256891149,
+      "learning_rate": 1.614314115308151e-06,
+      "loss": 0.6927,
+      "step": 812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.203078280589674,
+      "learning_rate": 1.6163021868787276e-06,
+      "loss": 0.6953,
+      "step": 813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.111125466210192,
+      "learning_rate": 1.618290258449304e-06,
+      "loss": 0.6921,
+      "step": 814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.0845965452657556,
+      "learning_rate": 1.6202783300198806e-06,
+      "loss": 0.694,
+      "step": 815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.789169743675614,
+      "learning_rate": 1.6222664015904572e-06,
+      "loss": 0.6914,
+      "step": 816
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.168326435341886,
+      "learning_rate": 1.6242544731610336e-06,
+      "loss": 0.6992,
+      "step": 817
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.416312026385379,
+      "learning_rate": 1.6262425447316103e-06,
+      "loss": 0.6973,
+      "step": 818
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.2006073183154085,
+      "learning_rate": 1.6282306163021869e-06,
+      "loss": 0.696,
+      "step": 819
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.694974248897912,
+      "learning_rate": 1.6302186878727633e-06,
+      "loss": 0.6862,
+      "step": 820
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2342980818642535,
+      "learning_rate": 1.6322067594433399e-06,
+      "loss": 0.6895,
+      "step": 821
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.016232699283681,
+      "learning_rate": 1.6341948310139165e-06,
+      "loss": 0.6934,
+      "step": 822
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.480752754489313,
+      "learning_rate": 1.6361829025844929e-06,
+      "loss": 0.707,
+      "step": 823
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.888109401924429,
+      "learning_rate": 1.6381709741550695e-06,
+      "loss": 0.6855,
+      "step": 824
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.102245392615632,
+      "learning_rate": 1.640159045725646e-06,
+      "loss": 0.6921,
+      "step": 825
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.815973491929639,
+      "learning_rate": 1.6421471172962227e-06,
+      "loss": 0.6947,
+      "step": 826
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.241995965827673,
+      "learning_rate": 1.6441351888667991e-06,
+      "loss": 0.6947,
+      "step": 827
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.905531560431864,
+      "learning_rate": 1.6461232604373755e-06,
+      "loss": 0.6921,
+      "step": 828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.477817967263282,
+      "learning_rate": 1.6481113320079523e-06,
+      "loss": 0.6973,
+      "step": 829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4678022331061817,
+      "learning_rate": 1.6500994035785287e-06,
+      "loss": 0.6947,
+      "step": 830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.42406347542297,
+      "learning_rate": 1.6520874751491053e-06,
+      "loss": 0.6908,
+      "step": 831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.268178733701501,
+      "learning_rate": 1.654075546719682e-06,
+      "loss": 0.6875,
+      "step": 832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.6479181928502165,
+      "learning_rate": 1.6560636182902584e-06,
+      "loss": 0.696,
+      "step": 833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.95319217918017,
+      "learning_rate": 1.658051689860835e-06,
+      "loss": 0.6868,
+      "step": 834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7550869964533624,
+      "learning_rate": 1.6600397614314114e-06,
+      "loss": 0.6979,
+      "step": 835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.201782050412735,
+      "learning_rate": 1.662027833001988e-06,
+      "loss": 0.6986,
+      "step": 836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.098151704629643,
+      "learning_rate": 1.6640159045725646e-06,
+      "loss": 0.7129,
+      "step": 837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.8408204257313665,
+      "learning_rate": 1.666003976143141e-06,
+      "loss": 0.6992,
+      "step": 838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.288241059253356,
+      "learning_rate": 1.6679920477137178e-06,
+      "loss": 0.6823,
+      "step": 839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2146650115410673,
+      "learning_rate": 1.6699801192842942e-06,
+      "loss": 0.6908,
+      "step": 840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.166939611881107,
+      "learning_rate": 1.6719681908548706e-06,
+      "loss": 0.6888,
+      "step": 841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.183287797176236,
+      "learning_rate": 1.6739562624254472e-06,
+      "loss": 0.6921,
+      "step": 842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.112974310600647,
+      "learning_rate": 1.6759443339960238e-06,
+      "loss": 0.6953,
+      "step": 843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7483603402552927,
+      "learning_rate": 1.6779324055666004e-06,
+      "loss": 0.694,
+      "step": 844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.364137008373639,
+      "learning_rate": 1.6799204771371768e-06,
+      "loss": 0.6992,
+      "step": 845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.796934809294173,
+      "learning_rate": 1.6819085487077532e-06,
+      "loss": 0.6914,
+      "step": 846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.044946171310735,
+      "learning_rate": 1.68389662027833e-06,
+      "loss": 0.6953,
+      "step": 847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4287484493844365,
+      "learning_rate": 1.6858846918489065e-06,
+      "loss": 0.6901,
+      "step": 848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.8512671533024125,
+      "learning_rate": 1.6878727634194829e-06,
+      "loss": 0.6973,
+      "step": 849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.555857703472488,
+      "learning_rate": 1.6898608349900597e-06,
+      "loss": 0.6868,
+      "step": 850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.636012271866163,
+      "learning_rate": 1.691848906560636e-06,
+      "loss": 0.6882,
+      "step": 851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.14776603769053,
+      "learning_rate": 1.6938369781312127e-06,
+      "loss": 0.694,
+      "step": 852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.350572689704162,
+      "learning_rate": 1.6958250497017893e-06,
+      "loss": 0.6947,
+      "step": 853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.555264810453602,
+      "learning_rate": 1.6978131212723657e-06,
+      "loss": 0.694,
+      "step": 854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.376179555087542,
+      "learning_rate": 1.6998011928429423e-06,
+      "loss": 0.6973,
+      "step": 855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4504591297740483,
+      "learning_rate": 1.7017892644135187e-06,
+      "loss": 0.6927,
+      "step": 856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6490423543718726,
+      "learning_rate": 1.7037773359840955e-06,
+      "loss": 0.6914,
+      "step": 857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.958504608183791,
+      "learning_rate": 1.705765407554672e-06,
+      "loss": 0.7038,
+      "step": 858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4465968932213764,
+      "learning_rate": 1.7077534791252483e-06,
+      "loss": 0.6875,
+      "step": 859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.923110782232334,
+      "learning_rate": 1.7097415506958251e-06,
+      "loss": 0.6901,
+      "step": 860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.2555047403152715,
+      "learning_rate": 1.7117296222664015e-06,
+      "loss": 0.7103,
+      "step": 861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.344862543399948,
+      "learning_rate": 1.713717693836978e-06,
+      "loss": 0.6953,
+      "step": 862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.240120278013498,
+      "learning_rate": 1.7157057654075546e-06,
+      "loss": 0.6921,
+      "step": 863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.323403567178968,
+      "learning_rate": 1.7176938369781312e-06,
+      "loss": 0.6934,
+      "step": 864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.052378316269431,
+      "learning_rate": 1.7196819085487078e-06,
+      "loss": 0.6882,
+      "step": 865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.992688593094907,
+      "learning_rate": 1.7216699801192842e-06,
+      "loss": 0.6934,
+      "step": 866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.004078207652702,
+      "learning_rate": 1.7236580516898606e-06,
+      "loss": 0.6953,
+      "step": 867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.401299343905316,
+      "learning_rate": 1.7256461232604374e-06,
+      "loss": 0.6966,
+      "step": 868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7520273499640306,
+      "learning_rate": 1.7276341948310138e-06,
+      "loss": 0.6914,
+      "step": 869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.036626063259436,
+      "learning_rate": 1.7296222664015904e-06,
+      "loss": 0.6947,
+      "step": 870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7441984781279734,
+      "learning_rate": 1.731610337972167e-06,
+      "loss": 0.6979,
+      "step": 871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.30973965034451,
+      "learning_rate": 1.7335984095427434e-06,
+      "loss": 0.7077,
+      "step": 872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.486208730388135,
+      "learning_rate": 1.73558648111332e-06,
+      "loss": 0.6855,
+      "step": 873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.921968898442971,
+      "learning_rate": 1.7375745526838964e-06,
+      "loss": 0.6986,
+      "step": 874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.451367604930032,
+      "learning_rate": 1.739562624254473e-06,
+      "loss": 0.6908,
+      "step": 875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.27837154708852,
+      "learning_rate": 1.7415506958250496e-06,
+      "loss": 0.6888,
+      "step": 876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.84877394856048,
+      "learning_rate": 1.743538767395626e-06,
+      "loss": 0.7148,
+      "step": 877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.2508256863175795,
+      "learning_rate": 1.7455268389662029e-06,
+      "loss": 0.7025,
+      "step": 878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.933498956363293,
+      "learning_rate": 1.7475149105367793e-06,
+      "loss": 0.6953,
+      "step": 879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.585763802745018,
+      "learning_rate": 1.7495029821073557e-06,
+      "loss": 0.6888,
+      "step": 880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.650566717731353,
+      "learning_rate": 1.7514910536779323e-06,
+      "loss": 0.6992,
+      "step": 881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.547887536753185,
+      "learning_rate": 1.7534791252485089e-06,
+      "loss": 0.6882,
+      "step": 882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1463896645877167,
+      "learning_rate": 1.7554671968190855e-06,
+      "loss": 0.6966,
+      "step": 883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.145493115407018,
+      "learning_rate": 1.7574552683896619e-06,
+      "loss": 0.6927,
+      "step": 884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.704999594535566,
+      "learning_rate": 1.7594433399602385e-06,
+      "loss": 0.7025,
+      "step": 885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.443099532638297,
+      "learning_rate": 1.7614314115308151e-06,
+      "loss": 0.6953,
+      "step": 886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.801918790388745,
+      "learning_rate": 1.7634194831013915e-06,
+      "loss": 0.7005,
+      "step": 887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7597522323247743,
+      "learning_rate": 1.765407554671968e-06,
+      "loss": 0.6914,
+      "step": 888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7854203661462174,
+      "learning_rate": 1.7673956262425447e-06,
+      "loss": 0.6986,
+      "step": 889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.740268298448499,
+      "learning_rate": 1.7693836978131211e-06,
+      "loss": 0.6888,
+      "step": 890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.142980007679268,
+      "learning_rate": 1.7713717693836977e-06,
+      "loss": 0.6999,
+      "step": 891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3643598292335815,
+      "learning_rate": 1.7733598409542744e-06,
+      "loss": 0.6927,
+      "step": 892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.759985147997545,
+      "learning_rate": 1.7753479125248508e-06,
+      "loss": 0.7031,
+      "step": 893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.368191892480012,
+      "learning_rate": 1.7773359840954274e-06,
+      "loss": 0.6855,
+      "step": 894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 15.134420868032946,
+      "learning_rate": 1.7793240556660038e-06,
+      "loss": 0.6836,
+      "step": 895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.419566128256565,
+      "learning_rate": 1.7813121272365806e-06,
+      "loss": 0.6953,
+      "step": 896
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.023914620215725,
+      "learning_rate": 1.783300198807157e-06,
+      "loss": 0.707,
+      "step": 897
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4512502150893682,
+      "learning_rate": 1.7852882703777334e-06,
+      "loss": 0.6836,
+      "step": 898
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 19.233359757096487,
+      "learning_rate": 1.7872763419483102e-06,
+      "loss": 0.7181,
+      "step": 899
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.692481405932554,
+      "learning_rate": 1.7892644135188866e-06,
+      "loss": 0.7103,
+      "step": 900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.263017577929976,
+      "learning_rate": 1.7912524850894632e-06,
+      "loss": 0.6927,
+      "step": 901
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.595389245834784,
+      "learning_rate": 1.7932405566600396e-06,
+      "loss": 0.6979,
+      "step": 902
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.317935169854547,
+      "learning_rate": 1.7952286282306162e-06,
+      "loss": 0.6934,
+      "step": 903
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.7791324390880465,
+      "learning_rate": 1.7972166998011928e-06,
+      "loss": 0.6986,
+      "step": 904
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.888575927155742,
+      "learning_rate": 1.7992047713717692e-06,
+      "loss": 0.7096,
+      "step": 905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.685215470306302,
+      "learning_rate": 1.8011928429423458e-06,
+      "loss": 0.7031,
+      "step": 906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.113720243186466,
+      "learning_rate": 1.8031809145129225e-06,
+      "loss": 0.6901,
+      "step": 907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.377371280473697,
+      "learning_rate": 1.8051689860834989e-06,
+      "loss": 0.6966,
+      "step": 908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.664667620433667,
+      "learning_rate": 1.8071570576540755e-06,
+      "loss": 0.6966,
+      "step": 909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7331593395876153,
+      "learning_rate": 1.809145129224652e-06,
+      "loss": 0.6927,
+      "step": 910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.53706180788123,
+      "learning_rate": 1.8111332007952285e-06,
+      "loss": 0.7038,
+      "step": 911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9243559893004265,
+      "learning_rate": 1.813121272365805e-06,
+      "loss": 0.6921,
+      "step": 912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5873251473390653,
+      "learning_rate": 1.8151093439363817e-06,
+      "loss": 0.694,
+      "step": 913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.688911650515929,
+      "learning_rate": 1.8170974155069583e-06,
+      "loss": 0.6979,
+      "step": 914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.297208287957993,
+      "learning_rate": 1.8190854870775347e-06,
+      "loss": 0.6901,
+      "step": 915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 15.505636881497116,
+      "learning_rate": 1.821073558648111e-06,
+      "loss": 0.7038,
+      "step": 916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.378347719476801,
+      "learning_rate": 1.823061630218688e-06,
+      "loss": 0.6934,
+      "step": 917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.937263683087334,
+      "learning_rate": 1.8250497017892643e-06,
+      "loss": 0.6953,
+      "step": 918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3795973449589685,
+      "learning_rate": 1.8270377733598407e-06,
+      "loss": 0.6888,
+      "step": 919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9109049499024344,
+      "learning_rate": 1.8290258449304175e-06,
+      "loss": 0.6953,
+      "step": 920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.387039622736997,
+      "learning_rate": 1.831013916500994e-06,
+      "loss": 0.6914,
+      "step": 921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7285302974104395,
+      "learning_rate": 1.8330019880715706e-06,
+      "loss": 0.7005,
+      "step": 922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9697974011415194,
+      "learning_rate": 1.834990059642147e-06,
+      "loss": 0.6934,
+      "step": 923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.209383703613921,
+      "learning_rate": 1.8369781312127236e-06,
+      "loss": 0.6934,
+      "step": 924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.480493816172714,
+      "learning_rate": 1.8389662027833002e-06,
+      "loss": 0.7005,
+      "step": 925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.56786775133432,
+      "learning_rate": 1.8409542743538766e-06,
+      "loss": 0.696,
+      "step": 926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.446553145302807,
+      "learning_rate": 1.8429423459244534e-06,
+      "loss": 0.6934,
+      "step": 927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.191238875811201,
+      "learning_rate": 1.8449304174950298e-06,
+      "loss": 0.6947,
+      "step": 928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.729379240651248,
+      "learning_rate": 1.8469184890656062e-06,
+      "loss": 0.6973,
+      "step": 929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.338064752072578,
+      "learning_rate": 1.8489065606361828e-06,
+      "loss": 0.6921,
+      "step": 930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.823985415617739,
+      "learning_rate": 1.8508946322067594e-06,
+      "loss": 0.6901,
+      "step": 931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0129590782576803,
+      "learning_rate": 1.8528827037773358e-06,
+      "loss": 0.6914,
+      "step": 932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.815092061487661,
+      "learning_rate": 1.8548707753479124e-06,
+      "loss": 0.6914,
+      "step": 933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.107460013451897,
+      "learning_rate": 1.856858846918489e-06,
+      "loss": 0.6947,
+      "step": 934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.324410589756834,
+      "learning_rate": 1.8588469184890656e-06,
+      "loss": 0.7083,
+      "step": 935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.907670997978956,
+      "learning_rate": 1.860834990059642e-06,
+      "loss": 0.6999,
+      "step": 936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.107926851275585,
+      "learning_rate": 1.8628230616302184e-06,
+      "loss": 0.696,
+      "step": 937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.211790977100659,
+      "learning_rate": 1.8648111332007953e-06,
+      "loss": 0.7044,
+      "step": 938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.212009696072005,
+      "learning_rate": 1.8667992047713717e-06,
+      "loss": 0.6966,
+      "step": 939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.108176027758403,
+      "learning_rate": 1.8687872763419483e-06,
+      "loss": 0.6895,
+      "step": 940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.452140350553258,
+      "learning_rate": 1.8707753479125249e-06,
+      "loss": 0.6836,
+      "step": 941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1429591141635713,
+      "learning_rate": 1.8727634194831013e-06,
+      "loss": 0.6855,
+      "step": 942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 26.96665132527698,
+      "learning_rate": 1.8747514910536779e-06,
+      "loss": 0.7324,
+      "step": 943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7487982317642337,
+      "learning_rate": 1.8767395626242543e-06,
+      "loss": 0.6836,
+      "step": 944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.673817536790449,
+      "learning_rate": 1.878727634194831e-06,
+      "loss": 0.6973,
+      "step": 945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.138354063842728,
+      "learning_rate": 1.8807157057654075e-06,
+      "loss": 0.6999,
+      "step": 946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.400058315331106,
+      "learning_rate": 1.882703777335984e-06,
+      "loss": 0.694,
+      "step": 947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.939466441615692,
+      "learning_rate": 1.8846918489065607e-06,
+      "loss": 0.6979,
+      "step": 948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.454498584064968,
+      "learning_rate": 1.8866799204771371e-06,
+      "loss": 0.696,
+      "step": 949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.0852354280009,
+      "learning_rate": 1.8886679920477135e-06,
+      "loss": 0.7031,
+      "step": 950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.316087498116654,
+      "learning_rate": 1.8906560636182901e-06,
+      "loss": 0.6868,
+      "step": 951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4427317234696493,
+      "learning_rate": 1.8926441351888668e-06,
+      "loss": 0.6803,
+      "step": 952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.32645785235573,
+      "learning_rate": 1.8946322067594434e-06,
+      "loss": 0.6908,
+      "step": 953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 14.28341041332783,
+      "learning_rate": 1.8966202783300198e-06,
+      "loss": 0.7129,
+      "step": 954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.053775899630727,
+      "learning_rate": 1.8986083499005962e-06,
+      "loss": 0.7018,
+      "step": 955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.654887654306456,
+      "learning_rate": 1.900596421471173e-06,
+      "loss": 0.6986,
+      "step": 956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.78805953649366,
+      "learning_rate": 1.9025844930417494e-06,
+      "loss": 0.7057,
+      "step": 957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.101018212950043,
+      "learning_rate": 1.9045725646123258e-06,
+      "loss": 0.696,
+      "step": 958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.510518061211453,
+      "learning_rate": 1.9065606361829026e-06,
+      "loss": 0.6875,
+      "step": 959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5667045304070333,
+      "learning_rate": 1.908548707753479e-06,
+      "loss": 0.6934,
+      "step": 960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.668021883283828,
+      "learning_rate": 1.9105367793240556e-06,
+      "loss": 0.7096,
+      "step": 961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.995754625538203,
+      "learning_rate": 1.912524850894632e-06,
+      "loss": 0.7005,
+      "step": 962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.505631905080863,
+      "learning_rate": 1.9145129224652084e-06,
+      "loss": 0.694,
+      "step": 963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.484901594098195,
+      "learning_rate": 1.9165009940357852e-06,
+      "loss": 0.7233,
+      "step": 964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.698380885789234,
+      "learning_rate": 1.9184890656063616e-06,
+      "loss": 0.6999,
+      "step": 965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.866185858593128,
+      "learning_rate": 1.9204771371769385e-06,
+      "loss": 0.6999,
+      "step": 966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8943103297600585,
+      "learning_rate": 1.922465208747515e-06,
+      "loss": 0.6927,
+      "step": 967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.359312904663794,
+      "learning_rate": 1.9244532803180912e-06,
+      "loss": 0.6934,
+      "step": 968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.780516317145175,
+      "learning_rate": 1.926441351888668e-06,
+      "loss": 0.6855,
+      "step": 969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.458153069959146,
+      "learning_rate": 1.9284294234592445e-06,
+      "loss": 0.6888,
+      "step": 970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.989755080349235,
+      "learning_rate": 1.9304174950298213e-06,
+      "loss": 0.6849,
+      "step": 971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.66016755641419,
+      "learning_rate": 1.9324055666003977e-06,
+      "loss": 0.7174,
+      "step": 972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.67117899392022,
+      "learning_rate": 1.934393638170974e-06,
+      "loss": 0.7493,
+      "step": 973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.489191144365526,
+      "learning_rate": 1.9363817097415505e-06,
+      "loss": 0.6875,
+      "step": 974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.807026818030657,
+      "learning_rate": 1.9383697813121273e-06,
+      "loss": 0.6953,
+      "step": 975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.632523198503545,
+      "learning_rate": 1.9403578528827037e-06,
+      "loss": 0.7109,
+      "step": 976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.548432584874475,
+      "learning_rate": 1.94234592445328e-06,
+      "loss": 0.6921,
+      "step": 977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2076861200367723,
+      "learning_rate": 1.944333996023857e-06,
+      "loss": 0.6829,
+      "step": 978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.859934419987951,
+      "learning_rate": 1.9463220675944333e-06,
+      "loss": 0.696,
+      "step": 979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.023774474510033,
+      "learning_rate": 1.9483101391650097e-06,
+      "loss": 0.7051,
+      "step": 980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.610599491642174,
+      "learning_rate": 1.950298210735586e-06,
+      "loss": 0.7005,
+      "step": 981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.895048692446782,
+      "learning_rate": 1.952286282306163e-06,
+      "loss": 0.6882,
+      "step": 982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.454060064388314,
+      "learning_rate": 1.9542743538767393e-06,
+      "loss": 0.7031,
+      "step": 983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.82195238407596,
+      "learning_rate": 1.956262425447316e-06,
+      "loss": 0.7012,
+      "step": 984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.519786887660453,
+      "learning_rate": 1.9582504970178926e-06,
+      "loss": 0.7188,
+      "step": 985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.123442382374296,
+      "learning_rate": 1.960238568588469e-06,
+      "loss": 0.7012,
+      "step": 986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.867602290505366,
+      "learning_rate": 1.962226640159046e-06,
+      "loss": 0.6934,
+      "step": 987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.752153268937873,
+      "learning_rate": 1.964214711729622e-06,
+      "loss": 0.6816,
+      "step": 988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.484104747690235,
+      "learning_rate": 1.9662027833001986e-06,
+      "loss": 0.7005,
+      "step": 989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0951532231199543,
+      "learning_rate": 1.9681908548707754e-06,
+      "loss": 0.6914,
+      "step": 990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.084750556866899,
+      "learning_rate": 1.970178926441352e-06,
+      "loss": 0.7148,
+      "step": 991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.394004964666189,
+      "learning_rate": 1.9721669980119286e-06,
+      "loss": 0.6966,
+      "step": 992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.25705084846513,
+      "learning_rate": 1.974155069582505e-06,
+      "loss": 0.7077,
+      "step": 993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.208039831303938,
+      "learning_rate": 1.9761431411530814e-06,
+      "loss": 0.6927,
+      "step": 994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.647153453978154,
+      "learning_rate": 1.978131212723658e-06,
+      "loss": 0.6953,
+      "step": 995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.4066691627509265,
+      "learning_rate": 1.9801192842942347e-06,
+      "loss": 0.6986,
+      "step": 996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7467144898241598,
+      "learning_rate": 1.982107355864811e-06,
+      "loss": 0.6934,
+      "step": 997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.465321296554912,
+      "learning_rate": 1.9840954274353874e-06,
+      "loss": 0.7025,
+      "step": 998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9380069431693756,
+      "learning_rate": 1.986083499005964e-06,
+      "loss": 0.6927,
+      "step": 999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.964899707486991,
+      "learning_rate": 1.9880715705765407e-06,
+      "loss": 0.6966,
+      "step": 1000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.702601292751394,
+      "learning_rate": 1.990059642147117e-06,
+      "loss": 0.7044,
+      "step": 1001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.033127393068792,
+      "learning_rate": 1.9920477137176935e-06,
+      "loss": 0.7005,
+      "step": 1002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3626985941966243,
+      "learning_rate": 1.9940357852882703e-06,
+      "loss": 0.6934,
+      "step": 1003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.186095863539595,
+      "learning_rate": 1.9960238568588467e-06,
+      "loss": 0.6895,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.225176223828071,
+      "learning_rate": 1.9980119284294235e-06,
+      "loss": 0.6934,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8292797066799271,
+      "learning_rate": 2e-06,
+      "loss": 0.6934,
+      "step": 1006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.06997557449564,
+      "learning_rate": 1.9999999953334557e-06,
+      "loss": 0.6999,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.801547040411194,
+      "learning_rate": 1.999999981333823e-06,
+      "loss": 0.7012,
+      "step": 1008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5371437373558954,
+      "learning_rate": 1.9999999580011024e-06,
+      "loss": 0.6914,
+      "step": 1009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.366782017762354,
+      "learning_rate": 1.9999999253352933e-06,
+      "loss": 0.696,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.421677210542779,
+      "learning_rate": 1.9999998833363966e-06,
+      "loss": 0.6979,
+      "step": 1011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.800402821991273,
+      "learning_rate": 1.9999998320044128e-06,
+      "loss": 0.6875,
+      "step": 1012
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.209937443209005,
+      "learning_rate": 1.9999997713393418e-06,
+      "loss": 0.6829,
+      "step": 1013
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.431312669822333,
+      "learning_rate": 1.9999997013411844e-06,
+      "loss": 0.7083,
+      "step": 1014
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.3487598605755435,
+      "learning_rate": 1.9999996220099416e-06,
+      "loss": 0.6992,
+      "step": 1015
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.040252050926318,
+      "learning_rate": 1.9999995333456142e-06,
+      "loss": 0.7083,
+      "step": 1016
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.569140077976554,
+      "learning_rate": 1.999999435348202e-06,
+      "loss": 0.7142,
+      "step": 1017
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.4251909418188236,
+      "learning_rate": 1.9999993280177072e-06,
+      "loss": 0.6947,
+      "step": 1018
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.986405164623672,
+      "learning_rate": 1.99999921135413e-06,
+      "loss": 0.6908,
+      "step": 1019
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.982672591336462,
+      "learning_rate": 1.999999085357472e-06,
+      "loss": 0.6882,
+      "step": 1020
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.741972701781011,
+      "learning_rate": 1.999998950027734e-06,
+      "loss": 0.6908,
+      "step": 1021
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.683610942511333,
+      "learning_rate": 1.999998805364917e-06,
+      "loss": 0.6908,
+      "step": 1022
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.089404390858266,
+      "learning_rate": 1.9999986513690232e-06,
+      "loss": 0.6953,
+      "step": 1023
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.312837301763771,
+      "learning_rate": 1.999998488040053e-06,
+      "loss": 0.6875,
+      "step": 1024
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.3967329862305,
+      "learning_rate": 1.999998315378009e-06,
+      "loss": 0.6921,
+      "step": 1025
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.6675648220760895,
+      "learning_rate": 1.9999981333828917e-06,
+      "loss": 0.6934,
+      "step": 1026
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.572276601879375,
+      "learning_rate": 1.999997942054704e-06,
+      "loss": 0.6973,
+      "step": 1027
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.021227842198956,
+      "learning_rate": 1.999997741393447e-06,
+      "loss": 0.7031,
+      "step": 1028
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7073995141668146,
+      "learning_rate": 1.9999975313991223e-06,
+      "loss": 0.6855,
+      "step": 1029
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7707845282187926,
+      "learning_rate": 1.9999973120717326e-06,
+      "loss": 0.6992,
+      "step": 1030
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.712513347887992,
+      "learning_rate": 1.999997083411279e-06,
+      "loss": 0.7077,
+      "step": 1031
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6898046425554734,
+      "learning_rate": 1.999996845417765e-06,
+      "loss": 0.6947,
+      "step": 1032
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.472729570775615,
+      "learning_rate": 1.9999965980911916e-06,
+      "loss": 0.6953,
+      "step": 1033
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.119011581176359,
+      "learning_rate": 1.9999963414315613e-06,
+      "loss": 0.7018,
+      "step": 1034
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.342953698051504,
+      "learning_rate": 1.9999960754388767e-06,
+      "loss": 0.6986,
+      "step": 1035
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.373106531205004,
+      "learning_rate": 1.9999958001131407e-06,
+      "loss": 0.6908,
+      "step": 1036
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.95118280892305,
+      "learning_rate": 1.999995515454355e-06,
+      "loss": 0.6855,
+      "step": 1037
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.828028171833577,
+      "learning_rate": 1.999995221462523e-06,
+      "loss": 0.6797,
+      "step": 1038
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8056218497854815,
+      "learning_rate": 1.9999949181376474e-06,
+      "loss": 0.6901,
+      "step": 1039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.676499712737745,
+      "learning_rate": 1.9999946054797304e-06,
+      "loss": 0.6999,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1378962927328473,
+      "learning_rate": 1.999994283488775e-06,
+      "loss": 0.6979,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.486372586674187,
+      "learning_rate": 1.999993952164785e-06,
+      "loss": 0.6908,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.550477579060876,
+      "learning_rate": 1.999993611507763e-06,
+      "loss": 0.709,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.115908029392567,
+      "learning_rate": 1.9999932615177124e-06,
+      "loss": 0.696,
+      "step": 1044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.49215764946554,
+      "learning_rate": 1.9999929021946357e-06,
+      "loss": 0.6966,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7341376696346615,
+      "learning_rate": 1.999992533538537e-06,
+      "loss": 0.6855,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.702693218827038,
+      "learning_rate": 1.99999215554942e-06,
+      "loss": 0.7025,
+      "step": 1047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1422197661104967,
+      "learning_rate": 1.999991768227287e-06,
+      "loss": 0.6979,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.599565008256906,
+      "learning_rate": 1.999991371572143e-06,
+      "loss": 0.6849,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.131311747800264,
+      "learning_rate": 1.9999909655839907e-06,
+      "loss": 0.6973,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.460783430822108,
+      "learning_rate": 1.999990550262835e-06,
+      "loss": 0.6875,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1547616199794746,
+      "learning_rate": 1.9999901256086783e-06,
+      "loss": 0.7044,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0733281893798976,
+      "learning_rate": 1.9999896916215256e-06,
+      "loss": 0.6947,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0330888037459063,
+      "learning_rate": 1.9999892483013805e-06,
+      "loss": 0.6849,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.207153915036544,
+      "learning_rate": 1.9999887956482475e-06,
+      "loss": 0.6921,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.936306034891713,
+      "learning_rate": 1.9999883336621306e-06,
+      "loss": 0.7057,
+      "step": 1056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3574005347310862,
+      "learning_rate": 1.999987862343034e-06,
+      "loss": 0.6914,
+      "step": 1057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6197703033041064,
+      "learning_rate": 1.9999873816909622e-06,
+      "loss": 0.6921,
+      "step": 1058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.946024696556024,
+      "learning_rate": 1.99998689170592e-06,
+      "loss": 0.6973,
+      "step": 1059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.125268425717447,
+      "learning_rate": 1.9999863923879115e-06,
+      "loss": 0.6829,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.599275111520482,
+      "learning_rate": 1.999985883736942e-06,
+      "loss": 0.6849,
+      "step": 1061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.340330018109908,
+      "learning_rate": 1.9999853657530153e-06,
+      "loss": 0.6862,
+      "step": 1062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.190436644698012,
+      "learning_rate": 1.999984838436137e-06,
+      "loss": 0.6973,
+      "step": 1063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3648404494239506,
+      "learning_rate": 1.9999843017863117e-06,
+      "loss": 0.6868,
+      "step": 1064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8906334931854407,
+      "learning_rate": 1.9999837558035445e-06,
+      "loss": 0.6914,
+      "step": 1065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.960138270427374,
+      "learning_rate": 1.9999832004878405e-06,
+      "loss": 0.6771,
+      "step": 1066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6950856043915379,
+      "learning_rate": 1.999982635839205e-06,
+      "loss": 0.6914,
+      "step": 1067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4288420811769913,
+      "learning_rate": 1.999982061857643e-06,
+      "loss": 0.6914,
+      "step": 1068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.3267676124202,
+      "learning_rate": 1.9999814785431597e-06,
+      "loss": 0.7057,
+      "step": 1069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.997122500887164,
+      "learning_rate": 1.999980885895761e-06,
+      "loss": 0.7077,
+      "step": 1070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.355467509040908,
+      "learning_rate": 1.999980283915452e-06,
+      "loss": 0.6953,
+      "step": 1071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.468938521766761,
+      "learning_rate": 1.9999796726022394e-06,
+      "loss": 0.7135,
+      "step": 1072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.092185102579719,
+      "learning_rate": 1.9999790519561277e-06,
+      "loss": 0.7057,
+      "step": 1073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.65412007035965,
+      "learning_rate": 1.9999784219771227e-06,
+      "loss": 0.6953,
+      "step": 1074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.355596414639488,
+      "learning_rate": 1.999977782665231e-06,
+      "loss": 0.6921,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 21.166281625965155,
+      "learning_rate": 1.999977134020458e-06,
+      "loss": 0.7298,
+      "step": 1076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.461406970761206,
+      "learning_rate": 1.99997647604281e-06,
+      "loss": 0.6908,
+      "step": 1077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.713121176123474,
+      "learning_rate": 1.9999758087322933e-06,
+      "loss": 0.7214,
+      "step": 1078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.33504300089685,
+      "learning_rate": 1.999975132088914e-06,
+      "loss": 0.7181,
+      "step": 1079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.85947937926659,
+      "learning_rate": 1.999974446112678e-06,
+      "loss": 0.6979,
+      "step": 1080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.244352018271897,
+      "learning_rate": 1.999973750803592e-06,
+      "loss": 0.6868,
+      "step": 1081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.763459526534081,
+      "learning_rate": 1.999973046161663e-06,
+      "loss": 0.7038,
+      "step": 1082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 15.485122607292187,
+      "learning_rate": 1.999972332186897e-06,
+      "loss": 0.7051,
+      "step": 1083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.207732589726451,
+      "learning_rate": 1.999971608879301e-06,
+      "loss": 0.6875,
+      "step": 1084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.554304353470423,
+      "learning_rate": 1.9999708762388812e-06,
+      "loss": 0.7031,
+      "step": 1085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2008074637578723,
+      "learning_rate": 1.9999701342656446e-06,
+      "loss": 0.6992,
+      "step": 1086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.565145568102359,
+      "learning_rate": 1.9999693829595986e-06,
+      "loss": 0.6953,
+      "step": 1087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.929480238599367,
+      "learning_rate": 1.99996862232075e-06,
+      "loss": 0.6947,
+      "step": 1088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4340590630972843,
+      "learning_rate": 1.9999678523491057e-06,
+      "loss": 0.6908,
+      "step": 1089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8052291658461228,
+      "learning_rate": 1.999967073044673e-06,
+      "loss": 0.681,
+      "step": 1090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.176593627051166,
+      "learning_rate": 1.9999662844074593e-06,
+      "loss": 0.6888,
+      "step": 1091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3341003354068368,
+      "learning_rate": 1.9999654864374716e-06,
+      "loss": 0.6934,
+      "step": 1092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.153145981076235,
+      "learning_rate": 1.9999646791347174e-06,
+      "loss": 0.6947,
+      "step": 1093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3625530243765125,
+      "learning_rate": 1.999963862499205e-06,
+      "loss": 0.6947,
+      "step": 1094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.237755900420671,
+      "learning_rate": 1.999963036530941e-06,
+      "loss": 0.6849,
+      "step": 1095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.879942023361139,
+      "learning_rate": 1.9999622012299338e-06,
+      "loss": 0.694,
+      "step": 1096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9494018016163188,
+      "learning_rate": 1.999961356596191e-06,
+      "loss": 0.6921,
+      "step": 1097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.105014001298896,
+      "learning_rate": 1.9999605026297203e-06,
+      "loss": 0.6888,
+      "step": 1098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0583132346450617,
+      "learning_rate": 1.9999596393305298e-06,
+      "loss": 0.6992,
+      "step": 1099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1082907500851253,
+      "learning_rate": 1.9999587666986273e-06,
+      "loss": 0.6921,
+      "step": 1100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.592443649518697,
+      "learning_rate": 1.9999578847340214e-06,
+      "loss": 0.6966,
+      "step": 1101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.293147857952295,
+      "learning_rate": 1.99995699343672e-06,
+      "loss": 0.6953,
+      "step": 1102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.073635242636787,
+      "learning_rate": 1.999956092806732e-06,
+      "loss": 0.6947,
+      "step": 1103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.273002994714191,
+      "learning_rate": 1.999955182844065e-06,
+      "loss": 0.6927,
+      "step": 1104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.120205948984377,
+      "learning_rate": 1.999954263548728e-06,
+      "loss": 0.6882,
+      "step": 1105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2828269188515153,
+      "learning_rate": 1.9999533349207293e-06,
+      "loss": 0.6986,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6669140192811063,
+      "learning_rate": 1.9999523969600776e-06,
+      "loss": 0.694,
+      "step": 1107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.405711319125375,
+      "learning_rate": 1.9999514496667817e-06,
+      "loss": 0.6862,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.881777949832793,
+      "learning_rate": 1.999950493040851e-06,
+      "loss": 0.6868,
+      "step": 1109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.41491955023962,
+      "learning_rate": 1.9999495270822934e-06,
+      "loss": 0.6908,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.891205228384799,
+      "learning_rate": 1.9999485517911183e-06,
+      "loss": 0.6947,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.03462738798572,
+      "learning_rate": 1.9999475671673356e-06,
+      "loss": 0.6927,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.113472105657358,
+      "learning_rate": 1.999946573210953e-06,
+      "loss": 0.6895,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.240797708231301,
+      "learning_rate": 1.9999455699219815e-06,
+      "loss": 0.6895,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.769018044821202,
+      "learning_rate": 1.999944557300429e-06,
+      "loss": 0.6908,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.28548763203119,
+      "learning_rate": 1.9999435353463058e-06,
+      "loss": 0.6927,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.5960387885094,
+      "learning_rate": 1.999942504059621e-06,
+      "loss": 0.7012,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9639005003594,
+      "learning_rate": 1.9999414634403846e-06,
+      "loss": 0.6927,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.268709356194492,
+      "learning_rate": 1.9999404134886055e-06,
+      "loss": 0.6947,
+      "step": 1119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.442407388119941,
+      "learning_rate": 1.9999393542042945e-06,
+      "loss": 0.6868,
+      "step": 1120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.09281745482155,
+      "learning_rate": 1.999938285587461e-06,
+      "loss": 0.6927,
+      "step": 1121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.592562370312127,
+      "learning_rate": 1.9999372076381153e-06,
+      "loss": 0.7044,
+      "step": 1122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.292267745933659,
+      "learning_rate": 1.999936120356267e-06,
+      "loss": 0.6966,
+      "step": 1123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9324335616958215,
+      "learning_rate": 1.9999350237419265e-06,
+      "loss": 0.7031,
+      "step": 1124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9216260853603027,
+      "learning_rate": 1.9999339177951037e-06,
+      "loss": 0.7044,
+      "step": 1125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.065565419999506,
+      "learning_rate": 1.9999328025158095e-06,
+      "loss": 0.6921,
+      "step": 1126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.32093635421495,
+      "learning_rate": 1.999931677904054e-06,
+      "loss": 0.7077,
+      "step": 1127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7266447103548401,
+      "learning_rate": 1.9999305439598476e-06,
+      "loss": 0.6868,
+      "step": 1128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6201452676479646,
+      "learning_rate": 1.999929400683201e-06,
+      "loss": 0.6973,
+      "step": 1129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.693128303337555,
+      "learning_rate": 1.9999282480741252e-06,
+      "loss": 0.6953,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.463308825226945,
+      "learning_rate": 1.9999270861326304e-06,
+      "loss": 0.6947,
+      "step": 1131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.636111441031145,
+      "learning_rate": 1.999925914858728e-06,
+      "loss": 0.6842,
+      "step": 1132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.622987484593875,
+      "learning_rate": 1.999924734252428e-06,
+      "loss": 0.6999,
+      "step": 1133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.373111946928275,
+      "learning_rate": 1.9999235443137422e-06,
+      "loss": 0.6986,
+      "step": 1134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.308617443775636,
+      "learning_rate": 1.9999223450426817e-06,
+      "loss": 0.6849,
+      "step": 1135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.409164647338666,
+      "learning_rate": 1.9999211364392574e-06,
+      "loss": 0.6973,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.123339406931054,
+      "learning_rate": 1.9999199185034808e-06,
+      "loss": 0.694,
+      "step": 1137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3103314446455516,
+      "learning_rate": 1.999918691235363e-06,
+      "loss": 0.6947,
+      "step": 1138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7993947725072954,
+      "learning_rate": 1.999917454634916e-06,
+      "loss": 0.6921,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.866112368596856,
+      "learning_rate": 1.999916208702151e-06,
+      "loss": 0.6921,
+      "step": 1140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5800756497612052,
+      "learning_rate": 1.999914953437079e-06,
+      "loss": 0.6947,
+      "step": 1141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2132383305480718,
+      "learning_rate": 1.999913688839713e-06,
+      "loss": 0.6908,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.867811737760581,
+      "learning_rate": 1.9999124149100637e-06,
+      "loss": 0.6927,
+      "step": 1143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.224760080111763,
+      "learning_rate": 1.999911131648143e-06,
+      "loss": 0.6966,
+      "step": 1144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4555319442809755,
+      "learning_rate": 1.9999098390539644e-06,
+      "loss": 0.6849,
+      "step": 1145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8688398326297655,
+      "learning_rate": 1.9999085371275382e-06,
+      "loss": 0.6895,
+      "step": 1146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.692880023560281,
+      "learning_rate": 1.999907225868877e-06,
+      "loss": 0.6901,
+      "step": 1147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.984078387968019,
+      "learning_rate": 1.9999059052779935e-06,
+      "loss": 0.6973,
+      "step": 1148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.70640874704877,
+      "learning_rate": 1.9999045753549e-06,
+      "loss": 0.6973,
+      "step": 1149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.95764234206401,
+      "learning_rate": 1.999903236099608e-06,
+      "loss": 0.6855,
+      "step": 1150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.626211123211836,
+      "learning_rate": 1.9999018875121312e-06,
+      "loss": 0.6901,
+      "step": 1151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.132150016883938,
+      "learning_rate": 1.9999005295924816e-06,
+      "loss": 0.7025,
+      "step": 1152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.5792927725200485,
+      "learning_rate": 1.999899162340672e-06,
+      "loss": 0.696,
+      "step": 1153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.027206116333765,
+      "learning_rate": 1.9998977857567147e-06,
+      "loss": 0.6914,
+      "step": 1154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1241899961615154,
+      "learning_rate": 1.9998963998406233e-06,
+      "loss": 0.7038,
+      "step": 1155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.070500594075202,
+      "learning_rate": 1.99989500459241e-06,
+      "loss": 0.6914,
+      "step": 1156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.405804580580018,
+      "learning_rate": 1.9998936000120886e-06,
+      "loss": 0.6849,
+      "step": 1157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.202190728814944,
+      "learning_rate": 1.9998921860996716e-06,
+      "loss": 0.7044,
+      "step": 1158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.111036619756035,
+      "learning_rate": 1.9998907628551724e-06,
+      "loss": 0.7051,
+      "step": 1159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.926859624173719,
+      "learning_rate": 1.9998893302786045e-06,
+      "loss": 0.7005,
+      "step": 1160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8749291842652838,
+      "learning_rate": 1.999887888369981e-06,
+      "loss": 0.6895,
+      "step": 1161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2675493119964891,
+      "learning_rate": 1.9998864371293148e-06,
+      "loss": 0.6823,
+      "step": 1162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3377159721767071,
+      "learning_rate": 1.9998849765566207e-06,
+      "loss": 0.6882,
+      "step": 1163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.4495386689291045,
+      "learning_rate": 1.9998835066519115e-06,
+      "loss": 0.6823,
+      "step": 1164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.5099027131317975,
+      "learning_rate": 1.999882027415201e-06,
+      "loss": 0.694,
+      "step": 1165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.148281112362167,
+      "learning_rate": 1.9998805388465036e-06,
+      "loss": 0.7051,
+      "step": 1166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.148400739736749,
+      "learning_rate": 1.9998790409458325e-06,
+      "loss": 0.6947,
+      "step": 1167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.438837741029189,
+      "learning_rate": 1.9998775337132016e-06,
+      "loss": 0.6849,
+      "step": 1168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1116378917383614,
+      "learning_rate": 1.9998760171486253e-06,
+      "loss": 0.6888,
+      "step": 1169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.100321719508706,
+      "learning_rate": 1.9998744912521177e-06,
+      "loss": 0.6966,
+      "step": 1170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2602714523163256,
+      "learning_rate": 1.9998729560236935e-06,
+      "loss": 0.6797,
+      "step": 1171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.247934959260723,
+      "learning_rate": 1.9998714114633663e-06,
+      "loss": 0.7031,
+      "step": 1172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.6569328650555235,
+      "learning_rate": 1.9998698575711504e-06,
+      "loss": 0.6797,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6773676688478627,
+      "learning_rate": 1.9998682943470612e-06,
+      "loss": 0.6875,
+      "step": 1174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.532692883453713,
+      "learning_rate": 1.9998667217911127e-06,
+      "loss": 0.694,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4877735252505908,
+      "learning_rate": 1.999865139903319e-06,
+      "loss": 0.6829,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.121436558830522,
+      "learning_rate": 1.9998635486836966e-06,
+      "loss": 0.696,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.501992168468318,
+      "learning_rate": 1.9998619481322584e-06,
+      "loss": 0.6953,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.12063662584594,
+      "learning_rate": 1.9998603382490205e-06,
+      "loss": 0.6927,
+      "step": 1179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.729865536161565,
+      "learning_rate": 1.9998587190339978e-06,
+      "loss": 0.6868,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2984276017212013,
+      "learning_rate": 1.9998570904872054e-06,
+      "loss": 0.6849,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8889990216551422,
+      "learning_rate": 1.9998554526086583e-06,
+      "loss": 0.6973,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5977493647859826,
+      "learning_rate": 1.999853805398371e-06,
+      "loss": 0.694,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.713306461904431,
+      "learning_rate": 1.999852148856361e-06,
+      "loss": 0.696,
+      "step": 1184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4208176163483734,
+      "learning_rate": 1.9998504829826415e-06,
+      "loss": 0.6816,
+      "step": 1185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.667434111258073,
+      "learning_rate": 1.9998488077772294e-06,
+      "loss": 0.6992,
+      "step": 1186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.387019295765688,
+      "learning_rate": 1.99984712324014e-06,
+      "loss": 0.6875,
+      "step": 1187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.222444778071212,
+      "learning_rate": 1.999845429371389e-06,
+      "loss": 0.6908,
+      "step": 1188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.552313858954987,
+      "learning_rate": 1.999843726170992e-06,
+      "loss": 0.6966,
+      "step": 1189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9598457029092704,
+      "learning_rate": 1.9998420136389654e-06,
+      "loss": 0.6914,
+      "step": 1190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8769262919545064,
+      "learning_rate": 1.9998402917753245e-06,
+      "loss": 0.6979,
+      "step": 1191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.458183403288928,
+      "learning_rate": 1.999838560580086e-06,
+      "loss": 0.709,
+      "step": 1192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.689946583403858,
+      "learning_rate": 1.9998368200532657e-06,
+      "loss": 0.7135,
+      "step": 1193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3637289494686995,
+      "learning_rate": 1.9998350701948797e-06,
+      "loss": 0.6868,
+      "step": 1194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.772905123461479,
+      "learning_rate": 1.9998333110049447e-06,
+      "loss": 0.7122,
+      "step": 1195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.066918958811793,
+      "learning_rate": 1.9998315424834773e-06,
+      "loss": 0.6901,
+      "step": 1196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6406991957509494,
+      "learning_rate": 1.9998297646304935e-06,
+      "loss": 0.7064,
+      "step": 1197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.494386596196189,
+      "learning_rate": 1.99982797744601e-06,
+      "loss": 0.7038,
+      "step": 1198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.616809410004511,
+      "learning_rate": 1.9998261809300434e-06,
+      "loss": 0.6927,
+      "step": 1199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.477145452779399,
+      "learning_rate": 1.999824375082611e-06,
+      "loss": 0.6816,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.002883307869293,
+      "learning_rate": 1.999822559903729e-06,
+      "loss": 0.6986,
+      "step": 1201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.695323585949646,
+      "learning_rate": 1.999820735393415e-06,
+      "loss": 0.7161,
+      "step": 1202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.331742549770151,
+      "learning_rate": 1.999818901551685e-06,
+      "loss": 0.6908,
+      "step": 1203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.023467135499061,
+      "learning_rate": 1.9998170583785574e-06,
+      "loss": 0.7018,
+      "step": 1204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1813961376035205,
+      "learning_rate": 1.9998152058740485e-06,
+      "loss": 0.6803,
+      "step": 1205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.813279127866117,
+      "learning_rate": 1.9998133440381762e-06,
+      "loss": 0.6979,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.432925945731422,
+      "learning_rate": 1.9998114728709574e-06,
+      "loss": 0.6953,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3645925717720657,
+      "learning_rate": 1.999809592372409e-06,
+      "loss": 0.6953,
+      "step": 1208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.4150350968884,
+      "learning_rate": 1.99980770254255e-06,
+      "loss": 0.6999,
+      "step": 1209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9788356973942921,
+      "learning_rate": 1.999805803381397e-06,
+      "loss": 0.6908,
+      "step": 1210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.23548359788962,
+      "learning_rate": 1.999803894888968e-06,
+      "loss": 0.7012,
+      "step": 1211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.48532804602444,
+      "learning_rate": 1.999801977065281e-06,
+      "loss": 0.7012,
+      "step": 1212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.636968838103016,
+      "learning_rate": 1.999800049910354e-06,
+      "loss": 0.7096,
+      "step": 1213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.154523824214468,
+      "learning_rate": 1.9997981134242044e-06,
+      "loss": 0.6914,
+      "step": 1214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.986611007766323,
+      "learning_rate": 1.9997961676068505e-06,
+      "loss": 0.7018,
+      "step": 1215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.466604494787924,
+      "learning_rate": 1.999794212458311e-06,
+      "loss": 0.6979,
+      "step": 1216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2422484080020837,
+      "learning_rate": 1.9997922479786033e-06,
+      "loss": 0.6947,
+      "step": 1217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.744303436816323,
+      "learning_rate": 1.999790274167746e-06,
+      "loss": 0.6953,
+      "step": 1218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.835739474224289,
+      "learning_rate": 1.9997882910257577e-06,
+      "loss": 0.6992,
+      "step": 1219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.647085156674823,
+      "learning_rate": 1.999786298552657e-06,
+      "loss": 0.6901,
+      "step": 1220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.971176047140974,
+      "learning_rate": 1.9997842967484626e-06,
+      "loss": 0.6953,
+      "step": 1221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.223657024765664,
+      "learning_rate": 1.999782285613193e-06,
+      "loss": 0.696,
+      "step": 1222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6736473897080892,
+      "learning_rate": 1.9997802651468664e-06,
+      "loss": 0.6947,
+      "step": 1223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.77338787885395,
+      "learning_rate": 1.999778235349502e-06,
+      "loss": 0.6914,
+      "step": 1224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9687129239705199,
+      "learning_rate": 1.9997761962211196e-06,
+      "loss": 0.6953,
+      "step": 1225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.839872607290245,
+      "learning_rate": 1.999774147761737e-06,
+      "loss": 0.6986,
+      "step": 1226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.181880984220184,
+      "learning_rate": 1.999772089971374e-06,
+      "loss": 0.7077,
+      "step": 1227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.857169845698413,
+      "learning_rate": 1.99977002285005e-06,
+      "loss": 0.6953,
+      "step": 1228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5353980442142992,
+      "learning_rate": 1.999767946397784e-06,
+      "loss": 0.6862,
+      "step": 1229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1415436240028636,
+      "learning_rate": 1.999765860614595e-06,
+      "loss": 0.7025,
+      "step": 1230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.1380485777421,
+      "learning_rate": 1.9997637655005025e-06,
+      "loss": 0.6986,
+      "step": 1231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.168018878858367,
+      "learning_rate": 1.999761661055527e-06,
+      "loss": 0.6914,
+      "step": 1232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.612162747731336,
+      "learning_rate": 1.999759547279687e-06,
+      "loss": 0.6992,
+      "step": 1233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.622766172899571,
+      "learning_rate": 1.999757424173003e-06,
+      "loss": 0.6914,
+      "step": 1234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9770470780173028,
+      "learning_rate": 1.999755291735495e-06,
+      "loss": 0.6901,
+      "step": 1235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.049556702207654,
+      "learning_rate": 1.9997531499671817e-06,
+      "loss": 0.6934,
+      "step": 1236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.320300089583626,
+      "learning_rate": 1.9997509988680845e-06,
+      "loss": 0.696,
+      "step": 1237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.088392187966064,
+      "learning_rate": 1.9997488384382223e-06,
+      "loss": 0.6966,
+      "step": 1238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2107489532104567,
+      "learning_rate": 1.9997466686776157e-06,
+      "loss": 0.6927,
+      "step": 1239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.670673205376453,
+      "learning_rate": 1.9997444895862853e-06,
+      "loss": 0.7005,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.729533113636872,
+      "learning_rate": 1.9997423011642512e-06,
+      "loss": 0.7038,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.789357449723623,
+      "learning_rate": 1.9997401034115335e-06,
+      "loss": 0.6927,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.010746892037609,
+      "learning_rate": 1.999737896328153e-06,
+      "loss": 0.6934,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0664320122140882,
+      "learning_rate": 1.999735679914131e-06,
+      "loss": 0.6888,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.579495803631911,
+      "learning_rate": 1.9997334541694867e-06,
+      "loss": 0.6888,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.054798169429937,
+      "learning_rate": 1.999731219094242e-06,
+      "loss": 0.6986,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7930285160000805,
+      "learning_rate": 1.9997289746884175e-06,
+      "loss": 0.694,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8681443999272316,
+      "learning_rate": 1.9997267209520337e-06,
+      "loss": 0.6882,
+      "step": 1248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.472390037033551,
+      "learning_rate": 1.9997244578851124e-06,
+      "loss": 0.6979,
+      "step": 1249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.45181429090312,
+      "learning_rate": 1.999722185487674e-06,
+      "loss": 0.6921,
+      "step": 1250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.405767358155009,
+      "learning_rate": 1.9997199037597403e-06,
+      "loss": 0.6934,
+      "step": 1251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1321796032638491,
+      "learning_rate": 1.9997176127013323e-06,
+      "loss": 0.6882,
+      "step": 1252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.232206107865191,
+      "learning_rate": 1.9997153123124715e-06,
+      "loss": 0.6908,
+      "step": 1253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.287180239466589,
+      "learning_rate": 1.999713002593179e-06,
+      "loss": 0.6953,
+      "step": 1254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.156212370002754,
+      "learning_rate": 1.9997106835434767e-06,
+      "loss": 0.694,
+      "step": 1255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0741840937618974,
+      "learning_rate": 1.9997083551633863e-06,
+      "loss": 0.6927,
+      "step": 1256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.859232594702279,
+      "learning_rate": 1.999706017452929e-06,
+      "loss": 0.7018,
+      "step": 1257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0352310722440654,
+      "learning_rate": 1.9997036704121273e-06,
+      "loss": 0.6862,
+      "step": 1258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.743472893236556,
+      "learning_rate": 1.999701314041003e-06,
+      "loss": 0.6999,
+      "step": 1259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.781929098705826,
+      "learning_rate": 1.9996989483395777e-06,
+      "loss": 0.694,
+      "step": 1260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.153136937510018,
+      "learning_rate": 1.9996965733078743e-06,
+      "loss": 0.6895,
+      "step": 1261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8306797966532783,
+      "learning_rate": 1.9996941889459136e-06,
+      "loss": 0.6986,
+      "step": 1262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.90014173216667,
+      "learning_rate": 1.999691795253719e-06,
+      "loss": 0.6999,
+      "step": 1263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9594625827507681,
+      "learning_rate": 1.9996893922313126e-06,
+      "loss": 0.6888,
+      "step": 1264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.14362682392756,
+      "learning_rate": 1.999686979878716e-06,
+      "loss": 0.7096,
+      "step": 1265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6605323414023918,
+      "learning_rate": 1.9996845581959532e-06,
+      "loss": 0.6888,
+      "step": 1266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.139231519348176,
+      "learning_rate": 1.9996821271830457e-06,
+      "loss": 0.7142,
+      "step": 1267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0843785039271072,
+      "learning_rate": 1.9996796868400164e-06,
+      "loss": 0.6875,
+      "step": 1268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6018181722601916,
+      "learning_rate": 1.9996772371668884e-06,
+      "loss": 0.6855,
+      "step": 1269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.762104577095524,
+      "learning_rate": 1.9996747781636843e-06,
+      "loss": 0.6797,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.476519944529524,
+      "learning_rate": 1.9996723098304268e-06,
+      "loss": 0.6947,
+      "step": 1271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.510140985254167,
+      "learning_rate": 1.99966983216714e-06,
+      "loss": 0.6992,
+      "step": 1272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.975974311486059,
+      "learning_rate": 1.9996673451738456e-06,
+      "loss": 0.6901,
+      "step": 1273
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.396368395862599,
+      "learning_rate": 1.9996648488505673e-06,
+      "loss": 0.6914,
+      "step": 1274
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3106436102571353,
+      "learning_rate": 1.999662343197329e-06,
+      "loss": 0.6927,
+      "step": 1275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6122974535648966,
+      "learning_rate": 1.9996598282141533e-06,
+      "loss": 0.6934,
+      "step": 1276
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.257249902612633,
+      "learning_rate": 1.9996573039010642e-06,
+      "loss": 0.6921,
+      "step": 1277
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.807640439896453,
+      "learning_rate": 1.999654770258085e-06,
+      "loss": 0.694,
+      "step": 1278
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.575759846635215,
+      "learning_rate": 1.9996522272852397e-06,
+      "loss": 0.7051,
+      "step": 1279
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.64196492575006,
+      "learning_rate": 1.9996496749825513e-06,
+      "loss": 0.7122,
+      "step": 1280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.127270697302432,
+      "learning_rate": 1.9996471133500444e-06,
+      "loss": 0.7005,
+      "step": 1281
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.066129580383378,
+      "learning_rate": 1.9996445423877428e-06,
+      "loss": 0.7018,
+      "step": 1282
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.049409493543406,
+      "learning_rate": 1.9996419620956697e-06,
+      "loss": 0.6888,
+      "step": 1283
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8655735024874135,
+      "learning_rate": 1.99963937247385e-06,
+      "loss": 0.6875,
+      "step": 1284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.947494002884089,
+      "learning_rate": 1.999636773522308e-06,
+      "loss": 0.6908,
+      "step": 1285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.0890236445937544,
+      "learning_rate": 1.999634165241067e-06,
+      "loss": 0.7038,
+      "step": 1286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9869335573724656,
+      "learning_rate": 1.9996315476301524e-06,
+      "loss": 0.6927,
+      "step": 1287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.054760653732889,
+      "learning_rate": 1.999628920689588e-06,
+      "loss": 0.694,
+      "step": 1288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.654051156478234,
+      "learning_rate": 1.999626284419398e-06,
+      "loss": 0.7031,
+      "step": 1289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.581355788761137,
+      "learning_rate": 1.999623638819608e-06,
+      "loss": 0.6921,
+      "step": 1290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.3617304601292215,
+      "learning_rate": 1.999620983890242e-06,
+      "loss": 0.6914,
+      "step": 1291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.814202836529461,
+      "learning_rate": 1.9996183196313254e-06,
+      "loss": 0.696,
+      "step": 1292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.4843085017173125,
+      "learning_rate": 1.999615646042882e-06,
+      "loss": 0.6953,
+      "step": 1293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.113120086695793,
+      "learning_rate": 1.9996129631249378e-06,
+      "loss": 0.6855,
+      "step": 1294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3797217480070145,
+      "learning_rate": 1.9996102708775173e-06,
+      "loss": 0.6875,
+      "step": 1295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.068520073642747,
+      "learning_rate": 1.9996075693006456e-06,
+      "loss": 0.6979,
+      "step": 1296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.049757070952307,
+      "learning_rate": 1.9996048583943484e-06,
+      "loss": 0.7005,
+      "step": 1297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.465870318288912,
+      "learning_rate": 1.9996021381586503e-06,
+      "loss": 0.6849,
+      "step": 1298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.724118437911253,
+      "learning_rate": 1.9995994085935772e-06,
+      "loss": 0.7148,
+      "step": 1299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.296287527054576,
+      "learning_rate": 1.9995966696991545e-06,
+      "loss": 0.7181,
+      "step": 1300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.538296506226393,
+      "learning_rate": 1.9995939214754076e-06,
+      "loss": 0.7201,
+      "step": 1301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9520455308259126,
+      "learning_rate": 1.999591163922362e-06,
+      "loss": 0.6973,
+      "step": 1302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4895262147234547,
+      "learning_rate": 1.999588397040044e-06,
+      "loss": 0.6986,
+      "step": 1303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7404258374466436,
+      "learning_rate": 1.999585620828479e-06,
+      "loss": 0.6901,
+      "step": 1304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.402410375050735,
+      "learning_rate": 1.999582835287693e-06,
+      "loss": 0.7012,
+      "step": 1305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.296414451848538,
+      "learning_rate": 1.9995800404177117e-06,
+      "loss": 0.6973,
+      "step": 1306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.606303220903318,
+      "learning_rate": 1.999577236218562e-06,
+      "loss": 0.6816,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.306373523552896,
+      "learning_rate": 1.9995744226902692e-06,
+      "loss": 0.7038,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.787490040691799,
+      "learning_rate": 1.99957159983286e-06,
+      "loss": 0.6895,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.021838302249624,
+      "learning_rate": 1.9995687676463604e-06,
+      "loss": 0.6901,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.33482801334691,
+      "learning_rate": 1.9995659261307977e-06,
+      "loss": 0.6999,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1428061393689,
+      "learning_rate": 1.9995630752861975e-06,
+      "loss": 0.6882,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.4353747236627665,
+      "learning_rate": 1.9995602151125864e-06,
+      "loss": 0.694,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.142343908983896,
+      "learning_rate": 1.999557345609992e-06,
+      "loss": 0.6927,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.015649135381721,
+      "learning_rate": 1.99955446677844e-06,
+      "loss": 0.6816,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9089167177234908,
+      "learning_rate": 1.999551578617958e-06,
+      "loss": 0.6947,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9916049157789493,
+      "learning_rate": 1.9995486811285728e-06,
+      "loss": 0.6738,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.231007788519676,
+      "learning_rate": 1.9995457743103113e-06,
+      "loss": 0.7135,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.333495271337855,
+      "learning_rate": 1.9995428581632007e-06,
+      "loss": 0.6914,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9931078727363534,
+      "learning_rate": 1.9995399326872682e-06,
+      "loss": 0.6895,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.7279735217706165,
+      "learning_rate": 1.9995369978825413e-06,
+      "loss": 0.7012,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.457980288386507,
+      "learning_rate": 1.9995340537490467e-06,
+      "loss": 0.6973,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1535462039995195,
+      "learning_rate": 1.999531100286813e-06,
+      "loss": 0.6882,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.7209132251517225,
+      "learning_rate": 1.9995281374958666e-06,
+      "loss": 0.6947,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.041346045421298,
+      "learning_rate": 1.999525165376236e-06,
+      "loss": 0.6921,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.821284703522376,
+      "learning_rate": 1.9995221839279485e-06,
+      "loss": 0.6966,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3169094661821898,
+      "learning_rate": 1.999519193151032e-06,
+      "loss": 0.6953,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2988998825990015,
+      "learning_rate": 1.9995161930455147e-06,
+      "loss": 0.6908,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.380739230361549,
+      "learning_rate": 1.999513183611424e-06,
+      "loss": 0.6947,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.586516280188938,
+      "learning_rate": 1.999510164848789e-06,
+      "loss": 0.6888,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6288614215216497,
+      "learning_rate": 1.999507136757637e-06,
+      "loss": 0.6868,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0358191289497845,
+      "learning_rate": 1.9995040993379965e-06,
+      "loss": 0.6947,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.908100699371478,
+      "learning_rate": 1.9995010525898954e-06,
+      "loss": 0.6986,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.102991223645948,
+      "learning_rate": 1.999497996513363e-06,
+      "loss": 0.7005,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.334604708842187,
+      "learning_rate": 1.9994949311084274e-06,
+      "loss": 0.6836,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.083411795557312,
+      "learning_rate": 1.999491856375117e-06,
+      "loss": 0.6888,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.097987227219584,
+      "learning_rate": 1.999488772313461e-06,
+      "loss": 0.6979,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1596776156825794,
+      "learning_rate": 1.999485678923488e-06,
+      "loss": 0.6934,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9323683299458003,
+      "learning_rate": 1.999482576205226e-06,
+      "loss": 0.6908,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.760677986941379,
+      "learning_rate": 1.9994794641587057e-06,
+      "loss": 0.6888,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.362544715245116,
+      "learning_rate": 1.9994763427839544e-06,
+      "loss": 0.6888,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.656320930663419,
+      "learning_rate": 1.9994732120810025e-06,
+      "loss": 0.6947,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.087475156553227,
+      "learning_rate": 1.9994700720498783e-06,
+      "loss": 0.6868,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.33713385621432,
+      "learning_rate": 1.999466922690612e-06,
+      "loss": 0.6836,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.288453691280976,
+      "learning_rate": 1.999463764003232e-06,
+      "loss": 0.6882,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.381092242878674,
+      "learning_rate": 1.9994605959877687e-06,
+      "loss": 0.6947,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8482473093385436,
+      "learning_rate": 1.999457418644251e-06,
+      "loss": 0.6882,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.647695332017683,
+      "learning_rate": 1.9994542319727095e-06,
+      "loss": 0.6908,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.888796396380525,
+      "learning_rate": 1.9994510359731723e-06,
+      "loss": 0.6953,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9270326173885076,
+      "learning_rate": 1.9994478306456707e-06,
+      "loss": 0.6908,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.243929726819895,
+      "learning_rate": 1.9994446159902345e-06,
+      "loss": 0.6947,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.435124783954313,
+      "learning_rate": 1.999441392006893e-06,
+      "loss": 0.7057,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.83963017526487,
+      "learning_rate": 1.999438158695676e-06,
+      "loss": 0.7005,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.271352178902257,
+      "learning_rate": 1.9994349160566147e-06,
+      "loss": 0.6888,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1583591269810745,
+      "learning_rate": 1.9994316640897393e-06,
+      "loss": 0.6921,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.380381840069722,
+      "learning_rate": 1.9994284027950795e-06,
+      "loss": 0.6979,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.452136783738231,
+      "learning_rate": 1.999425132172666e-06,
+      "loss": 0.709,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.46442592813191,
+      "learning_rate": 1.999421852222529e-06,
+      "loss": 0.6992,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5963977928197286,
+      "learning_rate": 1.9994185629447e-06,
+      "loss": 0.6868,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.63879693467515,
+      "learning_rate": 1.9994152643392093e-06,
+      "loss": 0.6855,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9571382087868398,
+      "learning_rate": 1.999411956406087e-06,
+      "loss": 0.6914,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4779328412614885,
+      "learning_rate": 1.9994086391453652e-06,
+      "loss": 0.6908,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.12266705249971,
+      "learning_rate": 1.9994053125570737e-06,
+      "loss": 0.6953,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1835715872091432,
+      "learning_rate": 1.999401976641244e-06,
+      "loss": 0.6999,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3016687799312983,
+      "learning_rate": 1.9993986313979074e-06,
+      "loss": 0.6882,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.917511275703195,
+      "learning_rate": 1.999395276827095e-06,
+      "loss": 0.6875,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2106454311332624,
+      "learning_rate": 1.9993919129288383e-06,
+      "loss": 0.6908,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.310291805035572,
+      "learning_rate": 1.999388539703168e-06,
+      "loss": 0.6829,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.429168606881153,
+      "learning_rate": 1.9993851571501164e-06,
+      "loss": 0.694,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.001206525988767,
+      "learning_rate": 1.999381765269715e-06,
+      "loss": 0.6947,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.3587401426970365,
+      "learning_rate": 1.999378364061995e-06,
+      "loss": 0.6914,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5299353773312183,
+      "learning_rate": 1.999374953526988e-06,
+      "loss": 0.694,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9859151292271555,
+      "learning_rate": 1.9993715336647267e-06,
+      "loss": 0.6875,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.247166597509331,
+      "learning_rate": 1.999368104475242e-06,
+      "loss": 0.7044,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9503879327995626,
+      "learning_rate": 1.9993646659585668e-06,
+      "loss": 0.6849,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0152628961774994,
+      "learning_rate": 1.9993612181147323e-06,
+      "loss": 0.6842,
+      "step": 1376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.078941763629482,
+      "learning_rate": 1.9993577609437717e-06,
+      "loss": 0.694,
+      "step": 1377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.659759900662239,
+      "learning_rate": 1.9993542944457167e-06,
+      "loss": 0.7044,
+      "step": 1378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.026285832605804,
+      "learning_rate": 1.9993508186205994e-06,
+      "loss": 0.6986,
+      "step": 1379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.267818771409319,
+      "learning_rate": 1.9993473334684526e-06,
+      "loss": 0.6934,
+      "step": 1380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0229882663812377,
+      "learning_rate": 1.9993438389893087e-06,
+      "loss": 0.7018,
+      "step": 1381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1615523089778144,
+      "learning_rate": 1.9993403351832005e-06,
+      "loss": 0.6888,
+      "step": 1382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6102151246732936,
+      "learning_rate": 1.999336822050161e-06,
+      "loss": 0.6901,
+      "step": 1383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.707922626199945,
+      "learning_rate": 1.9993332995902217e-06,
+      "loss": 0.6855,
+      "step": 1384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.291693610377123,
+      "learning_rate": 1.999329767803417e-06,
+      "loss": 0.6992,
+      "step": 1385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.62972351143295,
+      "learning_rate": 1.999326226689779e-06,
+      "loss": 0.694,
+      "step": 1386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7236064358296789,
+      "learning_rate": 1.999322676249341e-06,
+      "loss": 0.6829,
+      "step": 1387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.370766239780016,
+      "learning_rate": 1.9993191164821358e-06,
+      "loss": 0.7012,
+      "step": 1388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.144971691653064,
+      "learning_rate": 1.999315547388197e-06,
+      "loss": 0.6966,
+      "step": 1389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.456980046165912,
+      "learning_rate": 1.999311968967558e-06,
+      "loss": 0.7064,
+      "step": 1390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.8192011757707585,
+      "learning_rate": 1.9993083812202517e-06,
+      "loss": 0.6986,
+      "step": 1391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.393815201126012,
+      "learning_rate": 1.9993047841463123e-06,
+      "loss": 0.6973,
+      "step": 1392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.864807960931258,
+      "learning_rate": 1.999301177745773e-06,
+      "loss": 0.6921,
+      "step": 1393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2086541712553573,
+      "learning_rate": 1.9992975620186672e-06,
+      "loss": 0.6979,
+      "step": 1394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.372256497596961,
+      "learning_rate": 1.999293936965029e-06,
+      "loss": 0.6973,
+      "step": 1395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8481659987928563,
+      "learning_rate": 1.9992903025848917e-06,
+      "loss": 0.6979,
+      "step": 1396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.22787993865055,
+      "learning_rate": 1.9992866588782903e-06,
+      "loss": 0.6875,
+      "step": 1397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.968308898670227,
+      "learning_rate": 1.9992830058452577e-06,
+      "loss": 0.7018,
+      "step": 1398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.3345894238453235,
+      "learning_rate": 1.9992793434858287e-06,
+      "loss": 0.6973,
+      "step": 1399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.72630080141327,
+      "learning_rate": 1.999275671800037e-06,
+      "loss": 0.6849,
+      "step": 1400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.0998767153567925,
+      "learning_rate": 1.999271990787917e-06,
+      "loss": 0.6908,
+      "step": 1401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.910933972594582,
+      "learning_rate": 1.9992683004495037e-06,
+      "loss": 0.6842,
+      "step": 1402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.808772309092146,
+      "learning_rate": 1.9992646007848306e-06,
+      "loss": 0.6888,
+      "step": 1403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3173534659933797,
+      "learning_rate": 1.9992608917939322e-06,
+      "loss": 0.6934,
+      "step": 1404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.695833477945199,
+      "learning_rate": 1.9992571734768444e-06,
+      "loss": 0.6953,
+      "step": 1405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4754760115528986,
+      "learning_rate": 1.9992534458336008e-06,
+      "loss": 0.6927,
+      "step": 1406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.135721713350302,
+      "learning_rate": 1.999249708864236e-06,
+      "loss": 0.6901,
+      "step": 1407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.730005580133342,
+      "learning_rate": 1.999245962568785e-06,
+      "loss": 0.6947,
+      "step": 1408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3095089379911298,
+      "learning_rate": 1.999242206947284e-06,
+      "loss": 0.6901,
+      "step": 1409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.336873933932744,
+      "learning_rate": 1.9992384419997664e-06,
+      "loss": 0.694,
+      "step": 1410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.168439481729182,
+      "learning_rate": 1.9992346677262684e-06,
+      "loss": 0.6868,
+      "step": 1411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.690830641350171,
+      "learning_rate": 1.999230884126825e-06,
+      "loss": 0.6973,
+      "step": 1412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.8587240803626335,
+      "learning_rate": 1.999227091201471e-06,
+      "loss": 0.6966,
+      "step": 1413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.04496128926026,
+      "learning_rate": 1.9992232889502427e-06,
+      "loss": 0.681,
+      "step": 1414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.298348957834557,
+      "learning_rate": 1.9992194773731744e-06,
+      "loss": 0.694,
+      "step": 1415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.1621953217606915,
+      "learning_rate": 1.999215656470303e-06,
+      "loss": 0.6921,
+      "step": 1416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.7691785339472625,
+      "learning_rate": 1.999211826241663e-06,
+      "loss": 0.6901,
+      "step": 1417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.912923096481703,
+      "learning_rate": 1.999207986687291e-06,
+      "loss": 0.6816,
+      "step": 1418
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.857804353670968,
+      "learning_rate": 1.9992041378072223e-06,
+      "loss": 0.7031,
+      "step": 1419
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0207858464368178,
+      "learning_rate": 1.999200279601493e-06,
+      "loss": 0.7025,
+      "step": 1420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4483928215027646,
+      "learning_rate": 1.9991964120701394e-06,
+      "loss": 0.6953,
+      "step": 1421
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.492184381882627,
+      "learning_rate": 1.999192535213197e-06,
+      "loss": 0.707,
+      "step": 1422
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9284906663578417,
+      "learning_rate": 1.9991886490307025e-06,
+      "loss": 0.6855,
+      "step": 1423
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1616572208596527,
+      "learning_rate": 1.9991847535226917e-06,
+      "loss": 0.6927,
+      "step": 1424
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.676605184520193,
+      "learning_rate": 1.9991808486892013e-06,
+      "loss": 0.6999,
+      "step": 1425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9636216958048496,
+      "learning_rate": 1.999176934530268e-06,
+      "loss": 0.6862,
+      "step": 1426
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.980947716983463,
+      "learning_rate": 1.999173011045928e-06,
+      "loss": 0.7018,
+      "step": 1427
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0822602410672646,
+      "learning_rate": 1.9991690782362177e-06,
+      "loss": 0.6901,
+      "step": 1428
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.253262748887245,
+      "learning_rate": 1.999165136101174e-06,
+      "loss": 0.7038,
+      "step": 1429
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.915637808423142,
+      "learning_rate": 1.999161184640834e-06,
+      "loss": 0.6829,
+      "step": 1430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.328520577440131,
+      "learning_rate": 1.9991572238552337e-06,
+      "loss": 0.681,
+      "step": 1431
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4696451365236114,
+      "learning_rate": 1.9991532537444115e-06,
+      "loss": 0.6712,
+      "step": 1432
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.24023921479907,
+      "learning_rate": 1.9991492743084033e-06,
+      "loss": 0.6934,
+      "step": 1433
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.367698957957453,
+      "learning_rate": 1.9991452855472463e-06,
+      "loss": 0.7103,
+      "step": 1434
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6237104675398717,
+      "learning_rate": 1.9991412874609778e-06,
+      "loss": 0.6901,
+      "step": 1435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.290902033317405,
+      "learning_rate": 1.999137280049636e-06,
+      "loss": 0.6784,
+      "step": 1436
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.79148844343837,
+      "learning_rate": 1.999133263313257e-06,
+      "loss": 0.6966,
+      "step": 1437
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1506205384082024,
+      "learning_rate": 1.9991292372518793e-06,
+      "loss": 0.6888,
+      "step": 1438
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.908192062762571,
+      "learning_rate": 1.99912520186554e-06,
+      "loss": 0.694,
+      "step": 1439
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6295205609624253,
+      "learning_rate": 1.999121157154277e-06,
+      "loss": 0.6986,
+      "step": 1440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3088512527684473,
+      "learning_rate": 1.999117103118127e-06,
+      "loss": 0.6849,
+      "step": 1441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.250723279714798,
+      "learning_rate": 1.9991130397571297e-06,
+      "loss": 0.6908,
+      "step": 1442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1305960313536874,
+      "learning_rate": 1.9991089670713215e-06,
+      "loss": 0.6855,
+      "step": 1443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.593224823939013,
+      "learning_rate": 1.9991048850607415e-06,
+      "loss": 0.694,
+      "step": 1444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.504946385621945,
+      "learning_rate": 1.999100793725427e-06,
+      "loss": 0.6992,
+      "step": 1445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9090905998453783,
+      "learning_rate": 1.999096693065416e-06,
+      "loss": 0.6855,
+      "step": 1446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9879595653048917,
+      "learning_rate": 1.9990925830807476e-06,
+      "loss": 0.7012,
+      "step": 1447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4649529417059681,
+      "learning_rate": 1.9990884637714596e-06,
+      "loss": 0.6862,
+      "step": 1448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.691447858669326,
+      "learning_rate": 1.999084335137591e-06,
+      "loss": 0.6908,
+      "step": 1449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6248457642800878,
+      "learning_rate": 1.9990801971791793e-06,
+      "loss": 0.6973,
+      "step": 1450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3563952294687065,
+      "learning_rate": 1.9990760498962645e-06,
+      "loss": 0.7025,
+      "step": 1451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.381495588679319,
+      "learning_rate": 1.999071893288884e-06,
+      "loss": 0.7064,
+      "step": 1452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.558572861422781,
+      "learning_rate": 1.9990677273570773e-06,
+      "loss": 0.7064,
+      "step": 1453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.592402613188227,
+      "learning_rate": 1.9990635521008837e-06,
+      "loss": 0.6966,
+      "step": 1454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.579184911649686,
+      "learning_rate": 1.999059367520341e-06,
+      "loss": 0.6855,
+      "step": 1455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.431303756300698,
+      "learning_rate": 1.999055173615489e-06,
+      "loss": 0.7038,
+      "step": 1456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.13784639894286,
+      "learning_rate": 1.9990509703863667e-06,
+      "loss": 0.6934,
+      "step": 1457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.479105861444483,
+      "learning_rate": 1.9990467578330136e-06,
+      "loss": 0.6882,
+      "step": 1458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.256134638338795,
+      "learning_rate": 1.999042535955468e-06,
+      "loss": 0.6986,
+      "step": 1459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.060115163738568,
+      "learning_rate": 1.9990383047537707e-06,
+      "loss": 0.681,
+      "step": 1460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7505406396077473,
+      "learning_rate": 1.9990340642279604e-06,
+      "loss": 0.6895,
+      "step": 1461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.135624878872752,
+      "learning_rate": 1.999029814378077e-06,
+      "loss": 0.707,
+      "step": 1462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.585701746855526,
+      "learning_rate": 1.99902555520416e-06,
+      "loss": 0.6823,
+      "step": 1463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0488945039833766,
+      "learning_rate": 1.9990212867062486e-06,
+      "loss": 0.694,
+      "step": 1464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.155114039202514,
+      "learning_rate": 1.9990170088843843e-06,
+      "loss": 0.6829,
+      "step": 1465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7191715368276308,
+      "learning_rate": 1.999012721738605e-06,
+      "loss": 0.6888,
+      "step": 1466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9869717874914117,
+      "learning_rate": 1.999008425268952e-06,
+      "loss": 0.6901,
+      "step": 1467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.692726189469868,
+      "learning_rate": 1.9990041194754646e-06,
+      "loss": 0.6895,
+      "step": 1468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.321083723221184,
+      "learning_rate": 1.9989998043581835e-06,
+      "loss": 0.7031,
+      "step": 1469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2173074620702786,
+      "learning_rate": 1.998995479917149e-06,
+      "loss": 0.6836,
+      "step": 1470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7649288252967942,
+      "learning_rate": 1.9989911461524012e-06,
+      "loss": 0.6895,
+      "step": 1471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.181114195966454,
+      "learning_rate": 1.998986803063981e-06,
+      "loss": 0.6992,
+      "step": 1472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4363371574313684,
+      "learning_rate": 1.9989824506519286e-06,
+      "loss": 0.6999,
+      "step": 1473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.140351508335359,
+      "learning_rate": 1.9989780889162846e-06,
+      "loss": 0.6842,
+      "step": 1474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7905022137777103,
+      "learning_rate": 1.9989737178570894e-06,
+      "loss": 0.6992,
+      "step": 1475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.22086757010514,
+      "learning_rate": 1.9989693374743845e-06,
+      "loss": 0.694,
+      "step": 1476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2345930780297119,
+      "learning_rate": 1.9989649477682104e-06,
+      "loss": 0.6947,
+      "step": 1477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3756779669012205,
+      "learning_rate": 1.998960548738608e-06,
+      "loss": 0.6973,
+      "step": 1478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.30433761135794,
+      "learning_rate": 1.9989561403856185e-06,
+      "loss": 0.6862,
+      "step": 1479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8785420291651507,
+      "learning_rate": 1.9989517227092828e-06,
+      "loss": 0.6895,
+      "step": 1480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.872723880130971,
+      "learning_rate": 1.9989472957096427e-06,
+      "loss": 0.7057,
+      "step": 1481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.155010806175204,
+      "learning_rate": 1.998942859386739e-06,
+      "loss": 0.6953,
+      "step": 1482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.520662239077758,
+      "learning_rate": 1.998938413740613e-06,
+      "loss": 0.696,
+      "step": 1483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.660132682275333,
+      "learning_rate": 1.9989339587713067e-06,
+      "loss": 0.6908,
+      "step": 1484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6538546266801375,
+      "learning_rate": 1.9989294944788617e-06,
+      "loss": 0.6973,
+      "step": 1485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5344956739903575,
+      "learning_rate": 1.998925020863319e-06,
+      "loss": 0.6888,
+      "step": 1486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.85707919779448,
+      "learning_rate": 1.9989205379247206e-06,
+      "loss": 0.696,
+      "step": 1487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.976320106262834,
+      "learning_rate": 1.9989160456631088e-06,
+      "loss": 0.6895,
+      "step": 1488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.240571042001372,
+      "learning_rate": 1.998911544078525e-06,
+      "loss": 0.6927,
+      "step": 1489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.236716822426893,
+      "learning_rate": 1.9989070331710118e-06,
+      "loss": 0.6882,
+      "step": 1490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.404311902637231,
+      "learning_rate": 1.9989025129406104e-06,
+      "loss": 0.6914,
+      "step": 1491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.835022036586463,
+      "learning_rate": 1.998897983387364e-06,
+      "loss": 0.6908,
+      "step": 1492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.76728307846668,
+      "learning_rate": 1.998893444511314e-06,
+      "loss": 0.6927,
+      "step": 1493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.052613348044037,
+      "learning_rate": 1.998888896312503e-06,
+      "loss": 0.6901,
+      "step": 1494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4850082373742692,
+      "learning_rate": 1.9988843387909744e-06,
+      "loss": 0.6999,
+      "step": 1495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.908445880725529,
+      "learning_rate": 1.998879771946769e-06,
+      "loss": 0.6973,
+      "step": 1496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6926802055121524,
+      "learning_rate": 1.998875195779931e-06,
+      "loss": 0.6901,
+      "step": 1497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.889265356673741,
+      "learning_rate": 1.998870610290502e-06,
+      "loss": 0.694,
+      "step": 1498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2956619961509452,
+      "learning_rate": 1.9988660154785254e-06,
+      "loss": 0.6855,
+      "step": 1499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.85843534583009,
+      "learning_rate": 1.9988614113440444e-06,
+      "loss": 0.6908,
+      "step": 1500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.33773110098262,
+      "learning_rate": 1.998856797887101e-06,
+      "loss": 0.6999,
+      "step": 1501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.1668265829991595,
+      "learning_rate": 1.9988521751077387e-06,
+      "loss": 0.6764,
+      "step": 1502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.80851401889903,
+      "learning_rate": 1.998847543006001e-06,
+      "loss": 0.707,
+      "step": 1503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.118119438380875,
+      "learning_rate": 1.9988429015819306e-06,
+      "loss": 0.6927,
+      "step": 1504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.902865177160359,
+      "learning_rate": 1.9988382508355716e-06,
+      "loss": 0.6992,
+      "step": 1505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.155749987313715,
+      "learning_rate": 1.9988335907669662e-06,
+      "loss": 0.694,
+      "step": 1506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.494475652624614,
+      "learning_rate": 1.998828921376159e-06,
+      "loss": 0.707,
+      "step": 1507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.366087729928321,
+      "learning_rate": 1.998824242663193e-06,
+      "loss": 0.6888,
+      "step": 1508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.236118968825025,
+      "learning_rate": 1.998819554628112e-06,
+      "loss": 0.6947,
+      "step": 1509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.307144419395531,
+      "learning_rate": 1.9988148572709597e-06,
+      "loss": 0.7025,
+      "step": 1510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.187924429859214,
+      "learning_rate": 1.99881015059178e-06,
+      "loss": 0.696,
+      "step": 1511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7054575518024198,
+      "learning_rate": 1.9988054345906173e-06,
+      "loss": 0.6992,
+      "step": 1512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.775694970408079,
+      "learning_rate": 1.9988007092675143e-06,
+      "loss": 0.6868,
+      "step": 1513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.343438991173024,
+      "learning_rate": 1.9987959746225165e-06,
+      "loss": 0.6829,
+      "step": 1514
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6860879142869721,
+      "learning_rate": 1.9987912306556674e-06,
+      "loss": 0.6895,
+      "step": 1515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8729819539144689,
+      "learning_rate": 1.9987864773670115e-06,
+      "loss": 0.6868,
+      "step": 1516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9283908031681842,
+      "learning_rate": 1.998781714756593e-06,
+      "loss": 0.6992,
+      "step": 1517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.977755674504611,
+      "learning_rate": 1.998776942824457e-06,
+      "loss": 0.6992,
+      "step": 1518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7788188399253542,
+      "learning_rate": 1.998772161570647e-06,
+      "loss": 0.6934,
+      "step": 1519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.098854203152109,
+      "learning_rate": 1.9987673709952076e-06,
+      "loss": 0.6875,
+      "step": 1520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.740205941355486,
+      "learning_rate": 1.9987625710981848e-06,
+      "loss": 0.6999,
+      "step": 1521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8726968589614698,
+      "learning_rate": 1.998757761879622e-06,
+      "loss": 0.6908,
+      "step": 1522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.352247531292313,
+      "learning_rate": 1.998752943339565e-06,
+      "loss": 0.6882,
+      "step": 1523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0416261146523942,
+      "learning_rate": 1.9987481154780584e-06,
+      "loss": 0.6875,
+      "step": 1524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.157950799133916,
+      "learning_rate": 1.998743278295147e-06,
+      "loss": 0.6999,
+      "step": 1525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.903312088936948,
+      "learning_rate": 1.9987384317908766e-06,
+      "loss": 0.6973,
+      "step": 1526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.583080539075241,
+      "learning_rate": 1.9987335759652922e-06,
+      "loss": 0.6934,
+      "step": 1527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5340326191452216,
+      "learning_rate": 1.998728710818439e-06,
+      "loss": 0.6973,
+      "step": 1528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.699189413875399,
+      "learning_rate": 1.998723836350362e-06,
+      "loss": 0.7005,
+      "step": 1529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7843916675773394,
+      "learning_rate": 1.9987189525611074e-06,
+      "loss": 0.694,
+      "step": 1530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.379179760625495,
+      "learning_rate": 1.9987140594507205e-06,
+      "loss": 0.694,
+      "step": 1531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.139223742460426,
+      "learning_rate": 1.9987091570192472e-06,
+      "loss": 0.6921,
+      "step": 1532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.038470341631544,
+      "learning_rate": 1.9987042452667324e-06,
+      "loss": 0.6947,
+      "step": 1533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.519043598297152,
+      "learning_rate": 1.998699324193223e-06,
+      "loss": 0.6947,
+      "step": 1534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7410713482483762,
+      "learning_rate": 1.9986943937987647e-06,
+      "loss": 0.6875,
+      "step": 1535
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.469017526990286,
+      "learning_rate": 1.9986894540834028e-06,
+      "loss": 0.6953,
+      "step": 1536
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8880282111989484,
+      "learning_rate": 1.998684505047184e-06,
+      "loss": 0.6908,
+      "step": 1537
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9873381161048735,
+      "learning_rate": 1.9986795466901546e-06,
+      "loss": 0.6914,
+      "step": 1538
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9438905555989945,
+      "learning_rate": 1.9986745790123606e-06,
+      "loss": 0.6895,
+      "step": 1539
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.76083225924013,
+      "learning_rate": 1.9986696020138486e-06,
+      "loss": 0.6908,
+      "step": 1540
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.824374020722826,
+      "learning_rate": 1.9986646156946645e-06,
+      "loss": 0.6953,
+      "step": 1541
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5186004784780445,
+      "learning_rate": 1.998659620054855e-06,
+      "loss": 0.6888,
+      "step": 1542
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5857597719825511,
+      "learning_rate": 1.9986546150944677e-06,
+      "loss": 0.6842,
+      "step": 1543
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.725387017719347,
+      "learning_rate": 1.9986496008135486e-06,
+      "loss": 0.7031,
+      "step": 1544
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.569918251324989,
+      "learning_rate": 1.9986445772121437e-06,
+      "loss": 0.7057,
+      "step": 1545
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8265146908632692,
+      "learning_rate": 1.998639544290301e-06,
+      "loss": 0.6934,
+      "step": 1546
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5519952165550555,
+      "learning_rate": 1.9986345020480677e-06,
+      "loss": 0.6927,
+      "step": 1547
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.080543318930328,
+      "learning_rate": 1.9986294504854893e-06,
+      "loss": 0.6953,
+      "step": 1548
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.246605901880578,
+      "learning_rate": 1.9986243896026145e-06,
+      "loss": 0.6927,
+      "step": 1549
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.999942661009999,
+      "learning_rate": 1.9986193193994896e-06,
+      "loss": 0.6921,
+      "step": 1550
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.113024421214025,
+      "learning_rate": 1.9986142398761627e-06,
+      "loss": 0.7012,
+      "step": 1551
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.947071652858249,
+      "learning_rate": 1.9986091510326807e-06,
+      "loss": 0.6966,
+      "step": 1552
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.675297567527166,
+      "learning_rate": 1.9986040528690906e-06,
+      "loss": 0.6986,
+      "step": 1553
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.830767704950079,
+      "learning_rate": 1.998598945385441e-06,
+      "loss": 0.6992,
+      "step": 1554
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.552290728559268,
+      "learning_rate": 1.9985938285817798e-06,
+      "loss": 0.6921,
+      "step": 1555
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.147455377316412,
+      "learning_rate": 1.998588702458153e-06,
+      "loss": 0.6947,
+      "step": 1556
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.743104247072329,
+      "learning_rate": 1.99858356701461e-06,
+      "loss": 0.6868,
+      "step": 1557
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7502941282094118,
+      "learning_rate": 1.9985784222511983e-06,
+      "loss": 0.6979,
+      "step": 1558
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1305367453853248,
+      "learning_rate": 1.998573268167966e-06,
+      "loss": 0.6947,
+      "step": 1559
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6861034043419411,
+      "learning_rate": 1.9985681047649607e-06,
+      "loss": 0.6895,
+      "step": 1560
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6458015220654232,
+      "learning_rate": 1.998562932042231e-06,
+      "loss": 0.6888,
+      "step": 1561
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.46144657906583,
+      "learning_rate": 1.9985577499998256e-06,
+      "loss": 0.7005,
+      "step": 1562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0152310691816928,
+      "learning_rate": 1.998552558637792e-06,
+      "loss": 0.6921,
+      "step": 1563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.627741924089239,
+      "learning_rate": 1.9985473579561792e-06,
+      "loss": 0.6908,
+      "step": 1564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.255235829810926,
+      "learning_rate": 1.9985421479550357e-06,
+      "loss": 0.6901,
+      "step": 1565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3494389126895694,
+      "learning_rate": 1.99853692863441e-06,
+      "loss": 0.6947,
+      "step": 1566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1460994405208043,
+      "learning_rate": 1.9985316999943505e-06,
+      "loss": 0.6868,
+      "step": 1567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4084413003414373,
+      "learning_rate": 1.9985264620349067e-06,
+      "loss": 0.6882,
+      "step": 1568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.687557630674605,
+      "learning_rate": 1.998521214756127e-06,
+      "loss": 0.6868,
+      "step": 1569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4799908063906995,
+      "learning_rate": 1.9985159581580602e-06,
+      "loss": 0.6908,
+      "step": 1570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4215022144351395,
+      "learning_rate": 1.9985106922407562e-06,
+      "loss": 0.6921,
+      "step": 1571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.565014955617805,
+      "learning_rate": 1.998505417004263e-06,
+      "loss": 0.6914,
+      "step": 1572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.517504934573657,
+      "learning_rate": 1.998500132448631e-06,
+      "loss": 0.6914,
+      "step": 1573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5359858101084367,
+      "learning_rate": 1.998494838573909e-06,
+      "loss": 0.6895,
+      "step": 1574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5549559454041229,
+      "learning_rate": 1.998489535380146e-06,
+      "loss": 0.6973,
+      "step": 1575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1942567684412024,
+      "learning_rate": 1.998484222867392e-06,
+      "loss": 0.6914,
+      "step": 1576
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.400683344337876,
+      "learning_rate": 1.9984789010356965e-06,
+      "loss": 0.6953,
+      "step": 1577
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0783367419777454,
+      "learning_rate": 1.998473569885109e-06,
+      "loss": 0.7012,
+      "step": 1578
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6963836635757945,
+      "learning_rate": 1.9984682294156796e-06,
+      "loss": 0.6979,
+      "step": 1579
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4068140699898493,
+      "learning_rate": 1.9984628796274576e-06,
+      "loss": 0.6849,
+      "step": 1580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.290381960016264,
+      "learning_rate": 1.9984575205204933e-06,
+      "loss": 0.696,
+      "step": 1581
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.340362553008042,
+      "learning_rate": 1.998452152094837e-06,
+      "loss": 0.6979,
+      "step": 1582
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3370064005579032,
+      "learning_rate": 1.998446774350538e-06,
+      "loss": 0.6979,
+      "step": 1583
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8861791204493215,
+      "learning_rate": 1.9984413872876474e-06,
+      "loss": 0.6855,
+      "step": 1584
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.852711729472746,
+      "learning_rate": 1.998435990906215e-06,
+      "loss": 0.694,
+      "step": 1585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.44685151917126,
+      "learning_rate": 1.9984305852062907e-06,
+      "loss": 0.6992,
+      "step": 1586
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4189454533842563,
+      "learning_rate": 1.998425170187926e-06,
+      "loss": 0.6901,
+      "step": 1587
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.452967935859514,
+      "learning_rate": 1.9984197458511704e-06,
+      "loss": 0.7005,
+      "step": 1588
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8411422121489402,
+      "learning_rate": 1.9984143121960755e-06,
+      "loss": 0.6953,
+      "step": 1589
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6236646724892793,
+      "learning_rate": 1.998408869222691e-06,
+      "loss": 0.6862,
+      "step": 1590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8460367294446257,
+      "learning_rate": 1.9984034169310687e-06,
+      "loss": 0.6816,
+      "step": 1591
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.407571404430164,
+      "learning_rate": 1.998397955321259e-06,
+      "loss": 0.6868,
+      "step": 1592
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6356123729107992,
+      "learning_rate": 1.9983924843933126e-06,
+      "loss": 0.6908,
+      "step": 1593
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.275019858775861,
+      "learning_rate": 1.998387004147281e-06,
+      "loss": 0.6849,
+      "step": 1594
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.404253541504839,
+      "learning_rate": 1.998381514583215e-06,
+      "loss": 0.694,
+      "step": 1595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.024806797410414,
+      "learning_rate": 1.9983760157011664e-06,
+      "loss": 0.7012,
+      "step": 1596
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.486182033835039,
+      "learning_rate": 1.9983705075011856e-06,
+      "loss": 0.6921,
+      "step": 1597
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.322936193216696,
+      "learning_rate": 1.9983649899833254e-06,
+      "loss": 0.6947,
+      "step": 1598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.794555007728654,
+      "learning_rate": 1.998359463147636e-06,
+      "loss": 0.6934,
+      "step": 1599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.908177061182766,
+      "learning_rate": 1.998353926994169e-06,
+      "loss": 0.6986,
+      "step": 1600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6457877768423204,
+      "learning_rate": 1.9983483815229773e-06,
+      "loss": 0.6862,
+      "step": 1601
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0309336261492552,
+      "learning_rate": 1.9983428267341113e-06,
+      "loss": 0.6908,
+      "step": 1602
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8714146780875534,
+      "learning_rate": 1.9983372626276237e-06,
+      "loss": 0.6934,
+      "step": 1603
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.30435298685771,
+      "learning_rate": 1.998331689203566e-06,
+      "loss": 0.6914,
+      "step": 1604
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.268348980171225,
+      "learning_rate": 1.9983261064619904e-06,
+      "loss": 0.6862,
+      "step": 1605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2901307050246993,
+      "learning_rate": 1.9983205144029493e-06,
+      "loss": 0.6816,
+      "step": 1606
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9179867435191638,
+      "learning_rate": 1.998314913026494e-06,
+      "loss": 0.6908,
+      "step": 1607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.208068285332532,
+      "learning_rate": 1.9983093023326778e-06,
+      "loss": 0.7064,
+      "step": 1608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1082117493754375,
+      "learning_rate": 1.998303682321552e-06,
+      "loss": 0.6947,
+      "step": 1609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.595879245570591,
+      "learning_rate": 1.99829805299317e-06,
+      "loss": 0.6868,
+      "step": 1610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.115481429542578,
+      "learning_rate": 1.998292414347584e-06,
+      "loss": 0.6927,
+      "step": 1611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.222653229764159,
+      "learning_rate": 1.998286766384847e-06,
+      "loss": 0.6862,
+      "step": 1612
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.200913433576645,
+      "learning_rate": 1.998281109105011e-06,
+      "loss": 0.7038,
+      "step": 1613
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.212193992644344,
+      "learning_rate": 1.9982754425081288e-06,
+      "loss": 0.696,
+      "step": 1614
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.501754666327898,
+      "learning_rate": 1.998269766594254e-06,
+      "loss": 0.6908,
+      "step": 1615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.946064345510439,
+      "learning_rate": 1.998264081363439e-06,
+      "loss": 0.6966,
+      "step": 1616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.087211222053248,
+      "learning_rate": 1.998258386815737e-06,
+      "loss": 0.6953,
+      "step": 1617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9765365442018128,
+      "learning_rate": 1.998252682951201e-06,
+      "loss": 0.6758,
+      "step": 1618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.495186050946474,
+      "learning_rate": 1.9982469697698846e-06,
+      "loss": 0.696,
+      "step": 1619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.203571827908287,
+      "learning_rate": 1.9982412472718413e-06,
+      "loss": 0.679,
+      "step": 1620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3931333360358749,
+      "learning_rate": 1.998235515457124e-06,
+      "loss": 0.6868,
+      "step": 1621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7196929798264255,
+      "learning_rate": 1.9982297743257862e-06,
+      "loss": 0.6855,
+      "step": 1622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.565883817835107,
+      "learning_rate": 1.9982240238778817e-06,
+      "loss": 0.6901,
+      "step": 1623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.569580187592717,
+      "learning_rate": 1.9982182641134643e-06,
+      "loss": 0.6953,
+      "step": 1624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.340102309528225,
+      "learning_rate": 1.9982124950325874e-06,
+      "loss": 0.696,
+      "step": 1625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.471030253261582,
+      "learning_rate": 1.9982067166353048e-06,
+      "loss": 0.681,
+      "step": 1626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.991380963675249,
+      "learning_rate": 1.9982009289216707e-06,
+      "loss": 0.6934,
+      "step": 1627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.607249341148323,
+      "learning_rate": 1.9981951318917394e-06,
+      "loss": 0.7057,
+      "step": 1628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.251186384191094,
+      "learning_rate": 1.9981893255455645e-06,
+      "loss": 0.6862,
+      "step": 1629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6865127947307337,
+      "learning_rate": 1.9981835098832e-06,
+      "loss": 0.6855,
+      "step": 1630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.048699197118374,
+      "learning_rate": 1.9981776849047008e-06,
+      "loss": 0.6842,
+      "step": 1631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4365875265588306,
+      "learning_rate": 1.998171850610121e-06,
+      "loss": 0.6895,
+      "step": 1632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.724215268923454,
+      "learning_rate": 1.9981660069995153e-06,
+      "loss": 0.6849,
+      "step": 1633
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3947253637787764,
+      "learning_rate": 1.998160154072938e-06,
+      "loss": 0.6895,
+      "step": 1634
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.326004436044123,
+      "learning_rate": 1.9981542918304435e-06,
+      "loss": 0.7161,
+      "step": 1635
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.155731769378703,
+      "learning_rate": 1.9981484202720863e-06,
+      "loss": 0.6914,
+      "step": 1636
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.58989171226234,
+      "learning_rate": 1.998142539397922e-06,
+      "loss": 0.6901,
+      "step": 1637
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8744460440956406,
+      "learning_rate": 1.9981366492080053e-06,
+      "loss": 0.6921,
+      "step": 1638
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.7156451938889905,
+      "learning_rate": 1.9981307497023907e-06,
+      "loss": 0.6901,
+      "step": 1639
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2277335393079185,
+      "learning_rate": 1.9981248408811335e-06,
+      "loss": 0.6901,
+      "step": 1640
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6315519602757287,
+      "learning_rate": 1.9981189227442893e-06,
+      "loss": 0.6973,
+      "step": 1641
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0758100761734088,
+      "learning_rate": 1.9981129952919123e-06,
+      "loss": 0.6914,
+      "step": 1642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.129532419892393,
+      "learning_rate": 1.998107058524059e-06,
+      "loss": 0.6921,
+      "step": 1643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8206675906290006,
+      "learning_rate": 1.998101112440784e-06,
+      "loss": 0.6829,
+      "step": 1644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.342844322566621,
+      "learning_rate": 1.998095157042143e-06,
+      "loss": 0.6842,
+      "step": 1645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.098992497143411,
+      "learning_rate": 1.9980891923281915e-06,
+      "loss": 0.6979,
+      "step": 1646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9625789964917293,
+      "learning_rate": 1.9980832182989856e-06,
+      "loss": 0.696,
+      "step": 1647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.035442956401762,
+      "learning_rate": 1.998077234954581e-06,
+      "loss": 0.6816,
+      "step": 1648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0501705085614965,
+      "learning_rate": 1.9980712422950326e-06,
+      "loss": 0.7031,
+      "step": 1649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.133108137506468,
+      "learning_rate": 1.9980652403203975e-06,
+      "loss": 0.6966,
+      "step": 1650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.294483669339859,
+      "learning_rate": 1.9980592290307314e-06,
+      "loss": 0.6901,
+      "step": 1651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9729553950110086,
+      "learning_rate": 1.99805320842609e-06,
+      "loss": 0.6868,
+      "step": 1652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2450217975116464,
+      "learning_rate": 1.9980471785065295e-06,
+      "loss": 0.6934,
+      "step": 1653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.21259511283591,
+      "learning_rate": 1.998041139272107e-06,
+      "loss": 0.6973,
+      "step": 1654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.7118678887976655,
+      "learning_rate": 1.9980350907228776e-06,
+      "loss": 0.6934,
+      "step": 1655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4225741404561323,
+      "learning_rate": 1.998029032858899e-06,
+      "loss": 0.6868,
+      "step": 1656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2493721603353753,
+      "learning_rate": 1.998022965680227e-06,
+      "loss": 0.6764,
+      "step": 1657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5228460805188365,
+      "learning_rate": 1.9980168891869185e-06,
+      "loss": 0.694,
+      "step": 1658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8133461729086762,
+      "learning_rate": 1.99801080337903e-06,
+      "loss": 0.6823,
+      "step": 1659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.769774113557354,
+      "learning_rate": 1.9980047082566183e-06,
+      "loss": 0.6914,
+      "step": 1660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0010870931287443,
+      "learning_rate": 1.9979986038197406e-06,
+      "loss": 0.6973,
+      "step": 1661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.176162478070075,
+      "learning_rate": 1.9979924900684536e-06,
+      "loss": 0.6875,
+      "step": 1662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.82348591010482,
+      "learning_rate": 1.9979863670028142e-06,
+      "loss": 0.6895,
+      "step": 1663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.027177941968828,
+      "learning_rate": 1.99798023462288e-06,
+      "loss": 0.6992,
+      "step": 1664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.817945809438226,
+      "learning_rate": 1.997974092928708e-06,
+      "loss": 0.6908,
+      "step": 1665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0753856827414967,
+      "learning_rate": 1.9979679419203553e-06,
+      "loss": 0.6999,
+      "step": 1666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5443525676291443,
+      "learning_rate": 1.9979617815978803e-06,
+      "loss": 0.6862,
+      "step": 1667
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.546565204157627,
+      "learning_rate": 1.997955611961339e-06,
+      "loss": 0.6999,
+      "step": 1668
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3007517607327679,
+      "learning_rate": 1.99794943301079e-06,
+      "loss": 0.681,
+      "step": 1669
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6921311659973253,
+      "learning_rate": 1.9979432447462904e-06,
+      "loss": 0.6842,
+      "step": 1670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.064131228141342,
+      "learning_rate": 1.9979370471678986e-06,
+      "loss": 0.668,
+      "step": 1671
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.270042700125399,
+      "learning_rate": 1.997930840275672e-06,
+      "loss": 0.6992,
+      "step": 1672
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5408794992957389,
+      "learning_rate": 1.9979246240696684e-06,
+      "loss": 0.6816,
+      "step": 1673
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.479269617334028,
+      "learning_rate": 1.9979183985499463e-06,
+      "loss": 0.6979,
+      "step": 1674
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9581459056257842,
+      "learning_rate": 1.9979121637165633e-06,
+      "loss": 0.6868,
+      "step": 1675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.855720688261195,
+      "learning_rate": 1.997905919569578e-06,
+      "loss": 0.6979,
+      "step": 1676
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8365189740512924,
+      "learning_rate": 1.9978996661090485e-06,
+      "loss": 0.6947,
+      "step": 1677
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.685634819765307,
+      "learning_rate": 1.997893403335033e-06,
+      "loss": 0.7044,
+      "step": 1678
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.282323274554267,
+      "learning_rate": 1.9978871312475902e-06,
+      "loss": 0.6927,
+      "step": 1679
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9007650066797157,
+      "learning_rate": 1.9978808498467787e-06,
+      "loss": 0.6973,
+      "step": 1680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.731492378976467,
+      "learning_rate": 1.997874559132657e-06,
+      "loss": 0.6901,
+      "step": 1681
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1071475306279925,
+      "learning_rate": 1.9978682591052838e-06,
+      "loss": 0.6855,
+      "step": 1682
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5381640978465945,
+      "learning_rate": 1.9978619497647178e-06,
+      "loss": 0.6882,
+      "step": 1683
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.457767049228258,
+      "learning_rate": 1.9978556311110177e-06,
+      "loss": 0.7064,
+      "step": 1684
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.610392584250535,
+      "learning_rate": 1.997849303144243e-06,
+      "loss": 0.7103,
+      "step": 1685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3494868122469272,
+      "learning_rate": 1.9978429658644522e-06,
+      "loss": 0.6888,
+      "step": 1686
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6646815806652946,
+      "learning_rate": 1.9978366192717054e-06,
+      "loss": 0.6921,
+      "step": 1687
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7410519245757201,
+      "learning_rate": 1.997830263366061e-06,
+      "loss": 0.6875,
+      "step": 1688
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4916402249873393,
+      "learning_rate": 1.997823898147578e-06,
+      "loss": 0.6947,
+      "step": 1689
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9585999806720356,
+      "learning_rate": 1.9978175236163167e-06,
+      "loss": 0.6862,
+      "step": 1690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.779386171738725,
+      "learning_rate": 1.997811139772336e-06,
+      "loss": 0.6914,
+      "step": 1691
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.493897224284139,
+      "learning_rate": 1.997804746615696e-06,
+      "loss": 0.6868,
+      "step": 1692
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.861236157848409,
+      "learning_rate": 1.997798344146456e-06,
+      "loss": 0.6895,
+      "step": 1693
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0461672836730442,
+      "learning_rate": 1.9977919323646758e-06,
+      "loss": 0.6849,
+      "step": 1694
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2518969128109747,
+      "learning_rate": 1.997785511270415e-06,
+      "loss": 0.6934,
+      "step": 1695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.406059025627673,
+      "learning_rate": 1.9977790808637344e-06,
+      "loss": 0.6914,
+      "step": 1696
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.3466288991572934,
+      "learning_rate": 1.9977726411446926e-06,
+      "loss": 0.6888,
+      "step": 1697
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.972945216073848,
+      "learning_rate": 1.9977661921133507e-06,
+      "loss": 0.696,
+      "step": 1698
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.266333369310395,
+      "learning_rate": 1.9977597337697692e-06,
+      "loss": 0.6999,
+      "step": 1699
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7954096992169426,
+      "learning_rate": 1.9977532661140075e-06,
+      "loss": 0.6966,
+      "step": 1700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6706086817385994,
+      "learning_rate": 1.9977467891461265e-06,
+      "loss": 0.681,
+      "step": 1701
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.493955110601981,
+      "learning_rate": 1.9977403028661863e-06,
+      "loss": 0.6784,
+      "step": 1702
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.046176234305001,
+      "learning_rate": 1.9977338072742475e-06,
+      "loss": 0.694,
+      "step": 1703
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.2050430255441755,
+      "learning_rate": 1.997727302370371e-06,
+      "loss": 0.6973,
+      "step": 1704
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9165798741680815,
+      "learning_rate": 1.9977207881546174e-06,
+      "loss": 0.6862,
+      "step": 1705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.974545277278057,
+      "learning_rate": 1.9977142646270475e-06,
+      "loss": 0.6992,
+      "step": 1706
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.4194016586406635,
+      "learning_rate": 1.997707731787722e-06,
+      "loss": 0.6953,
+      "step": 1707
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.1653137033887,
+      "learning_rate": 1.997701189636702e-06,
+      "loss": 0.6908,
+      "step": 1708
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5646101730088771,
+      "learning_rate": 1.9976946381740485e-06,
+      "loss": 0.6888,
+      "step": 1709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.0750690939107095,
+      "learning_rate": 1.997688077399823e-06,
+      "loss": 0.679,
+      "step": 1710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.224457441071677,
+      "learning_rate": 1.997681507314086e-06,
+      "loss": 0.6914,
+      "step": 1711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.372484454938346,
+      "learning_rate": 1.9976749279168998e-06,
+      "loss": 0.6934,
+      "step": 1712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7148499435186948,
+      "learning_rate": 1.9976683392083248e-06,
+      "loss": 0.6882,
+      "step": 1713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.750872240297091,
+      "learning_rate": 1.9976617411884233e-06,
+      "loss": 0.6953,
+      "step": 1714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.497533753151583,
+      "learning_rate": 1.9976551338572564e-06,
+      "loss": 0.6953,
+      "step": 1715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.119147856669996,
+      "learning_rate": 1.9976485172148858e-06,
+      "loss": 0.6875,
+      "step": 1716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.394086720284962,
+      "learning_rate": 1.9976418912613734e-06,
+      "loss": 0.7142,
+      "step": 1717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6226873700375197,
+      "learning_rate": 1.9976352559967814e-06,
+      "loss": 0.6706,
+      "step": 1718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.382407614416505,
+      "learning_rate": 1.9976286114211705e-06,
+      "loss": 0.6992,
+      "step": 1719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.35107874584466,
+      "learning_rate": 1.997621957534604e-06,
+      "loss": 0.6758,
+      "step": 1720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4035281063762857,
+      "learning_rate": 1.997615294337144e-06,
+      "loss": 0.6934,
+      "step": 1721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.296239625427763,
+      "learning_rate": 1.9976086218288518e-06,
+      "loss": 0.6888,
+      "step": 1722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3308804432730155,
+      "learning_rate": 1.99760194000979e-06,
+      "loss": 0.6895,
+      "step": 1723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9095148328279062,
+      "learning_rate": 1.9975952488800213e-06,
+      "loss": 0.6849,
+      "step": 1724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.784512590657828,
+      "learning_rate": 1.9975885484396077e-06,
+      "loss": 0.7064,
+      "step": 1725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.209262368316938,
+      "learning_rate": 1.997581838688612e-06,
+      "loss": 0.6992,
+      "step": 1726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.082951126823297,
+      "learning_rate": 1.997575119627097e-06,
+      "loss": 0.6849,
+      "step": 1727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.052375056917112,
+      "learning_rate": 1.997568391255125e-06,
+      "loss": 0.679,
+      "step": 1728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.913467539833512,
+      "learning_rate": 1.997561653572759e-06,
+      "loss": 0.6999,
+      "step": 1729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.938791661826796,
+      "learning_rate": 1.997554906580062e-06,
+      "loss": 0.6875,
+      "step": 1730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5866727833732528,
+      "learning_rate": 1.9975481502770966e-06,
+      "loss": 0.7005,
+      "step": 1731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8270200316659735,
+      "learning_rate": 1.9975413846639266e-06,
+      "loss": 0.6895,
+      "step": 1732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.975061000904369,
+      "learning_rate": 1.997534609740614e-06,
+      "loss": 0.6882,
+      "step": 1733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.0964345335905896,
+      "learning_rate": 1.9975278255072228e-06,
+      "loss": 0.6895,
+      "step": 1734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.3693843343620316,
+      "learning_rate": 1.9975210319638164e-06,
+      "loss": 0.6966,
+      "step": 1735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8531611437742384,
+      "learning_rate": 1.997514229110458e-06,
+      "loss": 0.6777,
+      "step": 1736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.416044916361979,
+      "learning_rate": 1.997507416947211e-06,
+      "loss": 0.696,
+      "step": 1737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9916692737932127,
+      "learning_rate": 1.997500595474139e-06,
+      "loss": 0.6888,
+      "step": 1738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.212238054276413,
+      "learning_rate": 1.9974937646913057e-06,
+      "loss": 0.6836,
+      "step": 1739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3163543866766587,
+      "learning_rate": 1.997486924598775e-06,
+      "loss": 0.6992,
+      "step": 1740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.337699268193681,
+      "learning_rate": 1.9974800751966105e-06,
+      "loss": 0.7005,
+      "step": 1741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.50079428754215,
+      "learning_rate": 1.9974732164848766e-06,
+      "loss": 0.6836,
+      "step": 1742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7040055422363456,
+      "learning_rate": 1.9974663484636364e-06,
+      "loss": 0.6992,
+      "step": 1743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8170346790101009,
+      "learning_rate": 1.997459471132955e-06,
+      "loss": 0.6855,
+      "step": 1744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.234104219270103,
+      "learning_rate": 1.997452584492896e-06,
+      "loss": 0.7044,
+      "step": 1745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.265170923681604,
+      "learning_rate": 1.997445688543524e-06,
+      "loss": 0.6966,
+      "step": 1746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7244007768476572,
+      "learning_rate": 1.997438783284903e-06,
+      "loss": 0.6784,
+      "step": 1747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5784947214961316,
+      "learning_rate": 1.9974318687170974e-06,
+      "loss": 0.6986,
+      "step": 1748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9521640937677553,
+      "learning_rate": 1.997424944840172e-06,
+      "loss": 0.6927,
+      "step": 1749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.165430748975872,
+      "learning_rate": 1.997418011654192e-06,
+      "loss": 0.6823,
+      "step": 1750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6363373063349678,
+      "learning_rate": 1.9974110691592205e-06,
+      "loss": 0.6953,
+      "step": 1751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6392700780198126,
+      "learning_rate": 1.997404117355324e-06,
+      "loss": 0.6829,
+      "step": 1752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5846631347852327,
+      "learning_rate": 1.9973971562425665e-06,
+      "loss": 0.6947,
+      "step": 1753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.686545962127944,
+      "learning_rate": 1.997390185821013e-06,
+      "loss": 0.6986,
+      "step": 1754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.185680148764041,
+      "learning_rate": 1.997383206090729e-06,
+      "loss": 0.6849,
+      "step": 1755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3935474859542625,
+      "learning_rate": 1.9973762170517793e-06,
+      "loss": 0.6875,
+      "step": 1756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.657229887511287,
+      "learning_rate": 1.9973692187042288e-06,
+      "loss": 0.6862,
+      "step": 1757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4852855762145727,
+      "learning_rate": 1.9973622110481436e-06,
+      "loss": 0.6973,
+      "step": 1758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.033273091478141,
+      "learning_rate": 1.9973551940835886e-06,
+      "loss": 0.6882,
+      "step": 1759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6438373277100433,
+      "learning_rate": 1.997348167810629e-06,
+      "loss": 0.694,
+      "step": 1760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6871113432650409,
+      "learning_rate": 1.997341132229331e-06,
+      "loss": 0.6901,
+      "step": 1761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.096555649728867,
+      "learning_rate": 1.99733408733976e-06,
+      "loss": 0.6921,
+      "step": 1762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6169829799847172,
+      "learning_rate": 1.997327033141982e-06,
+      "loss": 0.6888,
+      "step": 1763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.18899015148802,
+      "learning_rate": 1.9973199696360623e-06,
+      "loss": 0.6947,
+      "step": 1764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.972831518197226,
+      "learning_rate": 1.997312896822067e-06,
+      "loss": 0.6875,
+      "step": 1765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.342368122517657,
+      "learning_rate": 1.9973058147000626e-06,
+      "loss": 0.6973,
+      "step": 1766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.623865043498567,
+      "learning_rate": 1.9972987232701146e-06,
+      "loss": 0.6836,
+      "step": 1767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.787001941161843,
+      "learning_rate": 1.9972916225322897e-06,
+      "loss": 0.7031,
+      "step": 1768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.820517246483959,
+      "learning_rate": 1.9972845124866535e-06,
+      "loss": 0.6823,
+      "step": 1769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8561942136063438,
+      "learning_rate": 1.997277393133273e-06,
+      "loss": 0.6855,
+      "step": 1770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.819000375726261,
+      "learning_rate": 1.9972702644722146e-06,
+      "loss": 0.6868,
+      "step": 1771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.631515252285526,
+      "learning_rate": 1.9972631265035443e-06,
+      "loss": 0.6921,
+      "step": 1772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2068231264775475,
+      "learning_rate": 1.997255979227329e-06,
+      "loss": 0.6973,
+      "step": 1773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7782017362437023,
+      "learning_rate": 1.9972488226436353e-06,
+      "loss": 0.6999,
+      "step": 1774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6408663677826503,
+      "learning_rate": 1.9972416567525304e-06,
+      "loss": 0.6842,
+      "step": 1775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.099510967985226,
+      "learning_rate": 1.9972344815540813e-06,
+      "loss": 0.7005,
+      "step": 1776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.003273339459401,
+      "learning_rate": 1.997227297048354e-06,
+      "loss": 0.7012,
+      "step": 1777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.653673829512176,
+      "learning_rate": 1.997220103235416e-06,
+      "loss": 0.6855,
+      "step": 1778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0412659656747654,
+      "learning_rate": 1.9972129001153353e-06,
+      "loss": 0.6992,
+      "step": 1779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.2557769981833795,
+      "learning_rate": 1.997205687688178e-06,
+      "loss": 0.6914,
+      "step": 1780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9051840641123827,
+      "learning_rate": 1.997198465954012e-06,
+      "loss": 0.6908,
+      "step": 1781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.711262711170887,
+      "learning_rate": 1.9971912349129045e-06,
+      "loss": 0.6842,
+      "step": 1782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.6766579493085705,
+      "learning_rate": 1.997183994564923e-06,
+      "loss": 0.6836,
+      "step": 1783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.509212374292903,
+      "learning_rate": 1.997176744910135e-06,
+      "loss": 0.6855,
+      "step": 1784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.392145817719588,
+      "learning_rate": 1.997169485948608e-06,
+      "loss": 0.6797,
+      "step": 1785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.707356216736125,
+      "learning_rate": 1.9971622176804103e-06,
+      "loss": 0.6927,
+      "step": 1786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.640765224054396,
+      "learning_rate": 1.99715494010561e-06,
+      "loss": 0.7109,
+      "step": 1787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.570185564929984,
+      "learning_rate": 1.997147653224274e-06,
+      "loss": 0.6849,
+      "step": 1788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.802075798536925,
+      "learning_rate": 1.9971403570364704e-06,
+      "loss": 0.6908,
+      "step": 1789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9709269649205211,
+      "learning_rate": 1.9971330515422676e-06,
+      "loss": 0.6836,
+      "step": 1790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7189503477870511,
+      "learning_rate": 1.997125736741734e-06,
+      "loss": 0.6836,
+      "step": 1791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1616459812006188,
+      "learning_rate": 1.9971184126349382e-06,
+      "loss": 0.6927,
+      "step": 1792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.152126271325931,
+      "learning_rate": 1.997111079221948e-06,
+      "loss": 0.694,
+      "step": 1793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5314171793009081,
+      "learning_rate": 1.9971037365028317e-06,
+      "loss": 0.6921,
+      "step": 1794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.850546967332377,
+      "learning_rate": 1.997096384477658e-06,
+      "loss": 0.6888,
+      "step": 1795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.376039645538997,
+      "learning_rate": 1.9970890231464956e-06,
+      "loss": 0.7083,
+      "step": 1796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0947510154489883,
+      "learning_rate": 1.997081652509413e-06,
+      "loss": 0.6901,
+      "step": 1797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.277309123701361,
+      "learning_rate": 1.9970742725664793e-06,
+      "loss": 0.694,
+      "step": 1798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.33957817177092,
+      "learning_rate": 1.9970668833177633e-06,
+      "loss": 0.6927,
+      "step": 1799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.000696742253296,
+      "learning_rate": 1.997059484763334e-06,
+      "loss": 0.6842,
+      "step": 1800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6897326725364983,
+      "learning_rate": 1.99705207690326e-06,
+      "loss": 0.6823,
+      "step": 1801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7936201179844063,
+      "learning_rate": 1.997044659737611e-06,
+      "loss": 0.6882,
+      "step": 1802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.617304358381109,
+      "learning_rate": 1.997037233266456e-06,
+      "loss": 0.681,
+      "step": 1803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9259132358432596,
+      "learning_rate": 1.9970297974898642e-06,
+      "loss": 0.6973,
+      "step": 1804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.92925814616004,
+      "learning_rate": 1.9970223524079055e-06,
+      "loss": 0.6992,
+      "step": 1805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5226413241308028,
+      "learning_rate": 1.997014898020649e-06,
+      "loss": 0.6862,
+      "step": 1806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.057630119648912,
+      "learning_rate": 1.997007434328164e-06,
+      "loss": 0.6947,
+      "step": 1807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.308825868620024,
+      "learning_rate": 1.9969999613305206e-06,
+      "loss": 0.6986,
+      "step": 1808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5936845615183888,
+      "learning_rate": 1.9969924790277882e-06,
+      "loss": 0.6784,
+      "step": 1809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.21511251049453,
+      "learning_rate": 1.9969849874200372e-06,
+      "loss": 0.6868,
+      "step": 1810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.326643652977999,
+      "learning_rate": 1.9969774865073367e-06,
+      "loss": 0.6901,
+      "step": 1811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.20600371214184,
+      "learning_rate": 1.9969699762897573e-06,
+      "loss": 0.7031,
+      "step": 1812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0854367342669202,
+      "learning_rate": 1.996962456767369e-06,
+      "loss": 0.6784,
+      "step": 1813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.60441118441897,
+      "learning_rate": 1.996954927940242e-06,
+      "loss": 0.7103,
+      "step": 1814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.414465274849144,
+      "learning_rate": 1.9969473898084465e-06,
+      "loss": 0.7116,
+      "step": 1815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.16072729785614,
+      "learning_rate": 1.9969398423720527e-06,
+      "loss": 0.7051,
+      "step": 1816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.855196631204361,
+      "learning_rate": 1.996932285631131e-06,
+      "loss": 0.6966,
+      "step": 1817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.077452268921966,
+      "learning_rate": 1.9969247195857524e-06,
+      "loss": 0.694,
+      "step": 1818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0139134704569472,
+      "learning_rate": 1.996917144235987e-06,
+      "loss": 0.6764,
+      "step": 1819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1784726360547384,
+      "learning_rate": 1.996909559581906e-06,
+      "loss": 0.6882,
+      "step": 1820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.240215594954575,
+      "learning_rate": 1.99690196562358e-06,
+      "loss": 0.6934,
+      "step": 1821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.394766724826576,
+      "learning_rate": 1.9968943623610795e-06,
+      "loss": 0.681,
+      "step": 1822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9651335084166299,
+      "learning_rate": 1.996886749794476e-06,
+      "loss": 0.6823,
+      "step": 1823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1569368874526456,
+      "learning_rate": 1.99687912792384e-06,
+      "loss": 0.6868,
+      "step": 1824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5905814838115874,
+      "learning_rate": 1.9968714967492432e-06,
+      "loss": 0.6908,
+      "step": 1825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.285263884842644,
+      "learning_rate": 1.9968638562707563e-06,
+      "loss": 0.6829,
+      "step": 1826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.631035020141265,
+      "learning_rate": 1.996856206488451e-06,
+      "loss": 0.6836,
+      "step": 1827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.858558247776294,
+      "learning_rate": 1.996848547402399e-06,
+      "loss": 0.7253,
+      "step": 1828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.13746622115907,
+      "learning_rate": 1.9968408790126706e-06,
+      "loss": 0.7155,
+      "step": 1829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.976536693651665,
+      "learning_rate": 1.9968332013193383e-06,
+      "loss": 0.7148,
+      "step": 1830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.758264684544402,
+      "learning_rate": 1.996825514322474e-06,
+      "loss": 0.6862,
+      "step": 1831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.062785539905678,
+      "learning_rate": 1.9968178180221488e-06,
+      "loss": 0.6966,
+      "step": 1832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4295458426529652,
+      "learning_rate": 1.996810112418435e-06,
+      "loss": 0.6908,
+      "step": 1833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.286709703183017,
+      "learning_rate": 1.996802397511404e-06,
+      "loss": 0.6849,
+      "step": 1834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.729585664792097,
+      "learning_rate": 1.996794673301128e-06,
+      "loss": 0.7064,
+      "step": 1835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.74636538924312,
+      "learning_rate": 1.996786939787679e-06,
+      "loss": 0.7005,
+      "step": 1836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.120442166241165,
+      "learning_rate": 1.9967791969711296e-06,
+      "loss": 0.7142,
+      "step": 1837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.23511997923501,
+      "learning_rate": 1.9967714448515517e-06,
+      "loss": 0.7122,
+      "step": 1838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.815848118828664,
+      "learning_rate": 1.9967636834290177e-06,
+      "loss": 0.6868,
+      "step": 1839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 15.886435585425918,
+      "learning_rate": 1.9967559127036005e-06,
+      "loss": 0.7298,
+      "step": 1840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.102158989199764,
+      "learning_rate": 1.996748132675372e-06,
+      "loss": 0.6947,
+      "step": 1841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7854044855762727,
+      "learning_rate": 1.9967403433444047e-06,
+      "loss": 0.6777,
+      "step": 1842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.005878827771043,
+      "learning_rate": 1.9967325447107722e-06,
+      "loss": 0.6823,
+      "step": 1843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.147309681410732,
+      "learning_rate": 1.9967247367745462e-06,
+      "loss": 0.6882,
+      "step": 1844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8845538128866177,
+      "learning_rate": 1.9967169195358004e-06,
+      "loss": 0.6862,
+      "step": 1845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7872471926071163,
+      "learning_rate": 1.9967090929946075e-06,
+      "loss": 0.6849,
+      "step": 1846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6011196270162839,
+      "learning_rate": 1.99670125715104e-06,
+      "loss": 0.6862,
+      "step": 1847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.13816836093163,
+      "learning_rate": 1.996693412005172e-06,
+      "loss": 0.6868,
+      "step": 1848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.477994693742774,
+      "learning_rate": 1.996685557557076e-06,
+      "loss": 0.6999,
+      "step": 1849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.019945867455358,
+      "learning_rate": 1.9966776938068257e-06,
+      "loss": 0.6816,
+      "step": 1850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1869799744972072,
+      "learning_rate": 1.9966698207544943e-06,
+      "loss": 0.696,
+      "step": 1851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4433053678201366,
+      "learning_rate": 1.9966619384001554e-06,
+      "loss": 0.6855,
+      "step": 1852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.564802204959158,
+      "learning_rate": 1.9966540467438824e-06,
+      "loss": 0.6953,
+      "step": 1853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.688522866277068,
+      "learning_rate": 1.996646145785749e-06,
+      "loss": 0.6771,
+      "step": 1854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.32650200054695,
+      "learning_rate": 1.9966382355258287e-06,
+      "loss": 0.7018,
+      "step": 1855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.927511319762765,
+      "learning_rate": 1.9966303159641963e-06,
+      "loss": 0.7064,
+      "step": 1856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6620599294334943,
+      "learning_rate": 1.9966223871009245e-06,
+      "loss": 0.6914,
+      "step": 1857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.118484071102985,
+      "learning_rate": 1.9966144489360883e-06,
+      "loss": 0.7018,
+      "step": 1858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.401468656066619,
+      "learning_rate": 1.9966065014697605e-06,
+      "loss": 0.6855,
+      "step": 1859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.751733145682448,
+      "learning_rate": 1.996598544702017e-06,
+      "loss": 0.7038,
+      "step": 1860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.668417159615413,
+      "learning_rate": 1.9965905786329305e-06,
+      "loss": 0.6979,
+      "step": 1861
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.908337431090505,
+      "learning_rate": 1.996582603262576e-06,
+      "loss": 0.6908,
+      "step": 1862
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.9360482119529925,
+      "learning_rate": 1.9965746185910284e-06,
+      "loss": 0.6868,
+      "step": 1863
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1449463908543964,
+      "learning_rate": 1.9965666246183613e-06,
+      "loss": 0.6888,
+      "step": 1864
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.369760831957967,
+      "learning_rate": 1.9965586213446503e-06,
+      "loss": 0.6888,
+      "step": 1865
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.5196420983631027,
+      "learning_rate": 1.996550608769969e-06,
+      "loss": 0.6966,
+      "step": 1866
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.834207002087839,
+      "learning_rate": 1.996542586894393e-06,
+      "loss": 0.6895,
+      "step": 1867
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.470202003913432,
+      "learning_rate": 1.9965345557179967e-06,
+      "loss": 0.6999,
+      "step": 1868
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.086118735865163,
+      "learning_rate": 1.9965265152408555e-06,
+      "loss": 0.7018,
+      "step": 1869
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.9676629199414934,
+      "learning_rate": 1.9965184654630444e-06,
+      "loss": 0.6921,
+      "step": 1870
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9474743772134738,
+      "learning_rate": 1.996510406384638e-06,
+      "loss": 0.6914,
+      "step": 1871
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.126440495953518,
+      "learning_rate": 1.996502338005712e-06,
+      "loss": 0.681,
+      "step": 1872
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0433778154676114,
+      "learning_rate": 1.9964942603263417e-06,
+      "loss": 0.6966,
+      "step": 1873
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.699484871147893,
+      "learning_rate": 1.996486173346602e-06,
+      "loss": 0.6914,
+      "step": 1874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.384461529025496,
+      "learning_rate": 1.9964780770665686e-06,
+      "loss": 0.6979,
+      "step": 1875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2764021990052121,
+      "learning_rate": 1.9964699714863175e-06,
+      "loss": 0.7025,
+      "step": 1876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7493988205425888,
+      "learning_rate": 1.9964618566059237e-06,
+      "loss": 0.6901,
+      "step": 1877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.615755667870812,
+      "learning_rate": 1.996453732425464e-06,
+      "loss": 0.6888,
+      "step": 1878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3199750627847446,
+      "learning_rate": 1.9964455989450127e-06,
+      "loss": 0.6862,
+      "step": 1879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2389520369249323,
+      "learning_rate": 1.996437456164647e-06,
+      "loss": 0.6986,
+      "step": 1880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.122331511452429,
+      "learning_rate": 1.996429304084442e-06,
+      "loss": 0.6862,
+      "step": 1881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.143603309074773,
+      "learning_rate": 1.9964211427044746e-06,
+      "loss": 0.7077,
+      "step": 1882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.930489202351744,
+      "learning_rate": 1.9964129720248203e-06,
+      "loss": 0.6927,
+      "step": 1883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6402510000867419,
+      "learning_rate": 1.9964047920455556e-06,
+      "loss": 0.6868,
+      "step": 1884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5645001749695555,
+      "learning_rate": 1.9963966027667573e-06,
+      "loss": 0.6882,
+      "step": 1885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.693478430769394,
+      "learning_rate": 1.9963884041885008e-06,
+      "loss": 0.6901,
+      "step": 1886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.19566290009337,
+      "learning_rate": 1.9963801963108634e-06,
+      "loss": 0.6973,
+      "step": 1887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6351791331936364,
+      "learning_rate": 1.9963719791339216e-06,
+      "loss": 0.6829,
+      "step": 1888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.333981404910573,
+      "learning_rate": 1.996363752657752e-06,
+      "loss": 0.6849,
+      "step": 1889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.61981848025029,
+      "learning_rate": 1.9963555168824314e-06,
+      "loss": 0.6973,
+      "step": 1890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.460621659705765,
+      "learning_rate": 1.9963472718080367e-06,
+      "loss": 0.6979,
+      "step": 1891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.685836276849453,
+      "learning_rate": 1.996339017434644e-06,
+      "loss": 0.6868,
+      "step": 1892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.239662494035462,
+      "learning_rate": 1.996330753762332e-06,
+      "loss": 0.6908,
+      "step": 1893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.603496469697187,
+      "learning_rate": 1.996322480791177e-06,
+      "loss": 0.6855,
+      "step": 1894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.056147299474493,
+      "learning_rate": 1.996314198521256e-06,
+      "loss": 0.6986,
+      "step": 1895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.338204910773419,
+      "learning_rate": 1.9963059069526464e-06,
+      "loss": 0.6803,
+      "step": 1896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.626400949663979,
+      "learning_rate": 1.9962976060854257e-06,
+      "loss": 0.6921,
+      "step": 1897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8845670711338506,
+      "learning_rate": 1.996289295919671e-06,
+      "loss": 0.6855,
+      "step": 1898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.84389074042448,
+      "learning_rate": 1.996280976455461e-06,
+      "loss": 0.6868,
+      "step": 1899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.828585062150965,
+      "learning_rate": 1.996272647692872e-06,
+      "loss": 0.6849,
+      "step": 1900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9504482483595247,
+      "learning_rate": 1.996264309631982e-06,
+      "loss": 0.707,
+      "step": 1901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.670174495820472,
+      "learning_rate": 1.9962559622728695e-06,
+      "loss": 0.707,
+      "step": 1902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.086282584101712,
+      "learning_rate": 1.996247605615612e-06,
+      "loss": 0.707,
+      "step": 1903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.83719911486593,
+      "learning_rate": 1.9962392396602878e-06,
+      "loss": 0.6914,
+      "step": 1904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0044266005143085,
+      "learning_rate": 1.996230864406974e-06,
+      "loss": 0.6895,
+      "step": 1905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3680255517894664,
+      "learning_rate": 1.9962224798557497e-06,
+      "loss": 0.6849,
+      "step": 1906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1563352017662876,
+      "learning_rate": 1.996214086006693e-06,
+      "loss": 0.7025,
+      "step": 1907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5163612737206382,
+      "learning_rate": 1.996205682859882e-06,
+      "loss": 0.6875,
+      "step": 1908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6280165036848278,
+      "learning_rate": 1.9961972704153952e-06,
+      "loss": 0.6927,
+      "step": 1909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.261634276743366,
+      "learning_rate": 1.9961888486733114e-06,
+      "loss": 0.707,
+      "step": 1910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.353699290908804,
+      "learning_rate": 1.996180417633709e-06,
+      "loss": 0.6953,
+      "step": 1911
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.891270007416212,
+      "learning_rate": 1.9961719772966664e-06,
+      "loss": 0.6973,
+      "step": 1912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8728025191878053,
+      "learning_rate": 1.996163527662263e-06,
+      "loss": 0.6842,
+      "step": 1913
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5075655048400134,
+      "learning_rate": 1.996155068730577e-06,
+      "loss": 0.6921,
+      "step": 1914
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.86181834334752,
+      "learning_rate": 1.9961466005016878e-06,
+      "loss": 0.6829,
+      "step": 1915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1339522825491555,
+      "learning_rate": 1.9961381229756742e-06,
+      "loss": 0.6882,
+      "step": 1916
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.620767716545684,
+      "learning_rate": 1.996129636152615e-06,
+      "loss": 0.6849,
+      "step": 1917
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3745328946192923,
+      "learning_rate": 1.9961211400325907e-06,
+      "loss": 0.6842,
+      "step": 1918
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.492692221195482,
+      "learning_rate": 1.996112634615679e-06,
+      "loss": 0.6849,
+      "step": 1919
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.784627357594357,
+      "learning_rate": 1.9961041199019605e-06,
+      "loss": 0.6966,
+      "step": 1920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7349687338935216,
+      "learning_rate": 1.9960955958915137e-06,
+      "loss": 0.6888,
+      "step": 1921
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.160622704525065,
+      "learning_rate": 1.996087062584419e-06,
+      "loss": 0.6914,
+      "step": 1922
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.276249184779373,
+      "learning_rate": 1.9960785199807552e-06,
+      "loss": 0.6803,
+      "step": 1923
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.15907767817942,
+      "learning_rate": 1.996069968080603e-06,
+      "loss": 0.6797,
+      "step": 1924
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.453320320756302,
+      "learning_rate": 1.9960614068840413e-06,
+      "loss": 0.7005,
+      "step": 1925
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5908392047449198,
+      "learning_rate": 1.9960528363911508e-06,
+      "loss": 0.6751,
+      "step": 1926
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.4526139530227375,
+      "learning_rate": 1.9960442566020105e-06,
+      "loss": 0.6947,
+      "step": 1927
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7460881879518189,
+      "learning_rate": 1.9960356675167015e-06,
+      "loss": 0.6895,
+      "step": 1928
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.096238702013258,
+      "learning_rate": 1.9960270691353032e-06,
+      "loss": 0.694,
+      "step": 1929
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5749894085344498,
+      "learning_rate": 1.9960184614578967e-06,
+      "loss": 0.6927,
+      "step": 1930
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9218377573638525,
+      "learning_rate": 1.9960098444845615e-06,
+      "loss": 0.6888,
+      "step": 1931
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7019652516838155,
+      "learning_rate": 1.9960012182153785e-06,
+      "loss": 0.7044,
+      "step": 1932
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7254194187159759,
+      "learning_rate": 1.9959925826504277e-06,
+      "loss": 0.6849,
+      "step": 1933
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.856728572214746,
+      "learning_rate": 1.99598393778979e-06,
+      "loss": 0.6849,
+      "step": 1934
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.1163379436324,
+      "learning_rate": 1.9959752836335466e-06,
+      "loss": 0.6836,
+      "step": 1935
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.708432906291821,
+      "learning_rate": 1.9959666201817776e-06,
+      "loss": 0.6895,
+      "step": 1936
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7710740266396161,
+      "learning_rate": 1.995957947434564e-06,
+      "loss": 0.6966,
+      "step": 1937
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1705774608443935,
+      "learning_rate": 1.995949265391987e-06,
+      "loss": 0.6771,
+      "step": 1938
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3172252601989016,
+      "learning_rate": 1.9959405740541273e-06,
+      "loss": 0.6647,
+      "step": 1939
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8530515447323538,
+      "learning_rate": 1.995931873421066e-06,
+      "loss": 0.679,
+      "step": 1940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.295527182394645,
+      "learning_rate": 1.9959231634928847e-06,
+      "loss": 0.6823,
+      "step": 1941
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.973357170008156,
+      "learning_rate": 1.995914444269664e-06,
+      "loss": 0.7044,
+      "step": 1942
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.261028914164652,
+      "learning_rate": 1.9959057157514863e-06,
+      "loss": 0.6862,
+      "step": 1943
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.96327467302479,
+      "learning_rate": 1.9958969779384322e-06,
+      "loss": 0.679,
+      "step": 1944
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.4031760656654,
+      "learning_rate": 1.9958882308305832e-06,
+      "loss": 0.6973,
+      "step": 1945
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3500851769793778,
+      "learning_rate": 1.995879474428022e-06,
+      "loss": 0.6855,
+      "step": 1946
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.234159359955198,
+      "learning_rate": 1.995870708730829e-06,
+      "loss": 0.6901,
+      "step": 1947
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2323204904741218,
+      "learning_rate": 1.9958619337390868e-06,
+      "loss": 0.6882,
+      "step": 1948
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.809498119086575,
+      "learning_rate": 1.995853149452877e-06,
+      "loss": 0.6999,
+      "step": 1949
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.557884500079079,
+      "learning_rate": 1.995844355872282e-06,
+      "loss": 0.724,
+      "step": 1950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7443323879835331,
+      "learning_rate": 1.9958355529973832e-06,
+      "loss": 0.6862,
+      "step": 1951
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6141509585968927,
+      "learning_rate": 1.9958267408282636e-06,
+      "loss": 0.6842,
+      "step": 1952
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.030571982065727,
+      "learning_rate": 1.9958179193650046e-06,
+      "loss": 0.6771,
+      "step": 1953
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.905846908170793,
+      "learning_rate": 1.9958090886076885e-06,
+      "loss": 0.6895,
+      "step": 1954
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0654355593027525,
+      "learning_rate": 1.995800248556399e-06,
+      "loss": 0.6849,
+      "step": 1955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7699487990756637,
+      "learning_rate": 1.995791399211217e-06,
+      "loss": 0.7018,
+      "step": 1956
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.424491443056506,
+      "learning_rate": 1.995782540572226e-06,
+      "loss": 0.6803,
+      "step": 1957
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.630259567715457,
+      "learning_rate": 1.995773672639509e-06,
+      "loss": 0.6868,
+      "step": 1958
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7125763102251735,
+      "learning_rate": 1.9957647954131474e-06,
+      "loss": 0.6868,
+      "step": 1959
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6909407567367195,
+      "learning_rate": 1.9957559088932256e-06,
+      "loss": 0.6973,
+      "step": 1960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.747246562417498,
+      "learning_rate": 1.995747013079825e-06,
+      "loss": 0.6816,
+      "step": 1961
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.952571070855079,
+      "learning_rate": 1.99573810797303e-06,
+      "loss": 0.6868,
+      "step": 1962
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.423353296809203,
+      "learning_rate": 1.9957291935729235e-06,
+      "loss": 0.6934,
+      "step": 1963
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.657788276197407,
+      "learning_rate": 1.995720269879588e-06,
+      "loss": 0.6979,
+      "step": 1964
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.284360413534841,
+      "learning_rate": 1.995711336893107e-06,
+      "loss": 0.6966,
+      "step": 1965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.912613259632315,
+      "learning_rate": 1.9957023946135646e-06,
+      "loss": 0.6999,
+      "step": 1966
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7658659288805398,
+      "learning_rate": 1.9956934430410437e-06,
+      "loss": 0.6868,
+      "step": 1967
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.919863011119986,
+      "learning_rate": 1.9956844821756272e-06,
+      "loss": 0.6797,
+      "step": 1968
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.393153776045875,
+      "learning_rate": 1.9956755120173997e-06,
+      "loss": 0.6947,
+      "step": 1969
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.719280163340467,
+      "learning_rate": 1.9956665325664448e-06,
+      "loss": 0.6882,
+      "step": 1970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.678259680279034,
+      "learning_rate": 1.9956575438228455e-06,
+      "loss": 0.6836,
+      "step": 1971
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.868006440444137,
+      "learning_rate": 1.9956485457866867e-06,
+      "loss": 0.6927,
+      "step": 1972
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.747126413998155,
+      "learning_rate": 1.995639538458052e-06,
+      "loss": 0.6875,
+      "step": 1973
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8835950350061261,
+      "learning_rate": 1.9956305218370253e-06,
+      "loss": 0.6908,
+      "step": 1974
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.85570412564342,
+      "learning_rate": 1.995621495923691e-06,
+      "loss": 0.6882,
+      "step": 1975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.857765191539288,
+      "learning_rate": 1.9956124607181335e-06,
+      "loss": 0.6888,
+      "step": 1976
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8433322633419253,
+      "learning_rate": 1.995603416220436e-06,
+      "loss": 0.6888,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.145107297270967,
+      "learning_rate": 1.9955943624306846e-06,
+      "loss": 0.6895,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7693732202829348,
+      "learning_rate": 1.9955852993489625e-06,
+      "loss": 0.6836,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9277512244403014,
+      "learning_rate": 1.995576226975355e-06,
+      "loss": 0.6829,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4254109300355733,
+      "learning_rate": 1.9955671453099464e-06,
+      "loss": 0.6875,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.054021916369393,
+      "learning_rate": 1.995558054352822e-06,
+      "loss": 0.6901,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0560488274605175,
+      "learning_rate": 1.9955489541040657e-06,
+      "loss": 0.6953,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8643900959065074,
+      "learning_rate": 1.9955398445637634e-06,
+      "loss": 0.6875,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.504174284950175,
+      "learning_rate": 1.9955307257319993e-06,
+      "loss": 0.6803,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.615904781152501,
+      "learning_rate": 1.9955215976088592e-06,
+      "loss": 0.6888,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4978601550127804,
+      "learning_rate": 1.9955124601944276e-06,
+      "loss": 0.6849,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.082485108952971,
+      "learning_rate": 1.9955033134887908e-06,
+      "loss": 0.6947,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.032246435371706,
+      "learning_rate": 1.995494157492033e-06,
+      "loss": 0.6895,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7230928404605168,
+      "learning_rate": 1.9954849922042404e-06,
+      "loss": 0.6823,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7925944690981708,
+      "learning_rate": 1.9954758176254983e-06,
+      "loss": 0.6888,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.525433378470288,
+      "learning_rate": 1.995466633755892e-06,
+      "loss": 0.7025,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6783022990016727,
+      "learning_rate": 1.995457440595508e-06,
+      "loss": 0.6771,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.082143867118819,
+      "learning_rate": 1.9954482381444317e-06,
+      "loss": 0.694,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8944470797360826,
+      "learning_rate": 1.995439026402749e-06,
+      "loss": 0.6921,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.990696838661297,
+      "learning_rate": 1.9954298053705456e-06,
+      "loss": 0.7272,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.72717974060425,
+      "learning_rate": 1.995420575047908e-06,
+      "loss": 0.7096,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.803881545073114,
+      "learning_rate": 1.995411335434922e-06,
+      "loss": 0.7051,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.156991078650013,
+      "learning_rate": 1.995402086531674e-06,
+      "loss": 0.7077,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.152033311169977,
+      "learning_rate": 1.99539282833825e-06,
+      "loss": 0.6882,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9092221035453314,
+      "learning_rate": 1.995383560854737e-06,
+      "loss": 0.6966,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.547844913546583,
+      "learning_rate": 1.9953742840812214e-06,
+      "loss": 0.696,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9034034601832075,
+      "learning_rate": 1.9953649980177892e-06,
+      "loss": 0.6895,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.361358256025763,
+      "learning_rate": 1.995355702664528e-06,
+      "loss": 0.724,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2372921413990507,
+      "learning_rate": 1.9953463980215233e-06,
+      "loss": 0.6875,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.45747902310465,
+      "learning_rate": 1.995337084088863e-06,
+      "loss": 0.7051,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.113875162550306,
+      "learning_rate": 1.995327760866634e-06,
+      "loss": 0.6875,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.04517824435416,
+      "learning_rate": 1.995318428354922e-06,
+      "loss": 0.6849,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.920224868879356,
+      "learning_rate": 1.995309086553816e-06,
+      "loss": 0.6921,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.5067930457413405,
+      "learning_rate": 1.995299735463402e-06,
+      "loss": 0.6882,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.359690449831231,
+      "learning_rate": 1.9952903750837674e-06,
+      "loss": 0.6875,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.455096117509393,
+      "learning_rate": 1.995281005415e-06,
+      "loss": 0.694,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.231817309036877,
+      "learning_rate": 1.995271626457187e-06,
+      "loss": 0.6966,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.680034279443335,
+      "learning_rate": 1.9952622382104155e-06,
+      "loss": 0.709,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.568553459686599,
+      "learning_rate": 1.9952528406747738e-06,
+      "loss": 0.6953,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.363112363038669,
+      "learning_rate": 1.9952434338503494e-06,
+      "loss": 0.6914,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.092192954961947,
+      "learning_rate": 1.9952340177372295e-06,
+      "loss": 0.6966,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.118839387464089,
+      "learning_rate": 1.9952245923355032e-06,
+      "loss": 0.6829,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7804083239797605,
+      "learning_rate": 1.995215157645257e-06,
+      "loss": 0.6868,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.160734921234478,
+      "learning_rate": 1.9952057136665805e-06,
+      "loss": 0.696,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.257677240953326,
+      "learning_rate": 1.9951962603995606e-06,
+      "loss": 0.7005,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.639954529516167,
+      "learning_rate": 1.995186797844286e-06,
+      "loss": 0.7116,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.286008533981664,
+      "learning_rate": 1.995177326000845e-06,
+      "loss": 0.6927,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.762902588660575,
+      "learning_rate": 1.9951678448693262e-06,
+      "loss": 0.6868,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.756253486191087,
+      "learning_rate": 1.995158354449818e-06,
+      "loss": 0.707,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.345667565943277,
+      "learning_rate": 1.9951488547424084e-06,
+      "loss": 0.6816,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.725159893646314,
+      "learning_rate": 1.995139345747187e-06,
+      "loss": 0.7012,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.311623236822453,
+      "learning_rate": 1.995129827464242e-06,
+      "loss": 0.6823,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.5340475090147985,
+      "learning_rate": 1.995120299893662e-06,
+      "loss": 0.696,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.487341278465876,
+      "learning_rate": 1.9951107630355364e-06,
+      "loss": 0.6823,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8832009858088183,
+      "learning_rate": 1.995101216889954e-06,
+      "loss": 0.6986,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5676434378814575,
+      "learning_rate": 1.9950916614570042e-06,
+      "loss": 0.6895,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.761581428312644,
+      "learning_rate": 1.9950820967367755e-06,
+      "loss": 0.6953,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.643963584080869,
+      "learning_rate": 1.995072522729358e-06,
+      "loss": 0.6921,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.125433435534502,
+      "learning_rate": 1.9950629394348406e-06,
+      "loss": 0.6966,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.585322606386857,
+      "learning_rate": 1.9950533468533126e-06,
+      "loss": 0.6966,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5812881738918659,
+      "learning_rate": 1.9950437449848637e-06,
+      "loss": 0.6888,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.384283057275008,
+      "learning_rate": 1.9950341338295836e-06,
+      "loss": 0.6901,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6341321847754362,
+      "learning_rate": 1.995024513387562e-06,
+      "loss": 0.694,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.001355182830089,
+      "learning_rate": 1.9950148836588886e-06,
+      "loss": 0.694,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7124265257553761,
+      "learning_rate": 1.995005244643653e-06,
+      "loss": 0.6816,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.317650816756803,
+      "learning_rate": 1.994995596341946e-06,
+      "loss": 0.6882,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.054312913684877,
+      "learning_rate": 1.9949859387538568e-06,
+      "loss": 0.6921,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4258608366876286,
+      "learning_rate": 1.994976271879476e-06,
+      "loss": 0.6882,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6275386363189513,
+      "learning_rate": 1.9949665957188934e-06,
+      "loss": 0.6855,
+      "step": 2045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8862428319677573,
+      "learning_rate": 1.9949569102721997e-06,
+      "loss": 0.6758,
+      "step": 2046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.298749701967591,
+      "learning_rate": 1.994947215539485e-06,
+      "loss": 0.6973,
+      "step": 2047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8185325745705474,
+      "learning_rate": 1.9949375115208405e-06,
+      "loss": 0.6934,
+      "step": 2048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4503573954041318,
+      "learning_rate": 1.9949277982163555e-06,
+      "loss": 0.6855,
+      "step": 2049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.2956205415081365,
+      "learning_rate": 1.994918075626122e-06,
+      "loss": 0.6797,
+      "step": 2050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2738551365923259,
+      "learning_rate": 1.99490834375023e-06,
+      "loss": 0.6829,
+      "step": 2051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.600001360228916,
+      "learning_rate": 1.99489860258877e-06,
+      "loss": 0.6875,
+      "step": 2052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.183156788966657,
+      "learning_rate": 1.9948888521418335e-06,
+      "loss": 0.6823,
+      "step": 2053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.068161408867821,
+      "learning_rate": 1.9948790924095118e-06,
+      "loss": 0.6999,
+      "step": 2054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7982894192473213,
+      "learning_rate": 1.994869323391895e-06,
+      "loss": 0.6992,
+      "step": 2055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.863045846447953,
+      "learning_rate": 1.994859545089075e-06,
+      "loss": 0.694,
+      "step": 2056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.846845818957937,
+      "learning_rate": 1.9948497575011427e-06,
+      "loss": 0.7038,
+      "step": 2057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7758328000616714,
+      "learning_rate": 1.99483996062819e-06,
+      "loss": 0.6908,
+      "step": 2058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4756741427466027,
+      "learning_rate": 1.994830154470308e-06,
+      "loss": 0.6764,
+      "step": 2059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.592956030895011,
+      "learning_rate": 1.994820339027588e-06,
+      "loss": 0.6823,
+      "step": 2060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4456124638029932,
+      "learning_rate": 1.9948105143001217e-06,
+      "loss": 0.6803,
+      "step": 2061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.267826905128766,
+      "learning_rate": 1.9948006802880012e-06,
+      "loss": 0.7025,
+      "step": 2062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1525362073924537,
+      "learning_rate": 1.994790836991318e-06,
+      "loss": 0.6784,
+      "step": 2063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.663751417688394,
+      "learning_rate": 1.9947809844101634e-06,
+      "loss": 0.6849,
+      "step": 2064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.383155322468333,
+      "learning_rate": 1.9947711225446303e-06,
+      "loss": 0.7129,
+      "step": 2065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7062354148819434,
+      "learning_rate": 1.9947612513948106e-06,
+      "loss": 0.6868,
+      "step": 2066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.2864146226542745,
+      "learning_rate": 1.9947513709607958e-06,
+      "loss": 0.6895,
+      "step": 2067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.443775392294684,
+      "learning_rate": 1.9947414812426785e-06,
+      "loss": 0.6849,
+      "step": 2068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6710792873400595,
+      "learning_rate": 1.994731582240551e-06,
+      "loss": 0.6875,
+      "step": 2069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8348885362936054,
+      "learning_rate": 1.9947216739545056e-06,
+      "loss": 0.6895,
+      "step": 2070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.133164318366473,
+      "learning_rate": 1.994711756384635e-06,
+      "loss": 0.7057,
+      "step": 2071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.770695876380639,
+      "learning_rate": 1.9947018295310315e-06,
+      "loss": 0.696,
+      "step": 2072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.818196140200967,
+      "learning_rate": 1.994691893393788e-06,
+      "loss": 0.6966,
+      "step": 2073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0202607611252796,
+      "learning_rate": 1.994681947972997e-06,
+      "loss": 0.6868,
+      "step": 2074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.040139442168075,
+      "learning_rate": 1.9946719932687514e-06,
+      "loss": 0.6829,
+      "step": 2075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.712751633177436,
+      "learning_rate": 1.994662029281144e-06,
+      "loss": 0.6914,
+      "step": 2076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6823353807143185,
+      "learning_rate": 1.994652056010268e-06,
+      "loss": 0.6914,
+      "step": 2077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1601387137138648,
+      "learning_rate": 1.994642073456217e-06,
+      "loss": 0.6777,
+      "step": 2078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.490106671360172,
+      "learning_rate": 1.9946320816190826e-06,
+      "loss": 0.6862,
+      "step": 2079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.3448887702269285,
+      "learning_rate": 1.9946220804989596e-06,
+      "loss": 0.6882,
+      "step": 2080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.424001310876028,
+      "learning_rate": 1.9946120700959408e-06,
+      "loss": 0.7005,
+      "step": 2081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.155785122024268,
+      "learning_rate": 1.994602050410119e-06,
+      "loss": 0.6895,
+      "step": 2082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.749798656685194,
+      "learning_rate": 1.994592021441589e-06,
+      "loss": 0.7402,
+      "step": 2083
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.578842641677133,
+      "learning_rate": 1.9945819831904434e-06,
+      "loss": 0.694,
+      "step": 2084
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.700416161048151,
+      "learning_rate": 1.994571935656776e-06,
+      "loss": 0.6862,
+      "step": 2085
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.353735769810609,
+      "learning_rate": 1.994561878840681e-06,
+      "loss": 0.6934,
+      "step": 2086
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7395966661774404,
+      "learning_rate": 1.994551812742252e-06,
+      "loss": 0.6973,
+      "step": 2087
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.272285523246088,
+      "learning_rate": 1.9945417373615833e-06,
+      "loss": 0.7064,
+      "step": 2088
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.584929317478945,
+      "learning_rate": 1.994531652698768e-06,
+      "loss": 0.7044,
+      "step": 2089
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.085380799831157,
+      "learning_rate": 1.994521558753901e-06,
+      "loss": 0.681,
+      "step": 2090
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3346828712935213,
+      "learning_rate": 1.9945114555270767e-06,
+      "loss": 0.6647,
+      "step": 2091
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4129533608897633,
+      "learning_rate": 1.9945013430183887e-06,
+      "loss": 0.6927,
+      "step": 2092
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.808605288489463,
+      "learning_rate": 1.994491221227932e-06,
+      "loss": 0.7214,
+      "step": 2093
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.878503302309573,
+      "learning_rate": 1.9944810901558006e-06,
+      "loss": 0.7116,
+      "step": 2094
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.912496176473203,
+      "learning_rate": 1.994470949802089e-06,
+      "loss": 0.7051,
+      "step": 2095
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.396770571714642,
+      "learning_rate": 1.9944608001668924e-06,
+      "loss": 0.7057,
+      "step": 2096
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5606462826800245,
+      "learning_rate": 1.994450641250305e-06,
+      "loss": 0.6882,
+      "step": 2097
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.329153344028439,
+      "learning_rate": 1.994440473052422e-06,
+      "loss": 0.6927,
+      "step": 2098
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6605678929078531,
+      "learning_rate": 1.994430295573338e-06,
+      "loss": 0.6888,
+      "step": 2099
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.113261500299247,
+      "learning_rate": 1.9944201088131487e-06,
+      "loss": 0.6914,
+      "step": 2100
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5880637594148963,
+      "learning_rate": 1.994409912771948e-06,
+      "loss": 0.6921,
+      "step": 2101
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2798923135290963,
+      "learning_rate": 1.9943997074498315e-06,
+      "loss": 0.6816,
+      "step": 2102
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.654119098771072,
+      "learning_rate": 1.994389492846895e-06,
+      "loss": 0.6803,
+      "step": 2103
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.199313599514428,
+      "learning_rate": 1.994379268963233e-06,
+      "loss": 0.6947,
+      "step": 2104
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.851114309405642,
+      "learning_rate": 1.9943690357989417e-06,
+      "loss": 0.7031,
+      "step": 2105
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.220419920568698,
+      "learning_rate": 1.994358793354116e-06,
+      "loss": 0.681,
+      "step": 2106
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.98584162773109,
+      "learning_rate": 1.994348541628852e-06,
+      "loss": 0.7083,
+      "step": 2107
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0433817876059175,
+      "learning_rate": 1.994338280623245e-06,
+      "loss": 0.6921,
+      "step": 2108
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7915302427096433,
+      "learning_rate": 1.994328010337391e-06,
+      "loss": 0.6836,
+      "step": 2109
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6399370438282699,
+      "learning_rate": 1.994317730771385e-06,
+      "loss": 0.7064,
+      "step": 2110
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5503988747216986,
+      "learning_rate": 1.994307441925325e-06,
+      "loss": 0.6882,
+      "step": 2111
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.597942605622542,
+      "learning_rate": 1.9942971437993048e-06,
+      "loss": 0.6777,
+      "step": 2112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6389595284438983,
+      "learning_rate": 1.9942868363934216e-06,
+      "loss": 0.6823,
+      "step": 2113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.112761335468765,
+      "learning_rate": 1.994276519707771e-06,
+      "loss": 0.694,
+      "step": 2114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.256079865361053,
+      "learning_rate": 1.9942661937424504e-06,
+      "loss": 0.6927,
+      "step": 2115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.989820918350867,
+      "learning_rate": 1.9942558584975555e-06,
+      "loss": 0.694,
+      "step": 2116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5021041033888342,
+      "learning_rate": 1.9942455139731826e-06,
+      "loss": 0.6784,
+      "step": 2117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.211411721378317,
+      "learning_rate": 1.994235160169428e-06,
+      "loss": 0.7083,
+      "step": 2118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0393163733592456,
+      "learning_rate": 1.9942247970863893e-06,
+      "loss": 0.6927,
+      "step": 2119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6852619063555013,
+      "learning_rate": 1.9942144247241625e-06,
+      "loss": 0.6999,
+      "step": 2120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.776295997754243,
+      "learning_rate": 1.9942040430828443e-06,
+      "loss": 0.6888,
+      "step": 2121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9254284931912148,
+      "learning_rate": 1.994193652162532e-06,
+      "loss": 0.681,
+      "step": 2122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.543272595122522,
+      "learning_rate": 1.9941832519633223e-06,
+      "loss": 0.696,
+      "step": 2123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.049792294444601,
+      "learning_rate": 1.9941728424853125e-06,
+      "loss": 0.7012,
+      "step": 2124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.382386896432917,
+      "learning_rate": 1.9941624237285996e-06,
+      "loss": 0.6842,
+      "step": 2125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7955270672325234,
+      "learning_rate": 1.9941519956932807e-06,
+      "loss": 0.6875,
+      "step": 2126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7404221622733624,
+      "learning_rate": 1.9941415583794537e-06,
+      "loss": 0.6901,
+      "step": 2127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5555171105404972,
+      "learning_rate": 1.9941311117872155e-06,
+      "loss": 0.6953,
+      "step": 2128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5700488862188897,
+      "learning_rate": 1.994120655916664e-06,
+      "loss": 0.6934,
+      "step": 2129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.52458283869441,
+      "learning_rate": 1.9941101907678955e-06,
+      "loss": 0.6966,
+      "step": 2130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.149561956964611,
+      "learning_rate": 1.9940997163410095e-06,
+      "loss": 0.6888,
+      "step": 2131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.502647042932812,
+      "learning_rate": 1.9940892326361027e-06,
+      "loss": 0.6732,
+      "step": 2132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1024101252463365,
+      "learning_rate": 1.994078739653273e-06,
+      "loss": 0.6836,
+      "step": 2133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.745992811628698,
+      "learning_rate": 1.9940682373926192e-06,
+      "loss": 0.7148,
+      "step": 2134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.299384507796885,
+      "learning_rate": 1.994057725854238e-06,
+      "loss": 0.7005,
+      "step": 2135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.535001025142034,
+      "learning_rate": 1.994047205038228e-06,
+      "loss": 0.6973,
+      "step": 2136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.600565469445625,
+      "learning_rate": 1.9940366749446884e-06,
+      "loss": 0.6745,
+      "step": 2137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2039056649276487,
+      "learning_rate": 1.9940261355737157e-06,
+      "loss": 0.6979,
+      "step": 2138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7447022415051809,
+      "learning_rate": 1.9940155869254096e-06,
+      "loss": 0.6895,
+      "step": 2139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.523029006600892,
+      "learning_rate": 1.994005028999868e-06,
+      "loss": 0.696,
+      "step": 2140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.493312344079601,
+      "learning_rate": 1.9939944617971897e-06,
+      "loss": 0.6901,
+      "step": 2141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.607744978245882,
+      "learning_rate": 1.9939838853174732e-06,
+      "loss": 0.6855,
+      "step": 2142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6261402375276124,
+      "learning_rate": 1.993973299560817e-06,
+      "loss": 0.6745,
+      "step": 2143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8889716687456835,
+      "learning_rate": 1.99396270452732e-06,
+      "loss": 0.6771,
+      "step": 2144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8106358285679502,
+      "learning_rate": 1.993952100217081e-06,
+      "loss": 0.6953,
+      "step": 2145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.352645258122751,
+      "learning_rate": 1.9939414866302e-06,
+      "loss": 0.6797,
+      "step": 2146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.700121439960679,
+      "learning_rate": 1.993930863766775e-06,
+      "loss": 0.7044,
+      "step": 2147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7545887794329897,
+      "learning_rate": 1.9939202316269046e-06,
+      "loss": 0.6836,
+      "step": 2148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.492695815242632,
+      "learning_rate": 1.9939095902106892e-06,
+      "loss": 0.6829,
+      "step": 2149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.556374903893604,
+      "learning_rate": 1.9938989395182276e-06,
+      "loss": 0.6888,
+      "step": 2150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.987735533520493,
+      "learning_rate": 1.9938882795496194e-06,
+      "loss": 0.6895,
+      "step": 2151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.812515994201732,
+      "learning_rate": 1.993877610304964e-06,
+      "loss": 0.6992,
+      "step": 2152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.766116478520033,
+      "learning_rate": 1.993866931784361e-06,
+      "loss": 0.694,
+      "step": 2153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.315207898980258,
+      "learning_rate": 1.9938562439879105e-06,
+      "loss": 0.6621,
+      "step": 2154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.041173453602977,
+      "learning_rate": 1.9938455469157114e-06,
+      "loss": 0.6868,
+      "step": 2155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1258223265797733,
+      "learning_rate": 1.993834840567863e-06,
+      "loss": 0.6979,
+      "step": 2156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7905317266681554,
+      "learning_rate": 1.9938241249444673e-06,
+      "loss": 0.6764,
+      "step": 2157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.371708594732771,
+      "learning_rate": 1.993813400045623e-06,
+      "loss": 0.6934,
+      "step": 2158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.493948390560731,
+      "learning_rate": 1.99380266587143e-06,
+      "loss": 0.6986,
+      "step": 2159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.12311069258627,
+      "learning_rate": 1.993791922421989e-06,
+      "loss": 0.7012,
+      "step": 2160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9436113594107916,
+      "learning_rate": 1.9937811696974004e-06,
+      "loss": 0.6992,
+      "step": 2161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3673607061302313,
+      "learning_rate": 1.993770407697764e-06,
+      "loss": 0.7018,
+      "step": 2162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.658847689327028,
+      "learning_rate": 1.99375963642318e-06,
+      "loss": 0.6771,
+      "step": 2163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.87020084339545,
+      "learning_rate": 1.99374885587375e-06,
+      "loss": 0.6947,
+      "step": 2164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7457409902662344,
+      "learning_rate": 1.9937380660495738e-06,
+      "loss": 0.6797,
+      "step": 2165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.832011694452225,
+      "learning_rate": 1.9937272669507526e-06,
+      "loss": 0.6836,
+      "step": 2166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.820818304251833,
+      "learning_rate": 1.993716458577387e-06,
+      "loss": 0.7135,
+      "step": 2167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.414683475710164,
+      "learning_rate": 1.9937056409295778e-06,
+      "loss": 0.696,
+      "step": 2168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1640065440815346,
+      "learning_rate": 1.9936948140074256e-06,
+      "loss": 0.6921,
+      "step": 2169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.619055603631372,
+      "learning_rate": 1.993683977811032e-06,
+      "loss": 0.6882,
+      "step": 2170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.767654621168968,
+      "learning_rate": 1.993673132340498e-06,
+      "loss": 0.7012,
+      "step": 2171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.693632661079222,
+      "learning_rate": 1.9936622775959244e-06,
+      "loss": 0.6849,
+      "step": 2172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7692208279091427,
+      "learning_rate": 1.9936514135774134e-06,
+      "loss": 0.6797,
+      "step": 2173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6299987937048089,
+      "learning_rate": 1.993640540285066e-06,
+      "loss": 0.681,
+      "step": 2174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6619590465918539,
+      "learning_rate": 1.993629657718983e-06,
+      "loss": 0.6849,
+      "step": 2175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2447398219516865,
+      "learning_rate": 1.9936187658792673e-06,
+      "loss": 0.6725,
+      "step": 2176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.883013522765589,
+      "learning_rate": 1.993607864766019e-06,
+      "loss": 0.6875,
+      "step": 2177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7726782454043695,
+      "learning_rate": 1.9935969543793408e-06,
+      "loss": 0.679,
+      "step": 2178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.413285909322824,
+      "learning_rate": 1.9935860347193345e-06,
+      "loss": 0.6803,
+      "step": 2179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.252662784834964,
+      "learning_rate": 1.9935751057861017e-06,
+      "loss": 0.6999,
+      "step": 2180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5745574790869732,
+      "learning_rate": 1.9935641675797453e-06,
+      "loss": 0.6875,
+      "step": 2181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2246222537745581,
+      "learning_rate": 1.9935532201003658e-06,
+      "loss": 0.6888,
+      "step": 2182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9966419847495245,
+      "learning_rate": 1.9935422633480663e-06,
+      "loss": 0.6823,
+      "step": 2183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7194377782750362,
+      "learning_rate": 1.9935312973229495e-06,
+      "loss": 0.6999,
+      "step": 2184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.854237634755463,
+      "learning_rate": 1.993520322025117e-06,
+      "loss": 0.6862,
+      "step": 2185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6402124465243826,
+      "learning_rate": 1.9935093374546716e-06,
+      "loss": 0.6992,
+      "step": 2186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9287914395701322,
+      "learning_rate": 1.9934983436117154e-06,
+      "loss": 0.6888,
+      "step": 2187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6292821983492249,
+      "learning_rate": 1.9934873404963514e-06,
+      "loss": 0.6908,
+      "step": 2188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4242998395390254,
+      "learning_rate": 1.993476328108682e-06,
+      "loss": 0.6855,
+      "step": 2189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6929148670356424,
+      "learning_rate": 1.9934653064488103e-06,
+      "loss": 0.6947,
+      "step": 2190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.516629257379315,
+      "learning_rate": 1.9934542755168395e-06,
+      "loss": 0.6966,
+      "step": 2191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.563783104534915,
+      "learning_rate": 1.9934432353128716e-06,
+      "loss": 0.6849,
+      "step": 2192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.320336265445,
+      "learning_rate": 1.99343218583701e-06,
+      "loss": 0.6855,
+      "step": 2193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9682675806463187,
+      "learning_rate": 1.9934211270893585e-06,
+      "loss": 0.6882,
+      "step": 2194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5899030623705874,
+      "learning_rate": 1.9934100590700196e-06,
+      "loss": 0.7018,
+      "step": 2195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.891305161653392,
+      "learning_rate": 1.9933989817790965e-06,
+      "loss": 0.679,
+      "step": 2196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.742194484886591,
+      "learning_rate": 1.993387895216693e-06,
+      "loss": 0.6875,
+      "step": 2197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.56204442573362,
+      "learning_rate": 1.9933767993829123e-06,
+      "loss": 0.6745,
+      "step": 2198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.044499938657159,
+      "learning_rate": 1.9933656942778583e-06,
+      "loss": 0.6947,
+      "step": 2199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.930336924678023,
+      "learning_rate": 1.9933545799016343e-06,
+      "loss": 0.6868,
+      "step": 2200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4270942823785298,
+      "learning_rate": 1.9933434562543442e-06,
+      "loss": 0.6953,
+      "step": 2201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1157076392385075,
+      "learning_rate": 1.9933323233360917e-06,
+      "loss": 0.6999,
+      "step": 2202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0636803627438773,
+      "learning_rate": 1.9933211811469807e-06,
+      "loss": 0.6751,
+      "step": 2203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8451427765030215,
+      "learning_rate": 1.9933100296871156e-06,
+      "loss": 0.6882,
+      "step": 2204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.230688723308228,
+      "learning_rate": 1.9932988689566e-06,
+      "loss": 0.6842,
+      "step": 2205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.857625442752405,
+      "learning_rate": 1.993287698955538e-06,
+      "loss": 0.6908,
+      "step": 2206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.852961501683365,
+      "learning_rate": 1.9932765196840343e-06,
+      "loss": 0.6855,
+      "step": 2207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.025892959034675,
+      "learning_rate": 1.993265331142193e-06,
+      "loss": 0.6979,
+      "step": 2208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9099491630723764,
+      "learning_rate": 1.9932541333301184e-06,
+      "loss": 0.7025,
+      "step": 2209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6872875051754535,
+      "learning_rate": 1.993242926247915e-06,
+      "loss": 0.6895,
+      "step": 2210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6373581972127378,
+      "learning_rate": 1.9932317098956878e-06,
+      "loss": 0.6934,
+      "step": 2211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.185894222548832,
+      "learning_rate": 1.9932204842735413e-06,
+      "loss": 0.6647,
+      "step": 2212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.644617201697543,
+      "learning_rate": 1.99320924938158e-06,
+      "loss": 0.6908,
+      "step": 2213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.8786071939596924,
+      "learning_rate": 1.993198005219909e-06,
+      "loss": 0.6908,
+      "step": 2214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.100392856421475,
+      "learning_rate": 1.993186751788633e-06,
+      "loss": 0.7064,
+      "step": 2215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.675661861918666,
+      "learning_rate": 1.993175489087857e-06,
+      "loss": 0.694,
+      "step": 2216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8993535007554266,
+      "learning_rate": 1.993164217117687e-06,
+      "loss": 0.6823,
+      "step": 2217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0220853624615156,
+      "learning_rate": 1.993152935878227e-06,
+      "loss": 0.681,
+      "step": 2218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7694124129233306,
+      "learning_rate": 1.993141645369583e-06,
+      "loss": 0.6999,
+      "step": 2219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.10813239040174,
+      "learning_rate": 1.9931303455918603e-06,
+      "loss": 0.707,
+      "step": 2220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.691324897531362,
+      "learning_rate": 1.9931190365451644e-06,
+      "loss": 0.6999,
+      "step": 2221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.470810045590243,
+      "learning_rate": 1.9931077182296004e-06,
+      "loss": 0.6947,
+      "step": 2222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.5437045215206044,
+      "learning_rate": 1.993096390645274e-06,
+      "loss": 0.668,
+      "step": 2223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.7139427375484,
+      "learning_rate": 1.9930850537922917e-06,
+      "loss": 0.6882,
+      "step": 2224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3890151866918785,
+      "learning_rate": 1.9930737076707587e-06,
+      "loss": 0.6895,
+      "step": 2225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4463552671182922,
+      "learning_rate": 1.9930623522807807e-06,
+      "loss": 0.6979,
+      "step": 2226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.911240973851349,
+      "learning_rate": 1.993050987622464e-06,
+      "loss": 0.6914,
+      "step": 2227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.237078473459028,
+      "learning_rate": 1.993039613695915e-06,
+      "loss": 0.6953,
+      "step": 2228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2217792942812893,
+      "learning_rate": 1.9930282305012393e-06,
+      "loss": 0.6966,
+      "step": 2229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8912423111094836,
+      "learning_rate": 1.9930168380385427e-06,
+      "loss": 0.7018,
+      "step": 2230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9602712265165847,
+      "learning_rate": 1.9930054363079325e-06,
+      "loss": 0.6823,
+      "step": 2231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.524989064059023,
+      "learning_rate": 1.9929940253095148e-06,
+      "loss": 0.6927,
+      "step": 2232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.04593519553391,
+      "learning_rate": 1.992982605043396e-06,
+      "loss": 0.6855,
+      "step": 2233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.405621452132216,
+      "learning_rate": 1.992971175509683e-06,
+      "loss": 0.7038,
+      "step": 2234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6622529358346688,
+      "learning_rate": 1.992959736708482e-06,
+      "loss": 0.6803,
+      "step": 2235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8725074117294827,
+      "learning_rate": 1.9929482886399e-06,
+      "loss": 0.6758,
+      "step": 2236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.894361820810704,
+      "learning_rate": 1.992936831304044e-06,
+      "loss": 0.6868,
+      "step": 2237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.190575241060966,
+      "learning_rate": 1.9929253647010203e-06,
+      "loss": 0.6921,
+      "step": 2238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9752207682225762,
+      "learning_rate": 1.9929138888309368e-06,
+      "loss": 0.6888,
+      "step": 2239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.561220083584678,
+      "learning_rate": 1.9929024036939e-06,
+      "loss": 0.6875,
+      "step": 2240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6385739978115232,
+      "learning_rate": 1.992890909290017e-06,
+      "loss": 0.6829,
+      "step": 2241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7160125826680717,
+      "learning_rate": 1.9928794056193955e-06,
+      "loss": 0.6875,
+      "step": 2242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.813691600158747,
+      "learning_rate": 1.9928678926821427e-06,
+      "loss": 0.6927,
+      "step": 2243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.08489813081945,
+      "learning_rate": 1.9928563704783664e-06,
+      "loss": 0.6842,
+      "step": 2244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.674841762860167,
+      "learning_rate": 1.9928448390081734e-06,
+      "loss": 0.6829,
+      "step": 2245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7389777782958432,
+      "learning_rate": 1.992833298271672e-06,
+      "loss": 0.6901,
+      "step": 2246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1746677211388077,
+      "learning_rate": 1.9928217482689693e-06,
+      "loss": 0.6803,
+      "step": 2247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.034862729827823,
+      "learning_rate": 1.9928101890001734e-06,
+      "loss": 0.6667,
+      "step": 2248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.335795386263849,
+      "learning_rate": 1.9927986204653924e-06,
+      "loss": 0.6842,
+      "step": 2249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.459900007243406,
+      "learning_rate": 1.992787042664734e-06,
+      "loss": 0.6953,
+      "step": 2250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.690398869667616,
+      "learning_rate": 1.9927754555983066e-06,
+      "loss": 0.6862,
+      "step": 2251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5334006077595865,
+      "learning_rate": 1.992763859266218e-06,
+      "loss": 0.6875,
+      "step": 2252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8139741219076213,
+      "learning_rate": 1.992752253668576e-06,
+      "loss": 0.6901,
+      "step": 2253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6290716890214458,
+      "learning_rate": 1.99274063880549e-06,
+      "loss": 0.6947,
+      "step": 2254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.37987042440755,
+      "learning_rate": 1.9927290146770673e-06,
+      "loss": 0.6823,
+      "step": 2255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.733879051592735,
+      "learning_rate": 1.9927173812834175e-06,
+      "loss": 0.6803,
+      "step": 2256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.71290006123081,
+      "learning_rate": 1.992705738624648e-06,
+      "loss": 0.7038,
+      "step": 2257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2253052666038076,
+      "learning_rate": 1.9926940867008684e-06,
+      "loss": 0.7057,
+      "step": 2258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1594701944996562,
+      "learning_rate": 1.9926824255121874e-06,
+      "loss": 0.6947,
+      "step": 2259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8265243891545587,
+      "learning_rate": 1.9926707550587133e-06,
+      "loss": 0.709,
+      "step": 2260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.506752072694046,
+      "learning_rate": 1.992659075340555e-06,
+      "loss": 0.6882,
+      "step": 2261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.544044121728527,
+      "learning_rate": 1.992647386357822e-06,
+      "loss": 0.6947,
+      "step": 2262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.305233831697887,
+      "learning_rate": 1.992635688110623e-06,
+      "loss": 0.6803,
+      "step": 2263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.116519307157374,
+      "learning_rate": 1.9926239805990674e-06,
+      "loss": 0.6947,
+      "step": 2264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4685152728352278,
+      "learning_rate": 1.9926122638232643e-06,
+      "loss": 0.694,
+      "step": 2265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.223307139261506,
+      "learning_rate": 1.9926005377833233e-06,
+      "loss": 0.6934,
+      "step": 2266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2605626641778476,
+      "learning_rate": 1.9925888024793538e-06,
+      "loss": 0.6921,
+      "step": 2267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9502291520027989,
+      "learning_rate": 1.992577057911465e-06,
+      "loss": 0.6953,
+      "step": 2268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.32379508581916,
+      "learning_rate": 1.992565304079767e-06,
+      "loss": 0.6921,
+      "step": 2269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8785238681931542,
+      "learning_rate": 1.992553540984369e-06,
+      "loss": 0.6725,
+      "step": 2270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4347165251415621,
+      "learning_rate": 1.992541768625381e-06,
+      "loss": 0.6849,
+      "step": 2271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.143832571028357,
+      "learning_rate": 1.992529987002913e-06,
+      "loss": 0.6771,
+      "step": 2272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.927830999202969,
+      "learning_rate": 1.992518196117075e-06,
+      "loss": 0.6973,
+      "step": 2273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.671685377824438,
+      "learning_rate": 1.9925063959679765e-06,
+      "loss": 0.6901,
+      "step": 2274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0832242567717163,
+      "learning_rate": 1.992494586555728e-06,
+      "loss": 0.6882,
+      "step": 2275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.313836913515264,
+      "learning_rate": 1.99248276788044e-06,
+      "loss": 0.6966,
+      "step": 2276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0414865618147986,
+      "learning_rate": 1.992470939942223e-06,
+      "loss": 0.6803,
+      "step": 2277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3136649696114544,
+      "learning_rate": 1.9924591027411865e-06,
+      "loss": 0.6986,
+      "step": 2278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.049608567680274,
+      "learning_rate": 1.992447256277442e-06,
+      "loss": 0.6738,
+      "step": 2279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.36033366136174,
+      "learning_rate": 1.9924354005510985e-06,
+      "loss": 0.6953,
+      "step": 2280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.628671278934357,
+      "learning_rate": 1.992423535562268e-06,
+      "loss": 0.6934,
+      "step": 2281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.037281595458311,
+      "learning_rate": 1.992411661311061e-06,
+      "loss": 0.6836,
+      "step": 2282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.913575437574883,
+      "learning_rate": 1.9923997777975884e-06,
+      "loss": 0.7025,
+      "step": 2283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.998230790176866,
+      "learning_rate": 1.9923878850219604e-06,
+      "loss": 0.7005,
+      "step": 2284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0455580370656008,
+      "learning_rate": 1.992375982984289e-06,
+      "loss": 0.6803,
+      "step": 2285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.935221177736731,
+      "learning_rate": 1.9923640716846847e-06,
+      "loss": 0.6797,
+      "step": 2286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.773057780477317,
+      "learning_rate": 1.9923521511232585e-06,
+      "loss": 0.6908,
+      "step": 2287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6323952219347397,
+      "learning_rate": 1.992340221300122e-06,
+      "loss": 0.6849,
+      "step": 2288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8898065495989793,
+      "learning_rate": 1.9923282822153868e-06,
+      "loss": 0.6823,
+      "step": 2289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.747321196373675,
+      "learning_rate": 1.9923163338691636e-06,
+      "loss": 0.6849,
+      "step": 2290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.095010572132168,
+      "learning_rate": 1.9923043762615644e-06,
+      "loss": 0.6803,
+      "step": 2291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3941613700499444,
+      "learning_rate": 1.9922924093927004e-06,
+      "loss": 0.6842,
+      "step": 2292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.967353156313163,
+      "learning_rate": 1.992280433262684e-06,
+      "loss": 0.6875,
+      "step": 2293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4871467802912086,
+      "learning_rate": 1.9922684478716265e-06,
+      "loss": 0.6986,
+      "step": 2294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2362750151710877,
+      "learning_rate": 1.9922564532196397e-06,
+      "loss": 0.6641,
+      "step": 2295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7338515204920957,
+      "learning_rate": 1.992244449306836e-06,
+      "loss": 0.6823,
+      "step": 2296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.554037977384171,
+      "learning_rate": 1.992232436133327e-06,
+      "loss": 0.6966,
+      "step": 2297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.911124013074286,
+      "learning_rate": 1.992220413699225e-06,
+      "loss": 0.6999,
+      "step": 2298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.651910643519013,
+      "learning_rate": 1.992208382004641e-06,
+      "loss": 0.6934,
+      "step": 2299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.983009746382285,
+      "learning_rate": 1.99219634104969e-06,
+      "loss": 0.7051,
+      "step": 2300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9496113032164744,
+      "learning_rate": 1.9921842908344818e-06,
+      "loss": 0.6934,
+      "step": 2301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.943371883459314,
+      "learning_rate": 1.9921722313591305e-06,
+      "loss": 0.6829,
+      "step": 2302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4816611092904166,
+      "learning_rate": 1.9921601626237477e-06,
+      "loss": 0.6908,
+      "step": 2303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6176177068976274,
+      "learning_rate": 1.9921480846284462e-06,
+      "loss": 0.6888,
+      "step": 2304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1519241763573707,
+      "learning_rate": 1.992135997373339e-06,
+      "loss": 0.6888,
+      "step": 2305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.476402791546761,
+      "learning_rate": 1.9921239008585393e-06,
+      "loss": 0.6992,
+      "step": 2306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9518622672244236,
+      "learning_rate": 1.9921117950841587e-06,
+      "loss": 0.6868,
+      "step": 2307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2378999022109305,
+      "learning_rate": 1.9920996800503117e-06,
+      "loss": 0.6862,
+      "step": 2308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.452299854438484,
+      "learning_rate": 1.99208755575711e-06,
+      "loss": 0.6849,
+      "step": 2309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1798554515653494,
+      "learning_rate": 1.9920754222046677e-06,
+      "loss": 0.6849,
+      "step": 2310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6060081224609783,
+      "learning_rate": 1.9920632793930977e-06,
+      "loss": 0.681,
+      "step": 2311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.713670288982561,
+      "learning_rate": 1.9920511273225135e-06,
+      "loss": 0.6803,
+      "step": 2312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.343097783019866,
+      "learning_rate": 1.9920389659930288e-06,
+      "loss": 0.6836,
+      "step": 2313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8065378792795349,
+      "learning_rate": 1.992026795404756e-06,
+      "loss": 0.6862,
+      "step": 2314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.2856841709613045,
+      "learning_rate": 1.9920146155578096e-06,
+      "loss": 0.6934,
+      "step": 2315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0772636111970415,
+      "learning_rate": 1.992002426452303e-06,
+      "loss": 0.6829,
+      "step": 2316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.237011071970526,
+      "learning_rate": 1.99199022808835e-06,
+      "loss": 0.6868,
+      "step": 2317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8977306885338818,
+      "learning_rate": 1.9919780204660647e-06,
+      "loss": 0.6784,
+      "step": 2318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.436987403462461,
+      "learning_rate": 1.9919658035855605e-06,
+      "loss": 0.6973,
+      "step": 2319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3502802905856433,
+      "learning_rate": 1.991953577446952e-06,
+      "loss": 0.6777,
+      "step": 2320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.014034904756365,
+      "learning_rate": 1.991941342050353e-06,
+      "loss": 0.7038,
+      "step": 2321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8457710509927011,
+      "learning_rate": 1.9919290973958777e-06,
+      "loss": 0.7012,
+      "step": 2322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.636039435009982,
+      "learning_rate": 1.9919168434836403e-06,
+      "loss": 0.7018,
+      "step": 2323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.06673171141674,
+      "learning_rate": 1.9919045803137554e-06,
+      "loss": 0.7012,
+      "step": 2324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.31036030359721,
+      "learning_rate": 1.9918923078863375e-06,
+      "loss": 0.7077,
+      "step": 2325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.776510620727537,
+      "learning_rate": 1.9918800262015004e-06,
+      "loss": 0.6836,
+      "step": 2326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8904436520686976,
+      "learning_rate": 1.99186773525936e-06,
+      "loss": 0.6823,
+      "step": 2327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.934085915906719,
+      "learning_rate": 1.99185543506003e-06,
+      "loss": 0.7246,
+      "step": 2328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.714077319731466,
+      "learning_rate": 1.9918431256036255e-06,
+      "loss": 0.7077,
+      "step": 2329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6436917216103746,
+      "learning_rate": 1.9918308068902616e-06,
+      "loss": 0.696,
+      "step": 2330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.157862924556312,
+      "learning_rate": 1.9918184789200527e-06,
+      "loss": 0.6745,
+      "step": 2331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.472181095801415,
+      "learning_rate": 1.9918061416931146e-06,
+      "loss": 0.6901,
+      "step": 2332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.167223082985449,
+      "learning_rate": 1.9917937952095616e-06,
+      "loss": 0.6888,
+      "step": 2333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8069277735616647,
+      "learning_rate": 1.9917814394695098e-06,
+      "loss": 0.6816,
+      "step": 2334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.775345913348262,
+      "learning_rate": 1.991769074473074e-06,
+      "loss": 0.6908,
+      "step": 2335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5765318134631081,
+      "learning_rate": 1.99175670022037e-06,
+      "loss": 0.679,
+      "step": 2336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7691607434128964,
+      "learning_rate": 1.991744316711512e-06,
+      "loss": 0.6921,
+      "step": 2337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.179160236549375,
+      "learning_rate": 1.9917319239466175e-06,
+      "loss": 0.6888,
+      "step": 2338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9772501227003914,
+      "learning_rate": 1.991719521925801e-06,
+      "loss": 0.6784,
+      "step": 2339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6585465663997132,
+      "learning_rate": 1.9917071106491784e-06,
+      "loss": 0.6849,
+      "step": 2340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3183524043315844,
+      "learning_rate": 1.991694690116866e-06,
+      "loss": 0.6842,
+      "step": 2341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1821763842605333,
+      "learning_rate": 1.991682260328979e-06,
+      "loss": 0.6895,
+      "step": 2342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.838521368866493,
+      "learning_rate": 1.991669821285634e-06,
+      "loss": 0.6966,
+      "step": 2343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.414708481394875,
+      "learning_rate": 1.991657372986947e-06,
+      "loss": 0.707,
+      "step": 2344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.369101320292697,
+      "learning_rate": 1.991644915433034e-06,
+      "loss": 0.7025,
+      "step": 2345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6885073685842326,
+      "learning_rate": 1.991632448624011e-06,
+      "loss": 0.6908,
+      "step": 2346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.923307268010873,
+      "learning_rate": 1.9916199725599944e-06,
+      "loss": 0.6966,
+      "step": 2347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1471148860358538,
+      "learning_rate": 1.9916074872411014e-06,
+      "loss": 0.6797,
+      "step": 2348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.095058859246579,
+      "learning_rate": 1.9915949926674477e-06,
+      "loss": 0.6888,
+      "step": 2349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.581398190345814,
+      "learning_rate": 1.9915824888391506e-06,
+      "loss": 0.6986,
+      "step": 2350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0637824721023226,
+      "learning_rate": 1.991569975756326e-06,
+      "loss": 0.6868,
+      "step": 2351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9331366626695359,
+      "learning_rate": 1.991557453419092e-06,
+      "loss": 0.6849,
+      "step": 2352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.719258639469793,
+      "learning_rate": 1.9915449218275636e-06,
+      "loss": 0.6973,
+      "step": 2353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.361172541045516,
+      "learning_rate": 1.991532380981859e-06,
+      "loss": 0.6842,
+      "step": 2354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7782802076870772,
+      "learning_rate": 1.9915198308820955e-06,
+      "loss": 0.7012,
+      "step": 2355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.645311177527937,
+      "learning_rate": 1.9915072715283894e-06,
+      "loss": 0.6868,
+      "step": 2356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45311244108103604,
+      "learning_rate": 1.991494702920858e-06,
+      "loss": 0.6849,
+      "step": 2357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.3955219787872766,
+      "learning_rate": 1.991482125059619e-06,
+      "loss": 0.6829,
+      "step": 2358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2463520298662694,
+      "learning_rate": 1.9914695379447897e-06,
+      "loss": 0.6862,
+      "step": 2359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7220375715786762,
+      "learning_rate": 1.9914569415764875e-06,
+      "loss": 0.6908,
+      "step": 2360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6371019877846984,
+      "learning_rate": 1.9914443359548303e-06,
+      "loss": 0.6862,
+      "step": 2361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.72273289321425,
+      "learning_rate": 1.9914317210799352e-06,
+      "loss": 0.6986,
+      "step": 2362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2395283359596143,
+      "learning_rate": 1.99141909695192e-06,
+      "loss": 0.6862,
+      "step": 2363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6040482542572079,
+      "learning_rate": 1.991406463570903e-06,
+      "loss": 0.6712,
+      "step": 2364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.72296237359763,
+      "learning_rate": 1.9913938209370015e-06,
+      "loss": 0.6921,
+      "step": 2365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.762056507314355,
+      "learning_rate": 1.991381169050334e-06,
+      "loss": 0.6784,
+      "step": 2366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7271170560212206,
+      "learning_rate": 1.9913685079110186e-06,
+      "loss": 0.6725,
+      "step": 2367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.074174316011279,
+      "learning_rate": 1.9913558375191726e-06,
+      "loss": 0.6855,
+      "step": 2368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5949354362220207,
+      "learning_rate": 1.9913431578749152e-06,
+      "loss": 0.6934,
+      "step": 2369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.418178023189968,
+      "learning_rate": 1.9913304689783644e-06,
+      "loss": 0.6849,
+      "step": 2370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.35548319673339,
+      "learning_rate": 1.9913177708296393e-06,
+      "loss": 0.6882,
+      "step": 2371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9625451158890335,
+      "learning_rate": 1.9913050634288567e-06,
+      "loss": 0.6758,
+      "step": 2372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.964763012693123,
+      "learning_rate": 1.991292346776137e-06,
+      "loss": 0.7057,
+      "step": 2373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.749692294516227,
+      "learning_rate": 1.9912796208715983e-06,
+      "loss": 0.6979,
+      "step": 2374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.8986423088413105,
+      "learning_rate": 1.9912668857153587e-06,
+      "loss": 0.6895,
+      "step": 2375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.423591533609114,
+      "learning_rate": 1.9912541413075376e-06,
+      "loss": 0.7214,
+      "step": 2376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8013112197700947,
+      "learning_rate": 1.991241387648254e-06,
+      "loss": 0.6947,
+      "step": 2377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5569539405450574,
+      "learning_rate": 1.9912286247376274e-06,
+      "loss": 0.6797,
+      "step": 2378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2111193150773094,
+      "learning_rate": 1.991215852575776e-06,
+      "loss": 0.6986,
+      "step": 2379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.269252816751187,
+      "learning_rate": 1.991203071162819e-06,
+      "loss": 0.6699,
+      "step": 2380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.913521127095258,
+      "learning_rate": 1.9911902804988766e-06,
+      "loss": 0.6888,
+      "step": 2381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.570259280884476,
+      "learning_rate": 1.9911774805840675e-06,
+      "loss": 0.7083,
+      "step": 2382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3857210898458456,
+      "learning_rate": 1.9911646714185117e-06,
+      "loss": 0.681,
+      "step": 2383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.406032786867253,
+      "learning_rate": 1.991151853002328e-06,
+      "loss": 0.6725,
+      "step": 2384
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.977565790351673,
+      "learning_rate": 1.9911390253356363e-06,
+      "loss": 0.6927,
+      "step": 2385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0183733445131988,
+      "learning_rate": 1.9911261884185563e-06,
+      "loss": 0.6901,
+      "step": 2386
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.083822810631499,
+      "learning_rate": 1.991113342251208e-06,
+      "loss": 0.6562,
+      "step": 2387
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.28944118573875,
+      "learning_rate": 1.991100486833712e-06,
+      "loss": 0.6992,
+      "step": 2388
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.813179021549531,
+      "learning_rate": 1.9910876221661867e-06,
+      "loss": 0.694,
+      "step": 2389
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.183013346472238,
+      "learning_rate": 1.991074748248753e-06,
+      "loss": 0.7012,
+      "step": 2390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6052952500584277,
+      "learning_rate": 1.9910618650815315e-06,
+      "loss": 0.6862,
+      "step": 2391
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.736633709905756,
+      "learning_rate": 1.991048972664641e-06,
+      "loss": 0.7012,
+      "step": 2392
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9882256882324154,
+      "learning_rate": 1.991036070998204e-06,
+      "loss": 0.6849,
+      "step": 2393
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8717732826680542,
+      "learning_rate": 1.991023160082339e-06,
+      "loss": 0.6999,
+      "step": 2394
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.814056045632999,
+      "learning_rate": 1.9910102399171673e-06,
+      "loss": 0.707,
+      "step": 2395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1519946595215487,
+      "learning_rate": 1.9909973105028096e-06,
+      "loss": 0.6751,
+      "step": 2396
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.516383975920999,
+      "learning_rate": 1.990984371839386e-06,
+      "loss": 0.6953,
+      "step": 2397
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.305978381448393,
+      "learning_rate": 1.9909714239270177e-06,
+      "loss": 0.694,
+      "step": 2398
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.427928167504524,
+      "learning_rate": 1.9909584667658258e-06,
+      "loss": 0.6849,
+      "step": 2399
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.705015343842275,
+      "learning_rate": 1.9909455003559306e-06,
+      "loss": 0.6927,
+      "step": 2400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3550374341969866,
+      "learning_rate": 1.990932524697454e-06,
+      "loss": 0.6973,
+      "step": 2401
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6529971859164365,
+      "learning_rate": 1.9909195397905156e-06,
+      "loss": 0.6966,
+      "step": 2402
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.172481873794828,
+      "learning_rate": 1.990906545635238e-06,
+      "loss": 0.6836,
+      "step": 2403
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.196402323169406,
+      "learning_rate": 1.990893542231742e-06,
+      "loss": 0.6784,
+      "step": 2404
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.541393542638346,
+      "learning_rate": 1.990880529580149e-06,
+      "loss": 0.694,
+      "step": 2405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2323274672313453,
+      "learning_rate": 1.9908675076805802e-06,
+      "loss": 0.6855,
+      "step": 2406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.119114121513338,
+      "learning_rate": 1.9908544765331576e-06,
+      "loss": 0.6927,
+      "step": 2407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0068103032453524,
+      "learning_rate": 1.990841436138002e-06,
+      "loss": 0.6966,
+      "step": 2408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.124206148595375,
+      "learning_rate": 1.990828386495236e-06,
+      "loss": 0.6855,
+      "step": 2409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.0350839293117575,
+      "learning_rate": 1.990815327604981e-06,
+      "loss": 0.6895,
+      "step": 2410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.775156966703518,
+      "learning_rate": 1.9908022594673592e-06,
+      "loss": 0.6953,
+      "step": 2411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.630559836103916,
+      "learning_rate": 1.9907891820824917e-06,
+      "loss": 0.6868,
+      "step": 2412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.51708831587758,
+      "learning_rate": 1.9907760954505015e-06,
+      "loss": 0.6888,
+      "step": 2413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.638509719456267,
+      "learning_rate": 1.9907629995715103e-06,
+      "loss": 0.7077,
+      "step": 2414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0546937761544428,
+      "learning_rate": 1.9907498944456403e-06,
+      "loss": 0.6849,
+      "step": 2415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.852517228226061,
+      "learning_rate": 1.990736780073014e-06,
+      "loss": 0.6901,
+      "step": 2416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.350049284103575,
+      "learning_rate": 1.990723656453754e-06,
+      "loss": 0.6953,
+      "step": 2417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7407917077462464,
+      "learning_rate": 1.9907105235879823e-06,
+      "loss": 0.6842,
+      "step": 2418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4568428471564547,
+      "learning_rate": 1.9906973814758216e-06,
+      "loss": 0.6732,
+      "step": 2419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0980099492397213,
+      "learning_rate": 1.9906842301173946e-06,
+      "loss": 0.6862,
+      "step": 2420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7228362506623108,
+      "learning_rate": 1.990671069512824e-06,
+      "loss": 0.6868,
+      "step": 2421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.060907661308508,
+      "learning_rate": 1.9906578996622333e-06,
+      "loss": 0.6986,
+      "step": 2422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4026712230872165,
+      "learning_rate": 1.990644720565744e-06,
+      "loss": 0.6934,
+      "step": 2423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.030935426755939,
+      "learning_rate": 1.9906315322234804e-06,
+      "loss": 0.7135,
+      "step": 2424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6663004120809151,
+      "learning_rate": 1.990618334635565e-06,
+      "loss": 0.6888,
+      "step": 2425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.051692268575131,
+      "learning_rate": 1.990605127802121e-06,
+      "loss": 0.6882,
+      "step": 2426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6639793592614784,
+      "learning_rate": 1.9905919117232717e-06,
+      "loss": 0.6914,
+      "step": 2427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8057250878916925,
+      "learning_rate": 1.9905786863991406e-06,
+      "loss": 0.6895,
+      "step": 2428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.529140140163367,
+      "learning_rate": 1.990565451829851e-06,
+      "loss": 0.6901,
+      "step": 2429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.766092312271186,
+      "learning_rate": 1.9905522080155264e-06,
+      "loss": 0.6986,
+      "step": 2430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.881884538546903,
+      "learning_rate": 1.9905389549562904e-06,
+      "loss": 0.6823,
+      "step": 2431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9299842480295004,
+      "learning_rate": 1.990525692652267e-06,
+      "loss": 0.6667,
+      "step": 2432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.8484660996013025,
+      "learning_rate": 1.9905124211035793e-06,
+      "loss": 0.7064,
+      "step": 2433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.335072796553507,
+      "learning_rate": 1.990499140310352e-06,
+      "loss": 0.6855,
+      "step": 2434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.567434184931717,
+      "learning_rate": 1.990485850272708e-06,
+      "loss": 0.6823,
+      "step": 2435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7489772719650274,
+      "learning_rate": 1.9904725509907727e-06,
+      "loss": 0.679,
+      "step": 2436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.414622101707826,
+      "learning_rate": 1.990459242464669e-06,
+      "loss": 0.681,
+      "step": 2437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.941909127006602,
+      "learning_rate": 1.990445924694522e-06,
+      "loss": 0.7005,
+      "step": 2438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.634478378363209,
+      "learning_rate": 1.9904325976804555e-06,
+      "loss": 0.6849,
+      "step": 2439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.100449453001142,
+      "learning_rate": 1.9904192614225935e-06,
+      "loss": 0.6784,
+      "step": 2440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8225501584510908,
+      "learning_rate": 1.9904059159210615e-06,
+      "loss": 0.7012,
+      "step": 2441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.123599575650918,
+      "learning_rate": 1.9903925611759834e-06,
+      "loss": 0.7168,
+      "step": 2442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9232016891829709,
+      "learning_rate": 1.9903791971874834e-06,
+      "loss": 0.6771,
+      "step": 2443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.776691509353578,
+      "learning_rate": 1.990365823955687e-06,
+      "loss": 0.6706,
+      "step": 2444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.426914068420084,
+      "learning_rate": 1.9903524414807198e-06,
+      "loss": 0.6816,
+      "step": 2445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.071395166967945,
+      "learning_rate": 1.990339049762705e-06,
+      "loss": 0.6979,
+      "step": 2446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3358753345925634,
+      "learning_rate": 1.9903256488017676e-06,
+      "loss": 0.7109,
+      "step": 2447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.370873037547973,
+      "learning_rate": 1.990312238598034e-06,
+      "loss": 0.6927,
+      "step": 2448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.9713920644756975,
+      "learning_rate": 1.9902988191516285e-06,
+      "loss": 0.7005,
+      "step": 2449
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9564420462763736,
+      "learning_rate": 1.9902853904626765e-06,
+      "loss": 0.7057,
+      "step": 2450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.7750909577266,
+      "learning_rate": 1.9902719525313032e-06,
+      "loss": 0.7051,
+      "step": 2451
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7040609256600752,
+      "learning_rate": 1.9902585053576346e-06,
+      "loss": 0.6908,
+      "step": 2452
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45358365792047206,
+      "learning_rate": 1.990245048941796e-06,
+      "loss": 0.6803,
+      "step": 2453
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.639056699397374,
+      "learning_rate": 1.9902315832839122e-06,
+      "loss": 0.709,
+      "step": 2454
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3184729448969468,
+      "learning_rate": 1.99021810838411e-06,
+      "loss": 0.6953,
+      "step": 2455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8977666774821498,
+      "learning_rate": 1.9902046242425143e-06,
+      "loss": 0.6849,
+      "step": 2456
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8435245984188402,
+      "learning_rate": 1.9901911308592512e-06,
+      "loss": 0.6868,
+      "step": 2457
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5304279024331808,
+      "learning_rate": 1.9901776282344466e-06,
+      "loss": 0.6947,
+      "step": 2458
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.155596731123696,
+      "learning_rate": 1.990164116368227e-06,
+      "loss": 0.6934,
+      "step": 2459
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2006659536436732,
+      "learning_rate": 1.990150595260718e-06,
+      "loss": 0.6803,
+      "step": 2460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8340066213619642,
+      "learning_rate": 1.990137064912046e-06,
+      "loss": 0.6875,
+      "step": 2461
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.7478393076042735,
+      "learning_rate": 1.990123525322337e-06,
+      "loss": 0.6921,
+      "step": 2462
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.8443809911615485,
+      "learning_rate": 1.990109976491718e-06,
+      "loss": 0.6986,
+      "step": 2463
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9755522452975106,
+      "learning_rate": 1.9900964184203144e-06,
+      "loss": 0.6868,
+      "step": 2464
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1869806016008937,
+      "learning_rate": 1.9900828511082537e-06,
+      "loss": 0.6875,
+      "step": 2465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.779783717613751,
+      "learning_rate": 1.990069274555662e-06,
+      "loss": 0.6986,
+      "step": 2466
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6294631644937062,
+      "learning_rate": 1.9900556887626668e-06,
+      "loss": 0.6849,
+      "step": 2467
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.91282053933091,
+      "learning_rate": 1.9900420937293937e-06,
+      "loss": 0.6927,
+      "step": 2468
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.605068472637087,
+      "learning_rate": 1.9900284894559704e-06,
+      "loss": 0.6992,
+      "step": 2469
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1219484708551961,
+      "learning_rate": 1.990014875942524e-06,
+      "loss": 0.6764,
+      "step": 2470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.508883945726252,
+      "learning_rate": 1.9900012531891804e-06,
+      "loss": 0.6855,
+      "step": 2471
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.550505027566599,
+      "learning_rate": 1.989987621196068e-06,
+      "loss": 0.6921,
+      "step": 2472
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1806279030931466,
+      "learning_rate": 1.989973979963314e-06,
+      "loss": 0.6849,
+      "step": 2473
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.614877264529312,
+      "learning_rate": 1.9899603294910445e-06,
+      "loss": 0.6855,
+      "step": 2474
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.343655910128949,
+      "learning_rate": 1.989946669779388e-06,
+      "loss": 0.6953,
+      "step": 2475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.544477823431974,
+      "learning_rate": 1.989933000828472e-06,
+      "loss": 0.696,
+      "step": 2476
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1820093594172985,
+      "learning_rate": 1.9899193226384233e-06,
+      "loss": 0.6927,
+      "step": 2477
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.46641626006166,
+      "learning_rate": 1.9899056352093705e-06,
+      "loss": 0.7044,
+      "step": 2478
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.990002386271633,
+      "learning_rate": 1.9898919385414404e-06,
+      "loss": 0.7038,
+      "step": 2479
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.5314104933583175,
+      "learning_rate": 1.9898782326347614e-06,
+      "loss": 0.6927,
+      "step": 2480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8827268593009416,
+      "learning_rate": 1.9898645174894615e-06,
+      "loss": 0.694,
+      "step": 2481
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1256577840738355,
+      "learning_rate": 1.9898507931056685e-06,
+      "loss": 0.679,
+      "step": 2482
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.314393813642246,
+      "learning_rate": 1.9898370594835104e-06,
+      "loss": 0.6979,
+      "step": 2483
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3923716092039267,
+      "learning_rate": 1.9898233166231154e-06,
+      "loss": 0.6927,
+      "step": 2484
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.870837542310202,
+      "learning_rate": 1.989809564524612e-06,
+      "loss": 0.6823,
+      "step": 2485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.363497887558907,
+      "learning_rate": 1.9897958031881283e-06,
+      "loss": 0.6966,
+      "step": 2486
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.201004976679577,
+      "learning_rate": 1.989782032613793e-06,
+      "loss": 0.6895,
+      "step": 2487
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.106759051132948,
+      "learning_rate": 1.9897682528017345e-06,
+      "loss": 0.6895,
+      "step": 2488
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.552946773501974,
+      "learning_rate": 1.989754463752081e-06,
+      "loss": 0.6823,
+      "step": 2489
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.790618569149767,
+      "learning_rate": 1.989740665464962e-06,
+      "loss": 0.6797,
+      "step": 2490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4184648571678795,
+      "learning_rate": 1.9897268579405057e-06,
+      "loss": 0.694,
+      "step": 2491
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3241428204397026,
+      "learning_rate": 1.989713041178841e-06,
+      "loss": 0.6992,
+      "step": 2492
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4204896229942934,
+      "learning_rate": 1.989699215180097e-06,
+      "loss": 0.6803,
+      "step": 2493
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0625499754084213,
+      "learning_rate": 1.9896853799444026e-06,
+      "loss": 0.6738,
+      "step": 2494
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7694960925696954,
+      "learning_rate": 1.989671535471887e-06,
+      "loss": 0.6745,
+      "step": 2495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6032849241978036,
+      "learning_rate": 1.98965768176268e-06,
+      "loss": 0.6947,
+      "step": 2496
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.02787621920832,
+      "learning_rate": 1.98964381881691e-06,
+      "loss": 0.6986,
+      "step": 2497
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.319470313016433,
+      "learning_rate": 1.9896299466347063e-06,
+      "loss": 0.694,
+      "step": 2498
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.087994119514148,
+      "learning_rate": 1.989616065216199e-06,
+      "loss": 0.7305,
+      "step": 2499
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.390122342368475,
+      "learning_rate": 1.9896021745615176e-06,
+      "loss": 0.7109,
+      "step": 2500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.870564587983292,
+      "learning_rate": 1.9895882746707917e-06,
+      "loss": 0.7031,
+      "step": 2501
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.30563019271204,
+      "learning_rate": 1.9895743655441505e-06,
+      "loss": 0.6855,
+      "step": 2502
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0206765481800906,
+      "learning_rate": 1.9895604471817247e-06,
+      "loss": 0.6882,
+      "step": 2503
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.524134590533019,
+      "learning_rate": 1.9895465195836434e-06,
+      "loss": 0.6882,
+      "step": 2504
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.484072501908744,
+      "learning_rate": 1.989532582750037e-06,
+      "loss": 0.6934,
+      "step": 2505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.405899945420772,
+      "learning_rate": 1.9895186366810357e-06,
+      "loss": 0.6921,
+      "step": 2506
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.988694858722698,
+      "learning_rate": 1.9895046813767693e-06,
+      "loss": 0.6888,
+      "step": 2507
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.405451511544117,
+      "learning_rate": 1.989490716837368e-06,
+      "loss": 0.6986,
+      "step": 2508
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.13300120815804,
+      "learning_rate": 1.9894767430629627e-06,
+      "loss": 0.7018,
+      "step": 2509
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.278927858476665,
+      "learning_rate": 1.9894627600536834e-06,
+      "loss": 0.7077,
+      "step": 2510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2163170224776274,
+      "learning_rate": 1.9894487678096605e-06,
+      "loss": 0.6914,
+      "step": 2511
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.9633531091308205,
+      "learning_rate": 1.989434766331025e-06,
+      "loss": 0.6934,
+      "step": 2512
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.23942815388428,
+      "learning_rate": 1.9894207556179067e-06,
+      "loss": 0.7025,
+      "step": 2513
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1559010983803857,
+      "learning_rate": 1.9894067356704375e-06,
+      "loss": 0.6901,
+      "step": 2514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.309370961094347,
+      "learning_rate": 1.9893927064887477e-06,
+      "loss": 0.6862,
+      "step": 2515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2612793501344064,
+      "learning_rate": 1.989378668072968e-06,
+      "loss": 0.6901,
+      "step": 2516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7499554328727192,
+      "learning_rate": 1.98936462042323e-06,
+      "loss": 0.6979,
+      "step": 2517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.068414677793844,
+      "learning_rate": 1.9893505635396645e-06,
+      "loss": 0.6973,
+      "step": 2518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.301527112922328,
+      "learning_rate": 1.9893364974224025e-06,
+      "loss": 0.6973,
+      "step": 2519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0749205207714458,
+      "learning_rate": 1.989322422071576e-06,
+      "loss": 0.6882,
+      "step": 2520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8923598243039756,
+      "learning_rate": 1.9893083374873152e-06,
+      "loss": 0.6758,
+      "step": 2521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.200197691260833,
+      "learning_rate": 1.9892942436697525e-06,
+      "loss": 0.6986,
+      "step": 2522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.191720537369071,
+      "learning_rate": 1.989280140619019e-06,
+      "loss": 0.7025,
+      "step": 2523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.105171613913576,
+      "learning_rate": 1.9892660283352465e-06,
+      "loss": 0.6973,
+      "step": 2524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.501943415268559,
+      "learning_rate": 1.9892519068185667e-06,
+      "loss": 0.6986,
+      "step": 2525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.118973312549695,
+      "learning_rate": 1.9892377760691113e-06,
+      "loss": 0.6999,
+      "step": 2526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0712739832200344,
+      "learning_rate": 1.9892236360870124e-06,
+      "loss": 0.6868,
+      "step": 2527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.326682877995164,
+      "learning_rate": 1.989209486872402e-06,
+      "loss": 0.6855,
+      "step": 2528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.184204564768249,
+      "learning_rate": 1.9891953284254116e-06,
+      "loss": 0.6914,
+      "step": 2529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.228720731038672,
+      "learning_rate": 1.9891811607461736e-06,
+      "loss": 0.696,
+      "step": 2530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0880970320917565,
+      "learning_rate": 1.9891669838348206e-06,
+      "loss": 0.6914,
+      "step": 2531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.658612568081262,
+      "learning_rate": 1.9891527976914844e-06,
+      "loss": 0.6973,
+      "step": 2532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4229572433802278,
+      "learning_rate": 1.989138602316298e-06,
+      "loss": 0.6882,
+      "step": 2533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.746283910362651,
+      "learning_rate": 1.9891243977093935e-06,
+      "loss": 0.7051,
+      "step": 2534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.128162435232586,
+      "learning_rate": 1.9891101838709035e-06,
+      "loss": 0.696,
+      "step": 2535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.808877845251727,
+      "learning_rate": 1.9890959608009606e-06,
+      "loss": 0.6816,
+      "step": 2536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3961652390914865,
+      "learning_rate": 1.9890817284996977e-06,
+      "loss": 0.6901,
+      "step": 2537
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.044491323522878,
+      "learning_rate": 1.989067486967248e-06,
+      "loss": 0.6914,
+      "step": 2538
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5782453501324029,
+      "learning_rate": 1.9890532362037435e-06,
+      "loss": 0.6868,
+      "step": 2539
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.570686520973919,
+      "learning_rate": 1.989038976209318e-06,
+      "loss": 0.6862,
+      "step": 2540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.339204114613456,
+      "learning_rate": 1.989024706984103e-06,
+      "loss": 0.681,
+      "step": 2541
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9403896968106258,
+      "learning_rate": 1.9890104285282344e-06,
+      "loss": 0.6979,
+      "step": 2542
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.224262089170245,
+      "learning_rate": 1.9889961408418435e-06,
+      "loss": 0.679,
+      "step": 2543
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.966427092640978,
+      "learning_rate": 1.9889818439250644e-06,
+      "loss": 0.6888,
+      "step": 2544
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.577174362911111,
+      "learning_rate": 1.9889675377780296e-06,
+      "loss": 0.6986,
+      "step": 2545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.396374091397961,
+      "learning_rate": 1.988953222400874e-06,
+      "loss": 0.6771,
+      "step": 2546
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3447197087197975,
+      "learning_rate": 1.98893889779373e-06,
+      "loss": 0.6816,
+      "step": 2547
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.160639552481637,
+      "learning_rate": 1.988924563956732e-06,
+      "loss": 0.6895,
+      "step": 2548
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.08270093829534,
+      "learning_rate": 1.9889102208900143e-06,
+      "loss": 0.6797,
+      "step": 2549
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9904869476129323,
+      "learning_rate": 1.988895868593709e-06,
+      "loss": 0.6973,
+      "step": 2550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.360214720341294,
+      "learning_rate": 1.9888815070679516e-06,
+      "loss": 0.7051,
+      "step": 2551
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.003682656156149,
+      "learning_rate": 1.988867136312876e-06,
+      "loss": 0.6862,
+      "step": 2552
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3609459694407446,
+      "learning_rate": 1.988852756328615e-06,
+      "loss": 0.6855,
+      "step": 2553
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5091255388247373,
+      "learning_rate": 1.9888383671153048e-06,
+      "loss": 0.6842,
+      "step": 2554
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5271510329536506,
+      "learning_rate": 1.9888239686730783e-06,
+      "loss": 0.7044,
+      "step": 2555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6557364583367375,
+      "learning_rate": 1.98880956100207e-06,
+      "loss": 0.6849,
+      "step": 2556
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.474382556949493,
+      "learning_rate": 1.988795144102415e-06,
+      "loss": 0.6947,
+      "step": 2557
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.861421726870536,
+      "learning_rate": 1.988780717974247e-06,
+      "loss": 0.6966,
+      "step": 2558
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.383612523508205,
+      "learning_rate": 1.9887662826177015e-06,
+      "loss": 0.6836,
+      "step": 2559
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.297961513192983,
+      "learning_rate": 1.9887518380329126e-06,
+      "loss": 0.6966,
+      "step": 2560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.553644615588608,
+      "learning_rate": 1.9887373842200153e-06,
+      "loss": 0.6921,
+      "step": 2561
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.343507402624377,
+      "learning_rate": 1.988722921179145e-06,
+      "loss": 0.6992,
+      "step": 2562
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2911529929713519,
+      "learning_rate": 1.9887084489104358e-06,
+      "loss": 0.6836,
+      "step": 2563
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3729143338718164,
+      "learning_rate": 1.9886939674140237e-06,
+      "loss": 0.6725,
+      "step": 2564
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.698378273947798,
+      "learning_rate": 1.988679476690043e-06,
+      "loss": 0.7116,
+      "step": 2565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.355851502930417,
+      "learning_rate": 1.988664976738629e-06,
+      "loss": 0.7285,
+      "step": 2566
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.887997694497216,
+      "learning_rate": 1.9886504675599177e-06,
+      "loss": 0.6927,
+      "step": 2567
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.362733555006235,
+      "learning_rate": 1.988635949154044e-06,
+      "loss": 0.6921,
+      "step": 2568
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.567894301704592,
+      "learning_rate": 1.988621421521144e-06,
+      "loss": 0.6823,
+      "step": 2569
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1918692219291867,
+      "learning_rate": 1.988606884661352e-06,
+      "loss": 0.6745,
+      "step": 2570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1180804755172373,
+      "learning_rate": 1.9885923385748057e-06,
+      "loss": 0.696,
+      "step": 2571
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7959646298185701,
+      "learning_rate": 1.988577783261639e-06,
+      "loss": 0.6882,
+      "step": 2572
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.972592171104218,
+      "learning_rate": 1.988563218721988e-06,
+      "loss": 0.694,
+      "step": 2573
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.367959554050634,
+      "learning_rate": 1.9885486449559897e-06,
+      "loss": 0.6823,
+      "step": 2574
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6260904374768683,
+      "learning_rate": 1.9885340619637794e-06,
+      "loss": 0.6771,
+      "step": 2575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8123565044747347,
+      "learning_rate": 1.988519469745493e-06,
+      "loss": 0.7077,
+      "step": 2576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.201085954711838,
+      "learning_rate": 1.988504868301267e-06,
+      "loss": 0.6927,
+      "step": 2577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.833124187569013,
+      "learning_rate": 1.988490257631238e-06,
+      "loss": 0.6862,
+      "step": 2578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.203629904482312,
+      "learning_rate": 1.988475637735542e-06,
+      "loss": 0.6777,
+      "step": 2579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.377624226742618,
+      "learning_rate": 1.988461008614315e-06,
+      "loss": 0.6921,
+      "step": 2580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.419926700781408,
+      "learning_rate": 1.9884463702676943e-06,
+      "loss": 0.6986,
+      "step": 2581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.923988936200002,
+      "learning_rate": 1.9884317226958164e-06,
+      "loss": 0.694,
+      "step": 2582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.134002478786731,
+      "learning_rate": 1.9884170658988174e-06,
+      "loss": 0.6882,
+      "step": 2583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8001396234115418,
+      "learning_rate": 1.9884023998768348e-06,
+      "loss": 0.6934,
+      "step": 2584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9730952581815666,
+      "learning_rate": 1.9883877246300047e-06,
+      "loss": 0.6745,
+      "step": 2585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6431368729716742,
+      "learning_rate": 1.988373040158465e-06,
+      "loss": 0.6934,
+      "step": 2586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9782081261658017,
+      "learning_rate": 1.9883583464623523e-06,
+      "loss": 0.6947,
+      "step": 2587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.505876216294109,
+      "learning_rate": 1.988343643541804e-06,
+      "loss": 0.6868,
+      "step": 2588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.175055979481037,
+      "learning_rate": 1.9883289313969566e-06,
+      "loss": 0.6758,
+      "step": 2589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8637605272501063,
+      "learning_rate": 1.988314210027948e-06,
+      "loss": 0.6953,
+      "step": 2590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6215259331134593,
+      "learning_rate": 1.9882994794349154e-06,
+      "loss": 0.6771,
+      "step": 2591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.100364727793992,
+      "learning_rate": 1.988284739617997e-06,
+      "loss": 0.7044,
+      "step": 2592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.279948491369992,
+      "learning_rate": 1.9882699905773286e-06,
+      "loss": 0.6888,
+      "step": 2593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3428433801793462,
+      "learning_rate": 1.9882552323130497e-06,
+      "loss": 0.6862,
+      "step": 2594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.3523923641673905,
+      "learning_rate": 1.9882404648252974e-06,
+      "loss": 0.6829,
+      "step": 2595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.423684344977644,
+      "learning_rate": 1.988225688114209e-06,
+      "loss": 0.7044,
+      "step": 2596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.843085852199613,
+      "learning_rate": 1.988210902179923e-06,
+      "loss": 0.6979,
+      "step": 2597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.103747858253418,
+      "learning_rate": 1.9881961070225775e-06,
+      "loss": 0.6986,
+      "step": 2598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.929580848768335,
+      "learning_rate": 1.98818130264231e-06,
+      "loss": 0.707,
+      "step": 2599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9383082321797218,
+      "learning_rate": 1.9881664890392597e-06,
+      "loss": 0.6895,
+      "step": 2600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7627893381612147,
+      "learning_rate": 1.9881516662135634e-06,
+      "loss": 0.679,
+      "step": 2601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5407811695520812,
+      "learning_rate": 1.988136834165361e-06,
+      "loss": 0.6927,
+      "step": 2602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.058421154029782,
+      "learning_rate": 1.9881219928947893e-06,
+      "loss": 0.6784,
+      "step": 2603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.655288368205597,
+      "learning_rate": 1.988107142401988e-06,
+      "loss": 0.6595,
+      "step": 2604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.142414745754062,
+      "learning_rate": 1.9880922826870956e-06,
+      "loss": 0.7116,
+      "step": 2605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.5233586964609245,
+      "learning_rate": 1.98807741375025e-06,
+      "loss": 0.6921,
+      "step": 2606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5555740652791237,
+      "learning_rate": 1.988062535591591e-06,
+      "loss": 0.6875,
+      "step": 2607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.918239358004728,
+      "learning_rate": 1.9880476482112563e-06,
+      "loss": 0.7077,
+      "step": 2608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.690519138243026,
+      "learning_rate": 1.988032751609386e-06,
+      "loss": 0.7129,
+      "step": 2609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3445249332257476,
+      "learning_rate": 1.988017845786119e-06,
+      "loss": 0.6862,
+      "step": 2610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.71887353710108,
+      "learning_rate": 1.9880029307415935e-06,
+      "loss": 0.6921,
+      "step": 2611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.712302915036871,
+      "learning_rate": 1.9879880064759497e-06,
+      "loss": 0.6862,
+      "step": 2612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.267992196909362,
+      "learning_rate": 1.9879730729893264e-06,
+      "loss": 0.6914,
+      "step": 2613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2916644086429314,
+      "learning_rate": 1.9879581302818627e-06,
+      "loss": 0.6966,
+      "step": 2614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.838531717493355,
+      "learning_rate": 1.9879431783536983e-06,
+      "loss": 0.7005,
+      "step": 2615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.894343176095047,
+      "learning_rate": 1.9879282172049733e-06,
+      "loss": 0.6849,
+      "step": 2616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8670915889466251,
+      "learning_rate": 1.9879132468358267e-06,
+      "loss": 0.6882,
+      "step": 2617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0952748508773924,
+      "learning_rate": 1.9878982672463987e-06,
+      "loss": 0.7025,
+      "step": 2618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.07685596855473,
+      "learning_rate": 1.987883278436828e-06,
+      "loss": 0.6947,
+      "step": 2619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.756715606257146,
+      "learning_rate": 1.987868280407256e-06,
+      "loss": 0.7018,
+      "step": 2620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.085994334046859,
+      "learning_rate": 1.987853273157822e-06,
+      "loss": 0.6921,
+      "step": 2621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.4663661541679724,
+      "learning_rate": 1.987838256688666e-06,
+      "loss": 0.6823,
+      "step": 2622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.248598407849384,
+      "learning_rate": 1.987823230999928e-06,
+      "loss": 0.6855,
+      "step": 2623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5880769367801058,
+      "learning_rate": 1.9878081960917484e-06,
+      "loss": 0.6849,
+      "step": 2624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6667717391086113,
+      "learning_rate": 1.9877931519642675e-06,
+      "loss": 0.6921,
+      "step": 2625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.477423518412438,
+      "learning_rate": 1.987778098617626e-06,
+      "loss": 0.6868,
+      "step": 2626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0616554454686193,
+      "learning_rate": 1.9877630360519638e-06,
+      "loss": 0.6862,
+      "step": 2627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9676764262665398,
+      "learning_rate": 1.9877479642674222e-06,
+      "loss": 0.6862,
+      "step": 2628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9600444396355154,
+      "learning_rate": 1.9877328832641413e-06,
+      "loss": 0.6986,
+      "step": 2629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5349827435882775,
+      "learning_rate": 1.987717793042262e-06,
+      "loss": 0.6986,
+      "step": 2630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.634228599136488,
+      "learning_rate": 1.9877026936019253e-06,
+      "loss": 0.6764,
+      "step": 2631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5835123089100809,
+      "learning_rate": 1.9876875849432723e-06,
+      "loss": 0.6758,
+      "step": 2632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6192342819281551,
+      "learning_rate": 1.987672467066443e-06,
+      "loss": 0.6992,
+      "step": 2633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5564616991128437,
+      "learning_rate": 1.98765733997158e-06,
+      "loss": 0.6745,
+      "step": 2634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.77237579643837,
+      "learning_rate": 1.987642203658823e-06,
+      "loss": 0.6934,
+      "step": 2635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.998979745733603,
+      "learning_rate": 1.9876270581283144e-06,
+      "loss": 0.6797,
+      "step": 2636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4863051061961812,
+      "learning_rate": 1.9876119033801953e-06,
+      "loss": 0.6921,
+      "step": 2637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.995896910907803,
+      "learning_rate": 1.9875967394146065e-06,
+      "loss": 0.6966,
+      "step": 2638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6201819163698368,
+      "learning_rate": 1.9875815662316903e-06,
+      "loss": 0.6973,
+      "step": 2639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.7264928557656285,
+      "learning_rate": 1.987566383831588e-06,
+      "loss": 0.6986,
+      "step": 2640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5738890968150132,
+      "learning_rate": 1.987551192214441e-06,
+      "loss": 0.6875,
+      "step": 2641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.141881969545281,
+      "learning_rate": 1.9875359913803917e-06,
+      "loss": 0.6921,
+      "step": 2642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2367747412108219,
+      "learning_rate": 1.9875207813295817e-06,
+      "loss": 0.6816,
+      "step": 2643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8032678648765017,
+      "learning_rate": 1.9875055620621524e-06,
+      "loss": 0.6934,
+      "step": 2644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2559552551390594,
+      "learning_rate": 1.987490333578247e-06,
+      "loss": 0.6927,
+      "step": 2645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.886487870325657,
+      "learning_rate": 1.987475095878007e-06,
+      "loss": 0.6693,
+      "step": 2646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.619991433800253,
+      "learning_rate": 1.9874598489615743e-06,
+      "loss": 0.6966,
+      "step": 2647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2365361266484736,
+      "learning_rate": 1.9874445928290913e-06,
+      "loss": 0.6875,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7580965757838912,
+      "learning_rate": 1.987429327480701e-06,
+      "loss": 0.6842,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6546765680032807,
+      "learning_rate": 1.987414052916545e-06,
+      "loss": 0.6882,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9670806851383773,
+      "learning_rate": 1.9873987691367665e-06,
+      "loss": 0.6921,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.321901899800652,
+      "learning_rate": 1.9873834761415083e-06,
+      "loss": 0.7077,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3708037430503193,
+      "learning_rate": 1.9873681739309124e-06,
+      "loss": 0.6836,
+      "step": 2653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.727222950206262,
+      "learning_rate": 1.9873528625051223e-06,
+      "loss": 0.6947,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9121041766046236,
+      "learning_rate": 1.9873375418642804e-06,
+      "loss": 0.6979,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0180305955746314,
+      "learning_rate": 1.98732221200853e-06,
+      "loss": 0.6914,
+      "step": 2656
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6835117834728306,
+      "learning_rate": 1.987306872938014e-06,
+      "loss": 0.6973,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3716689197652405,
+      "learning_rate": 1.987291524652876e-06,
+      "loss": 0.6875,
+      "step": 2658
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.255876161675588,
+      "learning_rate": 1.9872761671532584e-06,
+      "loss": 0.6758,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2280988959835852,
+      "learning_rate": 1.9872608004393053e-06,
+      "loss": 0.6888,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3142767733801897,
+      "learning_rate": 1.9872454245111596e-06,
+      "loss": 0.6868,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.6370487876779,
+      "learning_rate": 1.9872300393689655e-06,
+      "loss": 0.7012,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.820782695693171,
+      "learning_rate": 1.987214645012866e-06,
+      "loss": 0.6764,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.674497908384114,
+      "learning_rate": 1.9871992414430045e-06,
+      "loss": 0.6771,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1048942397401436,
+      "learning_rate": 1.9871838286595253e-06,
+      "loss": 0.694,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9519413198965808,
+      "learning_rate": 1.9871684066625726e-06,
+      "loss": 0.6966,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4936300647417395,
+      "learning_rate": 1.9871529754522893e-06,
+      "loss": 0.6914,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.345638438193076,
+      "learning_rate": 1.98713753502882e-06,
+      "loss": 0.6986,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.3720812098801005,
+      "learning_rate": 1.987122085392309e-06,
+      "loss": 0.7031,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0391489284839466,
+      "learning_rate": 1.9871066265429004e-06,
+      "loss": 0.6829,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5045503921982175,
+      "learning_rate": 1.9870911584807385e-06,
+      "loss": 0.6953,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7026557752962024,
+      "learning_rate": 1.987075681205967e-06,
+      "loss": 0.6875,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.268882718826785,
+      "learning_rate": 1.9870601947187314e-06,
+      "loss": 0.679,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.626731427009179,
+      "learning_rate": 1.9870446990191754e-06,
+      "loss": 0.6875,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2504530052121015,
+      "learning_rate": 1.987029194107444e-06,
+      "loss": 0.6862,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.594838894076204,
+      "learning_rate": 1.9870136799836818e-06,
+      "loss": 0.6693,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.878473007558301,
+      "learning_rate": 1.9869981566480334e-06,
+      "loss": 0.7096,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9157385946034737,
+      "learning_rate": 1.9869826241006443e-06,
+      "loss": 0.6908,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5754086686427042,
+      "learning_rate": 1.9869670823416592e-06,
+      "loss": 0.6986,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8290270271324258,
+      "learning_rate": 1.9869515313712226e-06,
+      "loss": 0.681,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.748220198018641,
+      "learning_rate": 1.98693597118948e-06,
+      "loss": 0.6803,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6268594156246572,
+      "learning_rate": 1.986920401796577e-06,
+      "loss": 0.6712,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7102154609454012,
+      "learning_rate": 1.9869048231926587e-06,
+      "loss": 0.6823,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.385010573931936,
+      "learning_rate": 1.9868892353778707e-06,
+      "loss": 0.6862,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.0377772262049465,
+      "learning_rate": 1.9868736383523575e-06,
+      "loss": 0.6921,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.679484417286736,
+      "learning_rate": 1.9868580321162656e-06,
+      "loss": 0.6882,
+      "step": 2686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.07939731224141,
+      "learning_rate": 1.986842416669741e-06,
+      "loss": 0.6868,
+      "step": 2687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.112492586476916,
+      "learning_rate": 1.986826792012928e-06,
+      "loss": 0.681,
+      "step": 2688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.3441906884793293,
+      "learning_rate": 1.9868111581459736e-06,
+      "loss": 0.6895,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6846142220796325,
+      "learning_rate": 1.986795515069023e-06,
+      "loss": 0.6979,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.0017926520403515,
+      "learning_rate": 1.9867798627822233e-06,
+      "loss": 0.7031,
+      "step": 2691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9718905950482846,
+      "learning_rate": 1.9867642012857197e-06,
+      "loss": 0.6641,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8997910096511665,
+      "learning_rate": 1.986748530579658e-06,
+      "loss": 0.6888,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7191583375045364,
+      "learning_rate": 1.9867328506641856e-06,
+      "loss": 0.7018,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.219389891542809,
+      "learning_rate": 1.986717161539448e-06,
+      "loss": 0.6901,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.552618363196877,
+      "learning_rate": 1.986701463205592e-06,
+      "loss": 0.6934,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.26478701861701,
+      "learning_rate": 1.986685755662764e-06,
+      "loss": 0.7044,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.42749643515195,
+      "learning_rate": 1.9866700389111105e-06,
+      "loss": 0.6888,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.367789698522069,
+      "learning_rate": 1.9866543129507782e-06,
+      "loss": 0.6732,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.28198069911321,
+      "learning_rate": 1.9866385777819137e-06,
+      "loss": 0.6914,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.311004814212521,
+      "learning_rate": 1.9866228334046644e-06,
+      "loss": 0.6999,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5541657533982574,
+      "learning_rate": 1.986607079819177e-06,
+      "loss": 0.6862,
+      "step": 2702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.757408733355727,
+      "learning_rate": 1.9865913170255985e-06,
+      "loss": 0.679,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8874385504836464,
+      "learning_rate": 1.986575545024076e-06,
+      "loss": 0.6823,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8113989763619143,
+      "learning_rate": 1.9865597638147563e-06,
+      "loss": 0.681,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6553560509082157,
+      "learning_rate": 1.9865439733977875e-06,
+      "loss": 0.6836,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8354153736449231,
+      "learning_rate": 1.9865281737733165e-06,
+      "loss": 0.6973,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0546773919768007,
+      "learning_rate": 1.9865123649414906e-06,
+      "loss": 0.6914,
+      "step": 2708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.835377104922984,
+      "learning_rate": 1.9864965469024576e-06,
+      "loss": 0.6927,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7260727275038332,
+      "learning_rate": 1.9864807196563654e-06,
+      "loss": 0.6836,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.152448676418845,
+      "learning_rate": 1.986464883203361e-06,
+      "loss": 0.6895,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8515444681577244,
+      "learning_rate": 1.9864490375435928e-06,
+      "loss": 0.6901,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.172125528005953,
+      "learning_rate": 1.9864331826772084e-06,
+      "loss": 0.6888,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.492711778017652,
+      "learning_rate": 1.986417318604356e-06,
+      "loss": 0.6973,
+      "step": 2714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3630461110051915,
+      "learning_rate": 1.9864014453251838e-06,
+      "loss": 0.6738,
+      "step": 2715
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.089129343275885,
+      "learning_rate": 1.9863855628398388e-06,
+      "loss": 0.6751,
+      "step": 2716
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.967195283040302,
+      "learning_rate": 1.986369671148471e-06,
+      "loss": 0.6999,
+      "step": 2717
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.951154276549892,
+      "learning_rate": 1.9863537702512274e-06,
+      "loss": 0.6868,
+      "step": 2718
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.22213983741441,
+      "learning_rate": 1.986337860148257e-06,
+      "loss": 0.7012,
+      "step": 2719
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0031969398455267,
+      "learning_rate": 1.986321940839708e-06,
+      "loss": 0.6849,
+      "step": 2720
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6514609929133144,
+      "learning_rate": 1.986306012325729e-06,
+      "loss": 0.6908,
+      "step": 2721
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5484900366301101,
+      "learning_rate": 1.9862900746064694e-06,
+      "loss": 0.6732,
+      "step": 2722
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5241144229280511,
+      "learning_rate": 1.9862741276820765e-06,
+      "loss": 0.6875,
+      "step": 2723
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4937997570975363,
+      "learning_rate": 1.9862581715527004e-06,
+      "loss": 0.6686,
+      "step": 2724
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6003811510502914,
+      "learning_rate": 1.9862422062184893e-06,
+      "loss": 0.6777,
+      "step": 2725
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.842368798043899,
+      "learning_rate": 1.9862262316795924e-06,
+      "loss": 0.681,
+      "step": 2726
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.538346983459479,
+      "learning_rate": 1.986210247936159e-06,
+      "loss": 0.6953,
+      "step": 2727
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3917899658237087,
+      "learning_rate": 1.9861942549883386e-06,
+      "loss": 0.6777,
+      "step": 2728
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1614899697109404,
+      "learning_rate": 1.9861782528362795e-06,
+      "loss": 0.6667,
+      "step": 2729
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8288953217579222,
+      "learning_rate": 1.9861622414801316e-06,
+      "loss": 0.6966,
+      "step": 2730
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8534775802250759,
+      "learning_rate": 1.9861462209200445e-06,
+      "loss": 0.6875,
+      "step": 2731
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.109905375459795,
+      "learning_rate": 1.9861301911561673e-06,
+      "loss": 0.6797,
+      "step": 2732
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.2625627705722913,
+      "learning_rate": 1.98611415218865e-06,
+      "loss": 0.7077,
+      "step": 2733
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.7821169839398125,
+      "learning_rate": 1.986098104017642e-06,
+      "loss": 0.6999,
+      "step": 2734
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.878907596668762,
+      "learning_rate": 1.986082046643293e-06,
+      "loss": 0.7188,
+      "step": 2735
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.9842711772866,
+      "learning_rate": 1.9860659800657532e-06,
+      "loss": 0.707,
+      "step": 2736
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.659735051559842,
+      "learning_rate": 1.9860499042851726e-06,
+      "loss": 0.696,
+      "step": 2737
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.659683115471302,
+      "learning_rate": 1.986033819301701e-06,
+      "loss": 0.6888,
+      "step": 2738
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8590030324471565,
+      "learning_rate": 1.9860177251154882e-06,
+      "loss": 0.6842,
+      "step": 2739
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.917430947113688,
+      "learning_rate": 1.986001621726685e-06,
+      "loss": 0.6784,
+      "step": 2740
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7624334401317565,
+      "learning_rate": 1.985985509135442e-06,
+      "loss": 0.6855,
+      "step": 2741
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8466172796581894,
+      "learning_rate": 1.985969387341908e-06,
+      "loss": 0.6908,
+      "step": 2742
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.050366836334191,
+      "learning_rate": 1.9859532563462353e-06,
+      "loss": 0.6927,
+      "step": 2743
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.932701441874374,
+      "learning_rate": 1.9859371161485733e-06,
+      "loss": 0.6914,
+      "step": 2744
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44804174369772637,
+      "learning_rate": 1.9859209667490727e-06,
+      "loss": 0.6842,
+      "step": 2745
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.514797002070457,
+      "learning_rate": 1.985904808147885e-06,
+      "loss": 0.6914,
+      "step": 2746
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9542797708631272,
+      "learning_rate": 1.9858886403451606e-06,
+      "loss": 0.6745,
+      "step": 2747
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6384226097229617,
+      "learning_rate": 1.9858724633410503e-06,
+      "loss": 0.6868,
+      "step": 2748
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1396228929314802,
+      "learning_rate": 1.985856277135705e-06,
+      "loss": 0.6764,
+      "step": 2749
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.1380479814171895,
+      "learning_rate": 1.9858400817292756e-06,
+      "loss": 0.6953,
+      "step": 2750
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.429319024897904,
+      "learning_rate": 1.985823877121914e-06,
+      "loss": 0.6921,
+      "step": 2751
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.339529857152535,
+      "learning_rate": 1.9858076633137707e-06,
+      "loss": 0.6888,
+      "step": 2752
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.0784003687528125,
+      "learning_rate": 1.9857914403049975e-06,
+      "loss": 0.6927,
+      "step": 2753
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.074400743796269,
+      "learning_rate": 1.9857752080957456e-06,
+      "loss": 0.6836,
+      "step": 2754
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.223848592735763,
+      "learning_rate": 1.9857589666861667e-06,
+      "loss": 0.6895,
+      "step": 2755
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9884451161273982,
+      "learning_rate": 1.9857427160764115e-06,
+      "loss": 0.6882,
+      "step": 2756
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.60717124309206,
+      "learning_rate": 1.985726456266633e-06,
+      "loss": 0.6712,
+      "step": 2757
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4778580158763206,
+      "learning_rate": 1.985710187256982e-06,
+      "loss": 0.6764,
+      "step": 2758
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.731780131970611,
+      "learning_rate": 1.985693909047611e-06,
+      "loss": 0.6784,
+      "step": 2759
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.631549828715404,
+      "learning_rate": 1.9856776216386712e-06,
+      "loss": 0.7298,
+      "step": 2760
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.102163964733258,
+      "learning_rate": 1.985661325030315e-06,
+      "loss": 0.7331,
+      "step": 2761
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.841821353321926,
+      "learning_rate": 1.9856450192226943e-06,
+      "loss": 0.7116,
+      "step": 2762
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.310909472733324,
+      "learning_rate": 1.985628704215962e-06,
+      "loss": 0.7005,
+      "step": 2763
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.534057792793658,
+      "learning_rate": 1.9856123800102695e-06,
+      "loss": 0.6901,
+      "step": 2764
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7622786885280552,
+      "learning_rate": 1.9855960466057695e-06,
+      "loss": 0.6842,
+      "step": 2765
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.196802359194576,
+      "learning_rate": 1.985579704002615e-06,
+      "loss": 0.6823,
+      "step": 2766
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.427342407535922,
+      "learning_rate": 1.985563352200957e-06,
+      "loss": 0.6842,
+      "step": 2767
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.336694258914513,
+      "learning_rate": 1.9855469912009494e-06,
+      "loss": 0.6868,
+      "step": 2768
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.941506695959788,
+      "learning_rate": 1.985530621002745e-06,
+      "loss": 0.6842,
+      "step": 2769
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0306985579809607,
+      "learning_rate": 1.9855142416064957e-06,
+      "loss": 0.6934,
+      "step": 2770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.667996452414096,
+      "learning_rate": 1.9854978530123553e-06,
+      "loss": 0.6999,
+      "step": 2771
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.2964053999572505,
+      "learning_rate": 1.985481455220476e-06,
+      "loss": 0.6882,
+      "step": 2772
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5837814473607903,
+      "learning_rate": 1.985465048231011e-06,
+      "loss": 0.6764,
+      "step": 2773
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.236111897915404,
+      "learning_rate": 1.985448632044114e-06,
+      "loss": 0.709,
+      "step": 2774
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.755348156425577,
+      "learning_rate": 1.9854322066599367e-06,
+      "loss": 0.6921,
+      "step": 2775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5683292233671846,
+      "learning_rate": 1.985415772078635e-06,
+      "loss": 0.6973,
+      "step": 2776
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.661537131309071,
+      "learning_rate": 1.9853993283003598e-06,
+      "loss": 0.6927,
+      "step": 2777
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8487323223658496,
+      "learning_rate": 1.9853828753252656e-06,
+      "loss": 0.681,
+      "step": 2778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.424023328218876,
+      "learning_rate": 1.9853664131535058e-06,
+      "loss": 0.6921,
+      "step": 2779
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6036215598985739,
+      "learning_rate": 1.9853499417852343e-06,
+      "loss": 0.6842,
+      "step": 2780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.202977394696708,
+      "learning_rate": 1.9853334612206047e-06,
+      "loss": 0.6901,
+      "step": 2781
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.960533624940487,
+      "learning_rate": 1.985316971459771e-06,
+      "loss": 0.6882,
+      "step": 2782
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.117343319894245,
+      "learning_rate": 1.9853004725028863e-06,
+      "loss": 0.6927,
+      "step": 2783
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.592332812836963,
+      "learning_rate": 1.9852839643501055e-06,
+      "loss": 0.6836,
+      "step": 2784
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8812959296151113,
+      "learning_rate": 1.9852674470015823e-06,
+      "loss": 0.6862,
+      "step": 2785
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7218475436475636,
+      "learning_rate": 1.9852509204574708e-06,
+      "loss": 0.6947,
+      "step": 2786
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6144154900158058,
+      "learning_rate": 1.9852343847179255e-06,
+      "loss": 0.6953,
+      "step": 2787
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1193741440711225,
+      "learning_rate": 1.9852178397831003e-06,
+      "loss": 0.694,
+      "step": 2788
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.522958153784763,
+      "learning_rate": 1.98520128565315e-06,
+      "loss": 0.694,
+      "step": 2789
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.3295469625742,
+      "learning_rate": 1.985184722328229e-06,
+      "loss": 0.6986,
+      "step": 2790
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6141441852591305,
+      "learning_rate": 1.985168149808492e-06,
+      "loss": 0.6829,
+      "step": 2791
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7921908416960766,
+      "learning_rate": 1.9851515680940932e-06,
+      "loss": 0.6921,
+      "step": 2792
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5558193099973203,
+      "learning_rate": 1.985134977185188e-06,
+      "loss": 0.6862,
+      "step": 2793
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3221552166343626,
+      "learning_rate": 1.985118377081931e-06,
+      "loss": 0.6803,
+      "step": 2794
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.1479336768016575,
+      "learning_rate": 1.9851017677844766e-06,
+      "loss": 0.6823,
+      "step": 2795
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6295515268765236,
+      "learning_rate": 1.9850851492929807e-06,
+      "loss": 0.6725,
+      "step": 2796
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4888216822179035,
+      "learning_rate": 1.985068521607598e-06,
+      "loss": 0.6868,
+      "step": 2797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9451100930340404,
+      "learning_rate": 1.9850518847284833e-06,
+      "loss": 0.6719,
+      "step": 2798
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.205928282549369,
+      "learning_rate": 1.985035238655792e-06,
+      "loss": 0.681,
+      "step": 2799
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.5844167404223315,
+      "learning_rate": 1.9850185833896807e-06,
+      "loss": 0.694,
+      "step": 2800
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.665060502971122,
+      "learning_rate": 1.9850019189303035e-06,
+      "loss": 0.679,
+      "step": 2801
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7738133681608343,
+      "learning_rate": 1.9849852452778156e-06,
+      "loss": 0.7012,
+      "step": 2802
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.679612981918824,
+      "learning_rate": 1.984968562432374e-06,
+      "loss": 0.6908,
+      "step": 2803
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.024200979107528,
+      "learning_rate": 1.9849518703941335e-06,
+      "loss": 0.6628,
+      "step": 2804
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3324227959310293,
+      "learning_rate": 1.98493516916325e-06,
+      "loss": 0.6836,
+      "step": 2805
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0193929387953227,
+      "learning_rate": 1.9849184587398796e-06,
+      "loss": 0.6914,
+      "step": 2806
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.372834278037163,
+      "learning_rate": 1.984901739124178e-06,
+      "loss": 0.6686,
+      "step": 2807
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.247841541979525,
+      "learning_rate": 1.984885010316302e-06,
+      "loss": 0.6784,
+      "step": 2808
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8671810165050837,
+      "learning_rate": 1.984868272316406e-06,
+      "loss": 0.6686,
+      "step": 2809
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.495878249698834,
+      "learning_rate": 1.984851525124648e-06,
+      "loss": 0.6745,
+      "step": 2810
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.063618152023506,
+      "learning_rate": 1.9848347687411834e-06,
+      "loss": 0.6686,
+      "step": 2811
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8554489397687073,
+      "learning_rate": 1.984818003166169e-06,
+      "loss": 0.681,
+      "step": 2812
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1723416277576724,
+      "learning_rate": 1.984801228399761e-06,
+      "loss": 0.7103,
+      "step": 2813
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.420332902254207,
+      "learning_rate": 1.984784444442116e-06,
+      "loss": 0.6784,
+      "step": 2814
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6876444547270133,
+      "learning_rate": 1.9847676512933907e-06,
+      "loss": 0.7305,
+      "step": 2815
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1406922931153576,
+      "learning_rate": 1.9847508489537418e-06,
+      "loss": 0.7051,
+      "step": 2816
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4408440732891874,
+      "learning_rate": 1.9847340374233263e-06,
+      "loss": 0.6771,
+      "step": 2817
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.587280379546873,
+      "learning_rate": 1.984717216702301e-06,
+      "loss": 0.6934,
+      "step": 2818
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.076342760524867,
+      "learning_rate": 1.9847003867908227e-06,
+      "loss": 0.6908,
+      "step": 2819
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.686302628574128,
+      "learning_rate": 1.984683547689048e-06,
+      "loss": 0.6921,
+      "step": 2820
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6568980337666327,
+      "learning_rate": 1.9846666993971355e-06,
+      "loss": 0.6973,
+      "step": 2821
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.76877514055872,
+      "learning_rate": 1.9846498419152414e-06,
+      "loss": 0.6908,
+      "step": 2822
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.471143336506184,
+      "learning_rate": 1.984632975243523e-06,
+      "loss": 0.7083,
+      "step": 2823
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.410405917484212,
+      "learning_rate": 1.9846160993821386e-06,
+      "loss": 0.7064,
+      "step": 2824
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.309959108848512,
+      "learning_rate": 1.9845992143312444e-06,
+      "loss": 0.7116,
+      "step": 2825
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.585234890645064,
+      "learning_rate": 1.9845823200909987e-06,
+      "loss": 0.679,
+      "step": 2826
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.0985605534627965,
+      "learning_rate": 1.9845654166615593e-06,
+      "loss": 0.6986,
+      "step": 2827
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9206437117341134,
+      "learning_rate": 1.984548504043084e-06,
+      "loss": 0.7005,
+      "step": 2828
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.7362122055304265,
+      "learning_rate": 1.9845315822357303e-06,
+      "loss": 0.6914,
+      "step": 2829
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.074714489763068,
+      "learning_rate": 1.9845146512396562e-06,
+      "loss": 0.696,
+      "step": 2830
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8017566435732338,
+      "learning_rate": 1.98449771105502e-06,
+      "loss": 0.696,
+      "step": 2831
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.392953866532917,
+      "learning_rate": 1.984480761681979e-06,
+      "loss": 0.6895,
+      "step": 2832
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.403439869782384,
+      "learning_rate": 1.9844638031206923e-06,
+      "loss": 0.6849,
+      "step": 2833
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.534075420709103,
+      "learning_rate": 1.9844468353713182e-06,
+      "loss": 0.6842,
+      "step": 2834
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.590874140437125,
+      "learning_rate": 1.9844298584340143e-06,
+      "loss": 0.6829,
+      "step": 2835
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2902415618907783,
+      "learning_rate": 1.98441287230894e-06,
+      "loss": 0.694,
+      "step": 2836
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.478084336368275,
+      "learning_rate": 1.984395876996253e-06,
+      "loss": 0.6862,
+      "step": 2837
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.765842111156135,
+      "learning_rate": 1.9843788724961123e-06,
+      "loss": 0.6908,
+      "step": 2838
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9093188917186095,
+      "learning_rate": 1.9843618588086763e-06,
+      "loss": 0.6895,
+      "step": 2839
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7698835825266286,
+      "learning_rate": 1.984344835934104e-06,
+      "loss": 0.6803,
+      "step": 2840
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0109644368916113,
+      "learning_rate": 1.9843278038725544e-06,
+      "loss": 0.6895,
+      "step": 2841
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5014680143037489,
+      "learning_rate": 1.9843107626241867e-06,
+      "loss": 0.681,
+      "step": 2842
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.863681846398097,
+      "learning_rate": 1.9842937121891592e-06,
+      "loss": 0.6914,
+      "step": 2843
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.014957755284304,
+      "learning_rate": 1.9842766525676313e-06,
+      "loss": 0.6901,
+      "step": 2844
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9323124896237116,
+      "learning_rate": 1.984259583759763e-06,
+      "loss": 0.6829,
+      "step": 2845
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7948962071141144,
+      "learning_rate": 1.984242505765712e-06,
+      "loss": 0.6875,
+      "step": 2846
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8971376541307475,
+      "learning_rate": 1.9842254185856395e-06,
+      "loss": 0.6953,
+      "step": 2847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.15915280694592,
+      "learning_rate": 1.9842083222197037e-06,
+      "loss": 0.6836,
+      "step": 2848
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.142296534639556,
+      "learning_rate": 1.984191216668065e-06,
+      "loss": 0.6875,
+      "step": 2849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9337368373222449,
+      "learning_rate": 1.9841741019308824e-06,
+      "loss": 0.6914,
+      "step": 2850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.3782035351296145,
+      "learning_rate": 1.9841569780083157e-06,
+      "loss": 0.6947,
+      "step": 2851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6096856247817002,
+      "learning_rate": 1.984139844900525e-06,
+      "loss": 0.6868,
+      "step": 2852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.866845076098853,
+      "learning_rate": 1.9841227026076705e-06,
+      "loss": 0.6868,
+      "step": 2853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.409108875038111,
+      "learning_rate": 1.9841055511299112e-06,
+      "loss": 0.6882,
+      "step": 2854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5443903345292074,
+      "learning_rate": 1.984088390467408e-06,
+      "loss": 0.681,
+      "step": 2855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.751815042289893,
+      "learning_rate": 1.9840712206203208e-06,
+      "loss": 0.6934,
+      "step": 2856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6103679544976477,
+      "learning_rate": 1.9840540415888097e-06,
+      "loss": 0.6816,
+      "step": 2857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.939559713238672,
+      "learning_rate": 1.9840368533730356e-06,
+      "loss": 0.6953,
+      "step": 2858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.112185991895307,
+      "learning_rate": 1.984019655973158e-06,
+      "loss": 0.6966,
+      "step": 2859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1359174797802425,
+      "learning_rate": 1.9840024493893386e-06,
+      "loss": 0.7025,
+      "step": 2860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3901796551167567,
+      "learning_rate": 1.9839852336217367e-06,
+      "loss": 0.6849,
+      "step": 2861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.024008269444834,
+      "learning_rate": 1.983968008670514e-06,
+      "loss": 0.6803,
+      "step": 2862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.223270358188794,
+      "learning_rate": 1.9839507745358304e-06,
+      "loss": 0.707,
+      "step": 2863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9019642854210512,
+      "learning_rate": 1.983933531217848e-06,
+      "loss": 0.694,
+      "step": 2864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.036046558829472,
+      "learning_rate": 1.983916278716726e-06,
+      "loss": 0.6999,
+      "step": 2865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.441517441991927,
+      "learning_rate": 1.9838990170326268e-06,
+      "loss": 0.6882,
+      "step": 2866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.084624293963832,
+      "learning_rate": 1.9838817461657112e-06,
+      "loss": 0.6882,
+      "step": 2867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9900174878450365,
+      "learning_rate": 1.9838644661161403e-06,
+      "loss": 0.6719,
+      "step": 2868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.277037696343748,
+      "learning_rate": 1.983847176884075e-06,
+      "loss": 0.6875,
+      "step": 2869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4100739211661824,
+      "learning_rate": 1.983829878469677e-06,
+      "loss": 0.6868,
+      "step": 2870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.076270226452727,
+      "learning_rate": 1.983812570873108e-06,
+      "loss": 0.6927,
+      "step": 2871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8310104262529219,
+      "learning_rate": 1.983795254094529e-06,
+      "loss": 0.6855,
+      "step": 2872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3982457747415395,
+      "learning_rate": 1.983777928134102e-06,
+      "loss": 0.6966,
+      "step": 2873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5246796255081403,
+      "learning_rate": 1.983760592991989e-06,
+      "loss": 0.6797,
+      "step": 2874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8804392173450837,
+      "learning_rate": 1.983743248668351e-06,
+      "loss": 0.6842,
+      "step": 2875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4086110894928587,
+      "learning_rate": 1.98372589516335e-06,
+      "loss": 0.6882,
+      "step": 2876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.5067205461054005,
+      "learning_rate": 1.9837085324771487e-06,
+      "loss": 0.6732,
+      "step": 2877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4666578620041753,
+      "learning_rate": 1.983691160609908e-06,
+      "loss": 0.7005,
+      "step": 2878
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.9415460886607185,
+      "learning_rate": 1.9836737795617917e-06,
+      "loss": 0.6992,
+      "step": 2879
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.436168981780964,
+      "learning_rate": 1.9836563893329604e-06,
+      "loss": 0.6966,
+      "step": 2880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.426235191363586,
+      "learning_rate": 1.983638989923577e-06,
+      "loss": 0.6895,
+      "step": 2881
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.681513800371783,
+      "learning_rate": 1.983621581333804e-06,
+      "loss": 0.6966,
+      "step": 2882
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6286999802364128,
+      "learning_rate": 1.983604163563804e-06,
+      "loss": 0.6712,
+      "step": 2883
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.658821717781594,
+      "learning_rate": 1.983586736613739e-06,
+      "loss": 0.6895,
+      "step": 2884
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.594574947073663,
+      "learning_rate": 1.9835693004837722e-06,
+      "loss": 0.6842,
+      "step": 2885
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.584619281745614,
+      "learning_rate": 1.9835518551740664e-06,
+      "loss": 0.6712,
+      "step": 2886
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9189798820864774,
+      "learning_rate": 1.9835344006847838e-06,
+      "loss": 0.6947,
+      "step": 2887
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.481417150169232,
+      "learning_rate": 1.9835169370160878e-06,
+      "loss": 0.6784,
+      "step": 2888
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0083347460247145,
+      "learning_rate": 1.9834994641681414e-06,
+      "loss": 0.7038,
+      "step": 2889
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.605904371415717,
+      "learning_rate": 1.983481982141107e-06,
+      "loss": 0.6966,
+      "step": 2890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2916364536232898,
+      "learning_rate": 1.983464490935149e-06,
+      "loss": 0.6927,
+      "step": 2891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9243942950983404,
+      "learning_rate": 1.9834469905504295e-06,
+      "loss": 0.6888,
+      "step": 2892
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.1675771710051785,
+      "learning_rate": 1.983429480987112e-06,
+      "loss": 0.6862,
+      "step": 2893
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.538487072537848,
+      "learning_rate": 1.9834119622453604e-06,
+      "loss": 0.6816,
+      "step": 2894
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3751127499260443,
+      "learning_rate": 1.9833944343253384e-06,
+      "loss": 0.6914,
+      "step": 2895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.552077544908535,
+      "learning_rate": 1.983376897227209e-06,
+      "loss": 0.6953,
+      "step": 2896
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.764331220281935,
+      "learning_rate": 1.983359350951136e-06,
+      "loss": 0.6934,
+      "step": 2897
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.174791187751516,
+      "learning_rate": 1.9833417954972827e-06,
+      "loss": 0.7012,
+      "step": 2898
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.228034417054952,
+      "learning_rate": 1.983324230865814e-06,
+      "loss": 0.6803,
+      "step": 2899
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.388242724081222,
+      "learning_rate": 1.983306657056893e-06,
+      "loss": 0.6895,
+      "step": 2900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.349692227747338,
+      "learning_rate": 1.983289074070684e-06,
+      "loss": 0.6895,
+      "step": 2901
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3863414049508,
+      "learning_rate": 1.9832714819073518e-06,
+      "loss": 0.6914,
+      "step": 2902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.204626097901225,
+      "learning_rate": 1.9832538805670595e-06,
+      "loss": 0.6921,
+      "step": 2903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.3475404747315824,
+      "learning_rate": 1.9832362700499715e-06,
+      "loss": 0.6888,
+      "step": 2904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.4166165706593405,
+      "learning_rate": 1.9832186503562528e-06,
+      "loss": 0.6816,
+      "step": 2905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9383034069146707,
+      "learning_rate": 1.9832010214860674e-06,
+      "loss": 0.7031,
+      "step": 2906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0100201412852734,
+      "learning_rate": 1.9831833834395795e-06,
+      "loss": 0.6719,
+      "step": 2907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.443525120326846,
+      "learning_rate": 1.9831657362169545e-06,
+      "loss": 0.7207,
+      "step": 2908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.306117754931878,
+      "learning_rate": 1.983148079818357e-06,
+      "loss": 0.696,
+      "step": 2909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2389685375528456,
+      "learning_rate": 1.9831304142439507e-06,
+      "loss": 0.6829,
+      "step": 2910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3533171955572068,
+      "learning_rate": 1.983112739493902e-06,
+      "loss": 0.6849,
+      "step": 2911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8537546584749542,
+      "learning_rate": 1.9830950555683753e-06,
+      "loss": 0.6862,
+      "step": 2912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.84899886376441,
+      "learning_rate": 1.983077362467535e-06,
+      "loss": 0.6888,
+      "step": 2913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9208632820445138,
+      "learning_rate": 1.983059660191547e-06,
+      "loss": 0.6888,
+      "step": 2914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6453340308251554,
+      "learning_rate": 1.983041948740576e-06,
+      "loss": 0.6882,
+      "step": 2915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.158513085865327,
+      "learning_rate": 1.983024228114788e-06,
+      "loss": 0.6986,
+      "step": 2916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.230352402070574,
+      "learning_rate": 1.9830064983143474e-06,
+      "loss": 0.6953,
+      "step": 2917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.743048385381829,
+      "learning_rate": 1.9829887593394204e-06,
+      "loss": 0.6953,
+      "step": 2918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5502402931446573,
+      "learning_rate": 1.9829710111901728e-06,
+      "loss": 0.6823,
+      "step": 2919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2215204787121436,
+      "learning_rate": 1.9829532538667693e-06,
+      "loss": 0.6901,
+      "step": 2920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.611428629618499,
+      "learning_rate": 1.9829354873693764e-06,
+      "loss": 0.6823,
+      "step": 2921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.589828677148211,
+      "learning_rate": 1.98291771169816e-06,
+      "loss": 0.6758,
+      "step": 2922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.869295446781418,
+      "learning_rate": 1.982899926853285e-06,
+      "loss": 0.6725,
+      "step": 2923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6088386110085334,
+      "learning_rate": 1.9828821328349183e-06,
+      "loss": 0.6823,
+      "step": 2924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3440849046551606,
+      "learning_rate": 1.9828643296432256e-06,
+      "loss": 0.6855,
+      "step": 2925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.448663052051138,
+      "learning_rate": 1.9828465172783735e-06,
+      "loss": 0.6836,
+      "step": 2926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8886292744355426,
+      "learning_rate": 1.9828286957405278e-06,
+      "loss": 0.6921,
+      "step": 2927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.675350587327813,
+      "learning_rate": 1.982810865029855e-06,
+      "loss": 0.6777,
+      "step": 2928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.236924628609626,
+      "learning_rate": 1.9827930251465218e-06,
+      "loss": 0.6862,
+      "step": 2929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.912686651084785,
+      "learning_rate": 1.982775176090694e-06,
+      "loss": 0.6712,
+      "step": 2930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5194568001863912,
+      "learning_rate": 1.9827573178625383e-06,
+      "loss": 0.6862,
+      "step": 2931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.312797906560409,
+      "learning_rate": 1.9827394504622223e-06,
+      "loss": 0.696,
+      "step": 2932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.28969733982192,
+      "learning_rate": 1.9827215738899118e-06,
+      "loss": 0.7012,
+      "step": 2933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0874079779805752,
+      "learning_rate": 1.982703688145774e-06,
+      "loss": 0.6829,
+      "step": 2934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7048633908982191,
+      "learning_rate": 1.9826857932299755e-06,
+      "loss": 0.7031,
+      "step": 2935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0042145725106915,
+      "learning_rate": 1.9826678891426835e-06,
+      "loss": 0.6875,
+      "step": 2936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.62359593727324,
+      "learning_rate": 1.9826499758840656e-06,
+      "loss": 0.6803,
+      "step": 2937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.428462355769395,
+      "learning_rate": 1.9826320534542883e-06,
+      "loss": 0.6895,
+      "step": 2938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3074382277959615,
+      "learning_rate": 1.9826141218535197e-06,
+      "loss": 0.6751,
+      "step": 2939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.859843682762858,
+      "learning_rate": 1.9825961810819258e-06,
+      "loss": 0.6875,
+      "step": 2940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.161550730706648,
+      "learning_rate": 1.982578231139675e-06,
+      "loss": 0.7005,
+      "step": 2941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1967563341638043,
+      "learning_rate": 1.9825602720269354e-06,
+      "loss": 0.6797,
+      "step": 2942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.460090338348227,
+      "learning_rate": 1.982542303743873e-06,
+      "loss": 0.6641,
+      "step": 2943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.499817196100047,
+      "learning_rate": 1.982524326290657e-06,
+      "loss": 0.6947,
+      "step": 2944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.953142678372877,
+      "learning_rate": 1.9825063396674542e-06,
+      "loss": 0.6973,
+      "step": 2945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.761483987678275,
+      "learning_rate": 1.982488343874433e-06,
+      "loss": 0.6875,
+      "step": 2946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7586879932372312,
+      "learning_rate": 1.9824703389117614e-06,
+      "loss": 0.6908,
+      "step": 2947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9850632766021439,
+      "learning_rate": 1.982452324779607e-06,
+      "loss": 0.6797,
+      "step": 2948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2335448190073977,
+      "learning_rate": 1.982434301478138e-06,
+      "loss": 0.6654,
+      "step": 2949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.803104456942374,
+      "learning_rate": 1.982416269007523e-06,
+      "loss": 0.6784,
+      "step": 2950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.652986902244935,
+      "learning_rate": 1.9823982273679302e-06,
+      "loss": 0.7025,
+      "step": 2951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.754084792199685,
+      "learning_rate": 1.9823801765595275e-06,
+      "loss": 0.6855,
+      "step": 2952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.938286185736328,
+      "learning_rate": 1.9823621165824843e-06,
+      "loss": 0.6973,
+      "step": 2953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8190869807519975,
+      "learning_rate": 1.982344047436968e-06,
+      "loss": 0.6901,
+      "step": 2954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.40123442743233,
+      "learning_rate": 1.982325969123148e-06,
+      "loss": 0.7038,
+      "step": 2955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.25763890188319,
+      "learning_rate": 1.982307881641193e-06,
+      "loss": 0.6868,
+      "step": 2956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2506099084517297,
+      "learning_rate": 1.9822897849912715e-06,
+      "loss": 0.6849,
+      "step": 2957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.016384314284113,
+      "learning_rate": 1.9822716791735527e-06,
+      "loss": 0.7005,
+      "step": 2958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.605293932049143,
+      "learning_rate": 1.9822535641882054e-06,
+      "loss": 0.6849,
+      "step": 2959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7796442160775683,
+      "learning_rate": 1.982235440035399e-06,
+      "loss": 0.7116,
+      "step": 2960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.5289449185133,
+      "learning_rate": 1.982217306715302e-06,
+      "loss": 0.6849,
+      "step": 2961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0353120888406284,
+      "learning_rate": 1.982199164228084e-06,
+      "loss": 0.6732,
+      "step": 2962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.509309579487616,
+      "learning_rate": 1.982181012573915e-06,
+      "loss": 0.6784,
+      "step": 2963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.936904343334161,
+      "learning_rate": 1.982162851752963e-06,
+      "loss": 0.7005,
+      "step": 2964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5509396941786273,
+      "learning_rate": 1.9821446817653983e-06,
+      "loss": 0.6758,
+      "step": 2965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.972364033129919,
+      "learning_rate": 1.9821265026113908e-06,
+      "loss": 0.6803,
+      "step": 2966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3049406723365267,
+      "learning_rate": 1.9821083142911095e-06,
+      "loss": 0.696,
+      "step": 2967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.239893851389402,
+      "learning_rate": 1.9820901168047243e-06,
+      "loss": 0.6849,
+      "step": 2968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0181668307691347,
+      "learning_rate": 1.982071910152406e-06,
+      "loss": 0.6751,
+      "step": 2969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.040217771842623,
+      "learning_rate": 1.982053694334323e-06,
+      "loss": 0.6927,
+      "step": 2970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0452993493226437,
+      "learning_rate": 1.982035469350646e-06,
+      "loss": 0.6771,
+      "step": 2971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.440599746729316,
+      "learning_rate": 1.9820172352015453e-06,
+      "loss": 0.6738,
+      "step": 2972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6329439651920824,
+      "learning_rate": 1.981998991887191e-06,
+      "loss": 0.6816,
+      "step": 2973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.0261269626771865,
+      "learning_rate": 1.981980739407753e-06,
+      "loss": 0.7018,
+      "step": 2974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.959158090131071,
+      "learning_rate": 1.981962477763402e-06,
+      "loss": 0.6979,
+      "step": 2975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3334731164748646,
+      "learning_rate": 1.9819442069543085e-06,
+      "loss": 0.6738,
+      "step": 2976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.4899825835431555,
+      "learning_rate": 1.981925926980643e-06,
+      "loss": 0.6777,
+      "step": 2977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4735388022035956,
+      "learning_rate": 1.9819076378425755e-06,
+      "loss": 0.6882,
+      "step": 2978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8066834709537599,
+      "learning_rate": 1.9818893395402774e-06,
+      "loss": 0.6751,
+      "step": 2979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.998998907525343,
+      "learning_rate": 1.9818710320739193e-06,
+      "loss": 0.6823,
+      "step": 2980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.143724495836616,
+      "learning_rate": 1.981852715443672e-06,
+      "loss": 0.694,
+      "step": 2981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.0668050620519995,
+      "learning_rate": 1.981834389649707e-06,
+      "loss": 0.6927,
+      "step": 2982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.039039806802821,
+      "learning_rate": 1.981816054692194e-06,
+      "loss": 0.6901,
+      "step": 2983
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.474797907457845,
+      "learning_rate": 1.9817977105713053e-06,
+      "loss": 0.7103,
+      "step": 2984
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.363754289014343,
+      "learning_rate": 1.9817793572872117e-06,
+      "loss": 0.6979,
+      "step": 2985
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9583961178999942,
+      "learning_rate": 1.9817609948400847e-06,
+      "loss": 0.6855,
+      "step": 2986
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.919889704207995,
+      "learning_rate": 1.981742623230095e-06,
+      "loss": 0.7005,
+      "step": 2987
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7668654726565185,
+      "learning_rate": 1.9817242424574152e-06,
+      "loss": 0.6953,
+      "step": 2988
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6350404802972804,
+      "learning_rate": 1.981705852522216e-06,
+      "loss": 0.6849,
+      "step": 2989
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.544051687930288,
+      "learning_rate": 1.9816874534246694e-06,
+      "loss": 0.7005,
+      "step": 2990
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4930473935247102,
+      "learning_rate": 1.9816690451649464e-06,
+      "loss": 0.6686,
+      "step": 2991
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6715745134382116,
+      "learning_rate": 1.98165062774322e-06,
+      "loss": 0.6712,
+      "step": 2992
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4260709381435643,
+      "learning_rate": 1.981632201159661e-06,
+      "loss": 0.6823,
+      "step": 2993
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7000961780690039,
+      "learning_rate": 1.9816137654144423e-06,
+      "loss": 0.6836,
+      "step": 2994
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.851148847954693,
+      "learning_rate": 1.9815953205077355e-06,
+      "loss": 0.6901,
+      "step": 2995
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.729830574640234,
+      "learning_rate": 1.9815768664397125e-06,
+      "loss": 0.7051,
+      "step": 2996
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.005634936858878,
+      "learning_rate": 1.981558403210546e-06,
+      "loss": 0.6816,
+      "step": 2997
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.598300080061404,
+      "learning_rate": 1.981539930820408e-06,
+      "loss": 0.6829,
+      "step": 2998
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5122823996541779,
+      "learning_rate": 1.981521449269471e-06,
+      "loss": 0.6836,
+      "step": 2999
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.095502539072187,
+      "learning_rate": 1.9815029585579075e-06,
+      "loss": 0.6895,
+      "step": 3000
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7473129055130263,
+      "learning_rate": 1.98148445868589e-06,
+      "loss": 0.6934,
+      "step": 3001
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.190014414186361,
+      "learning_rate": 1.981465949653592e-06,
+      "loss": 0.6758,
+      "step": 3002
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6847211165585547,
+      "learning_rate": 1.9814474314611844e-06,
+      "loss": 0.6895,
+      "step": 3003
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.341259585037921,
+      "learning_rate": 1.9814289041088416e-06,
+      "loss": 0.6784,
+      "step": 3004
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5318821760324388,
+      "learning_rate": 1.981410367596736e-06,
+      "loss": 0.6855,
+      "step": 3005
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5404281238233862,
+      "learning_rate": 1.981391821925041e-06,
+      "loss": 0.6842,
+      "step": 3006
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.3512874756181805,
+      "learning_rate": 1.981373267093929e-06,
+      "loss": 0.6816,
+      "step": 3007
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.070356223054269,
+      "learning_rate": 1.9813547031035734e-06,
+      "loss": 0.707,
+      "step": 3008
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3232122530146664,
+      "learning_rate": 1.981336129954148e-06,
+      "loss": 0.7064,
+      "step": 3009
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1800787950993494,
+      "learning_rate": 1.981317547645825e-06,
+      "loss": 0.6816,
+      "step": 3010
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.746604948832256,
+      "learning_rate": 1.9812989561787794e-06,
+      "loss": 0.6719,
+      "step": 3011
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.665345387857187,
+      "learning_rate": 1.9812803555531832e-06,
+      "loss": 0.6849,
+      "step": 3012
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.26303624619338,
+      "learning_rate": 1.981261745769211e-06,
+      "loss": 0.6973,
+      "step": 3013
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.642725879884296,
+      "learning_rate": 1.9812431268270363e-06,
+      "loss": 0.7012,
+      "step": 3014
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.031519257166516,
+      "learning_rate": 1.981224498726832e-06,
+      "loss": 0.6875,
+      "step": 3015
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9982833082505334,
+      "learning_rate": 1.9812058614687736e-06,
+      "loss": 0.6914,
+      "step": 3016
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0288851025933043,
+      "learning_rate": 1.9811872150530337e-06,
+      "loss": 0.6882,
+      "step": 3017
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.268683834228295,
+      "learning_rate": 1.9811685594797867e-06,
+      "loss": 0.6888,
+      "step": 3018
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0887147207803434,
+      "learning_rate": 1.981149894749207e-06,
+      "loss": 0.6777,
+      "step": 3019
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.225650666921747,
+      "learning_rate": 1.9811312208614682e-06,
+      "loss": 0.6706,
+      "step": 3020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.885039701234593,
+      "learning_rate": 1.981112537816745e-06,
+      "loss": 0.6882,
+      "step": 3021
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.123319857493474,
+      "learning_rate": 1.981093845615212e-06,
+      "loss": 0.6816,
+      "step": 3022
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.08778630212537,
+      "learning_rate": 1.981075144257043e-06,
+      "loss": 0.6855,
+      "step": 3023
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9991380599653888,
+      "learning_rate": 1.981056433742413e-06,
+      "loss": 0.6888,
+      "step": 3024
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.500843690123799,
+      "learning_rate": 1.9810377140714968e-06,
+      "loss": 0.6803,
+      "step": 3025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.839228570160179,
+      "learning_rate": 1.981018985244469e-06,
+      "loss": 0.6777,
+      "step": 3026
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8671371565519894,
+      "learning_rate": 1.981000247261504e-06,
+      "loss": 0.6868,
+      "step": 3027
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5698405523029593,
+      "learning_rate": 1.9809815001227766e-06,
+      "loss": 0.6784,
+      "step": 3028
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2988255533424193,
+      "learning_rate": 1.980962743828463e-06,
+      "loss": 0.6816,
+      "step": 3029
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8827309517194194,
+      "learning_rate": 1.9809439783787364e-06,
+      "loss": 0.6868,
+      "step": 3030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.891610035360143,
+      "learning_rate": 1.980925203773773e-06,
+      "loss": 0.707,
+      "step": 3031
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.561779767771808,
+      "learning_rate": 1.9809064200137486e-06,
+      "loss": 0.6862,
+      "step": 3032
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3286953387100038,
+      "learning_rate": 1.9808876270988374e-06,
+      "loss": 0.6816,
+      "step": 3033
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.872056980126314,
+      "learning_rate": 1.9808688250292153e-06,
+      "loss": 0.6908,
+      "step": 3034
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2222248795447732,
+      "learning_rate": 1.9808500138050576e-06,
+      "loss": 0.6888,
+      "step": 3035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.481345347824915,
+      "learning_rate": 1.9808311934265397e-06,
+      "loss": 0.6947,
+      "step": 3036
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.579566988658866,
+      "learning_rate": 1.980812363893838e-06,
+      "loss": 0.7018,
+      "step": 3037
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.620609810909627,
+      "learning_rate": 1.980793525207128e-06,
+      "loss": 0.6751,
+      "step": 3038
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6499587938995328,
+      "learning_rate": 1.980774677366585e-06,
+      "loss": 0.6947,
+      "step": 3039
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.73001415913525,
+      "learning_rate": 1.980755820372385e-06,
+      "loss": 0.7142,
+      "step": 3040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.034940298166839,
+      "learning_rate": 1.9807369542247042e-06,
+      "loss": 0.7096,
+      "step": 3041
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6388102757567165,
+      "learning_rate": 1.980718078923719e-06,
+      "loss": 0.6934,
+      "step": 3042
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.642097218635058,
+      "learning_rate": 1.980699194469605e-06,
+      "loss": 0.6973,
+      "step": 3043
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.571573736339227,
+      "learning_rate": 1.9806803008625388e-06,
+      "loss": 0.6895,
+      "step": 3044
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.2308109888543015,
+      "learning_rate": 1.9806613981026967e-06,
+      "loss": 0.6862,
+      "step": 3045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2930830366725536,
+      "learning_rate": 1.9806424861902545e-06,
+      "loss": 0.6868,
+      "step": 3046
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.185624109974866,
+      "learning_rate": 1.98062356512539e-06,
+      "loss": 0.6908,
+      "step": 3047
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6173211158051823,
+      "learning_rate": 1.9806046349082787e-06,
+      "loss": 0.6816,
+      "step": 3048
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1448249015430365,
+      "learning_rate": 1.9805856955390975e-06,
+      "loss": 0.6842,
+      "step": 3049
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2351895998987525,
+      "learning_rate": 1.9805667470180235e-06,
+      "loss": 0.6914,
+      "step": 3050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4224246656823515,
+      "learning_rate": 1.9805477893452332e-06,
+      "loss": 0.7018,
+      "step": 3051
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.209761235118297,
+      "learning_rate": 1.9805288225209037e-06,
+      "loss": 0.6862,
+      "step": 3052
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8490707822653216,
+      "learning_rate": 1.980509846545212e-06,
+      "loss": 0.6836,
+      "step": 3053
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6386159549401704,
+      "learning_rate": 1.9804908614183353e-06,
+      "loss": 0.6712,
+      "step": 3054
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2122976882876313,
+      "learning_rate": 1.9804718671404504e-06,
+      "loss": 0.6914,
+      "step": 3055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3486249264636195,
+      "learning_rate": 1.9804528637117352e-06,
+      "loss": 0.6914,
+      "step": 3056
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.605385072947633,
+      "learning_rate": 1.9804338511323665e-06,
+      "loss": 0.694,
+      "step": 3057
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5957200727418956,
+      "learning_rate": 1.9804148294025224e-06,
+      "loss": 0.6875,
+      "step": 3058
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.929055666992203,
+      "learning_rate": 1.9803957985223795e-06,
+      "loss": 0.6875,
+      "step": 3059
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4504051351529255,
+      "learning_rate": 1.980376758492116e-06,
+      "loss": 0.6842,
+      "step": 3060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0061971564687076,
+      "learning_rate": 1.9803577093119092e-06,
+      "loss": 0.6862,
+      "step": 3061
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7195847209323913,
+      "learning_rate": 1.980338650981938e-06,
+      "loss": 0.6992,
+      "step": 3062
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5569311415913263,
+      "learning_rate": 1.9803195835023785e-06,
+      "loss": 0.6777,
+      "step": 3063
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8578800202553742,
+      "learning_rate": 1.9803005068734102e-06,
+      "loss": 0.6875,
+      "step": 3064
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.289533506676331,
+      "learning_rate": 1.980281421095211e-06,
+      "loss": 0.6999,
+      "step": 3065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.313166783725178,
+      "learning_rate": 1.980262326167958e-06,
+      "loss": 0.7012,
+      "step": 3066
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.150554952939769,
+      "learning_rate": 1.98024322209183e-06,
+      "loss": 0.6816,
+      "step": 3067
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8180990240893173,
+      "learning_rate": 1.9802241088670055e-06,
+      "loss": 0.6908,
+      "step": 3068
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.506800437297779,
+      "learning_rate": 1.980204986493663e-06,
+      "loss": 0.694,
+      "step": 3069
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4804253797473008,
+      "learning_rate": 1.98018585497198e-06,
+      "loss": 0.6797,
+      "step": 3070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1908431436041762,
+      "learning_rate": 1.980166714302136e-06,
+      "loss": 0.681,
+      "step": 3071
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48236764373473195,
+      "learning_rate": 1.9801475644843094e-06,
+      "loss": 0.6862,
+      "step": 3072
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.390393574196901,
+      "learning_rate": 1.9801284055186793e-06,
+      "loss": 0.7057,
+      "step": 3073
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4080996165653383,
+      "learning_rate": 1.9801092374054233e-06,
+      "loss": 0.6836,
+      "step": 3074
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.601176259288163,
+      "learning_rate": 1.9800900601447218e-06,
+      "loss": 0.6986,
+      "step": 3075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5490763487122594,
+      "learning_rate": 1.9800708737367527e-06,
+      "loss": 0.6875,
+      "step": 3076
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.80999226758673,
+      "learning_rate": 1.9800516781816955e-06,
+      "loss": 0.679,
+      "step": 3077
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9419144568328839,
+      "learning_rate": 1.9800324734797294e-06,
+      "loss": 0.6888,
+      "step": 3078
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.061369564914996,
+      "learning_rate": 1.980013259631034e-06,
+      "loss": 0.6784,
+      "step": 3079
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.27162224541536,
+      "learning_rate": 1.9799940366357877e-06,
+      "loss": 0.6712,
+      "step": 3080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.450923681868789,
+      "learning_rate": 1.9799748044941706e-06,
+      "loss": 0.6875,
+      "step": 3081
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.608364224197894,
+      "learning_rate": 1.9799555632063617e-06,
+      "loss": 0.6745,
+      "step": 3082
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.24115139384331,
+      "learning_rate": 1.979936312772541e-06,
+      "loss": 0.6751,
+      "step": 3083
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2770122702708595,
+      "learning_rate": 1.9799170531928883e-06,
+      "loss": 0.6855,
+      "step": 3084
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6007879974874349,
+      "learning_rate": 1.9798977844675824e-06,
+      "loss": 0.6836,
+      "step": 3085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8994664174253713,
+      "learning_rate": 1.9798785065968046e-06,
+      "loss": 0.6973,
+      "step": 3086
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.722164426947481,
+      "learning_rate": 1.9798592195807337e-06,
+      "loss": 0.7174,
+      "step": 3087
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.566801349011789,
+      "learning_rate": 1.97983992341955e-06,
+      "loss": 0.6927,
+      "step": 3088
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.067255822129543,
+      "learning_rate": 1.979820618113434e-06,
+      "loss": 0.6745,
+      "step": 3089
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.245193725312089,
+      "learning_rate": 1.979801303662565e-06,
+      "loss": 0.6686,
+      "step": 3090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.460472073798966,
+      "learning_rate": 1.979781980067124e-06,
+      "loss": 0.6947,
+      "step": 3091
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4888863832974013,
+      "learning_rate": 1.979762647327291e-06,
+      "loss": 0.6693,
+      "step": 3092
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5755371231368285,
+      "learning_rate": 1.9797433054432466e-06,
+      "loss": 0.6764,
+      "step": 3093
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1816493552453124,
+      "learning_rate": 1.9797239544151717e-06,
+      "loss": 0.6855,
+      "step": 3094
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.12593606572561,
+      "learning_rate": 1.979704594243246e-06,
+      "loss": 0.7214,
+      "step": 3095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.770985360455103,
+      "learning_rate": 1.979685224927651e-06,
+      "loss": 0.6855,
+      "step": 3096
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.996820863572172,
+      "learning_rate": 1.979665846468567e-06,
+      "loss": 0.679,
+      "step": 3097
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.40609538343781,
+      "learning_rate": 1.979646458866175e-06,
+      "loss": 0.7064,
+      "step": 3098
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.934518010256724,
+      "learning_rate": 1.979627062120656e-06,
+      "loss": 0.7018,
+      "step": 3099
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.463181670159533,
+      "learning_rate": 1.9796076562321907e-06,
+      "loss": 0.6992,
+      "step": 3100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8634566468414097,
+      "learning_rate": 1.979588241200961e-06,
+      "loss": 0.6797,
+      "step": 3101
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6623828959711031,
+      "learning_rate": 1.9795688170271473e-06,
+      "loss": 0.6855,
+      "step": 3102
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.280319351030144,
+      "learning_rate": 1.9795493837109312e-06,
+      "loss": 0.6908,
+      "step": 3103
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3468649334265073,
+      "learning_rate": 1.9795299412524945e-06,
+      "loss": 0.6784,
+      "step": 3104
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.595235206046532,
+      "learning_rate": 1.9795104896520177e-06,
+      "loss": 0.6934,
+      "step": 3105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7053996227419984,
+      "learning_rate": 1.9794910289096832e-06,
+      "loss": 0.6855,
+      "step": 3106
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.192819364515391,
+      "learning_rate": 1.979471559025672e-06,
+      "loss": 0.6855,
+      "step": 3107
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.844064114213031,
+      "learning_rate": 1.9794520800001665e-06,
+      "loss": 0.694,
+      "step": 3108
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2892493775719498,
+      "learning_rate": 1.979432591833348e-06,
+      "loss": 0.6771,
+      "step": 3109
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.890348921410966,
+      "learning_rate": 1.9794130945253988e-06,
+      "loss": 0.7044,
+      "step": 3110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8090223341101273,
+      "learning_rate": 1.9793935880765e-06,
+      "loss": 0.6895,
+      "step": 3111
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.029605692229435,
+      "learning_rate": 1.9793740724868343e-06,
+      "loss": 0.6888,
+      "step": 3112
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8095400605675465,
+      "learning_rate": 1.979354547756584e-06,
+      "loss": 0.6842,
+      "step": 3113
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2418810900796013,
+      "learning_rate": 1.979335013885931e-06,
+      "loss": 0.6771,
+      "step": 3114
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6418164356880279,
+      "learning_rate": 1.9793154708750577e-06,
+      "loss": 0.6875,
+      "step": 3115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.0467008757835796,
+      "learning_rate": 1.9792959187241463e-06,
+      "loss": 0.7005,
+      "step": 3116
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.095986423668685,
+      "learning_rate": 1.97927635743338e-06,
+      "loss": 0.6836,
+      "step": 3117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.123545668893634,
+      "learning_rate": 1.9792567870029404e-06,
+      "loss": 0.6849,
+      "step": 3118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4769749478176855,
+      "learning_rate": 1.979237207433011e-06,
+      "loss": 0.7005,
+      "step": 3119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4759592786363758,
+      "learning_rate": 1.9792176187237738e-06,
+      "loss": 0.6849,
+      "step": 3120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5444092936943171,
+      "learning_rate": 1.9791980208754124e-06,
+      "loss": 0.694,
+      "step": 3121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8001784854042797,
+      "learning_rate": 1.9791784138881085e-06,
+      "loss": 0.6888,
+      "step": 3122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1077847962964187,
+      "learning_rate": 1.9791587977620467e-06,
+      "loss": 0.6758,
+      "step": 3123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.120421271526948,
+      "learning_rate": 1.9791391724974087e-06,
+      "loss": 0.6914,
+      "step": 3124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5663823177729,
+      "learning_rate": 1.9791195380943783e-06,
+      "loss": 0.6849,
+      "step": 3125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5503785283606837,
+      "learning_rate": 1.979099894553139e-06,
+      "loss": 0.6836,
+      "step": 3126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.952104790536909,
+      "learning_rate": 1.9790802418738733e-06,
+      "loss": 0.6862,
+      "step": 3127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.373124378082684,
+      "learning_rate": 1.9790605800567656e-06,
+      "loss": 0.6803,
+      "step": 3128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.513180872535552,
+      "learning_rate": 1.9790409091019985e-06,
+      "loss": 0.6979,
+      "step": 3129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.309178305240091,
+      "learning_rate": 1.9790212290097564e-06,
+      "loss": 0.6823,
+      "step": 3130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.900546845770343,
+      "learning_rate": 1.9790015397802226e-06,
+      "loss": 0.6855,
+      "step": 3131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1943681065000575,
+      "learning_rate": 1.9789818414135805e-06,
+      "loss": 0.6842,
+      "step": 3132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7446065636686301,
+      "learning_rate": 1.9789621339100143e-06,
+      "loss": 0.7044,
+      "step": 3133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.885902753936155,
+      "learning_rate": 1.978942417269708e-06,
+      "loss": 0.6882,
+      "step": 3134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7895814502682835,
+      "learning_rate": 1.9789226914928458e-06,
+      "loss": 0.6836,
+      "step": 3135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.931185466702865,
+      "learning_rate": 1.978902956579611e-06,
+      "loss": 0.681,
+      "step": 3136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.419213592076695,
+      "learning_rate": 1.978883212530189e-06,
+      "loss": 0.6771,
+      "step": 3137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.848586021560974,
+      "learning_rate": 1.978863459344763e-06,
+      "loss": 0.694,
+      "step": 3138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6176643055528444,
+      "learning_rate": 1.978843697023518e-06,
+      "loss": 0.6641,
+      "step": 3139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.816241456110386,
+      "learning_rate": 1.9788239255666382e-06,
+      "loss": 0.7031,
+      "step": 3140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1503289482804866,
+      "learning_rate": 1.978804144974308e-06,
+      "loss": 0.6914,
+      "step": 3141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4837945471749851,
+      "learning_rate": 1.978784355246712e-06,
+      "loss": 0.6849,
+      "step": 3142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.290993868966678,
+      "learning_rate": 1.978764556384035e-06,
+      "loss": 0.7051,
+      "step": 3143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.957995389652617,
+      "learning_rate": 1.9787447483864625e-06,
+      "loss": 0.6908,
+      "step": 3144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.571557587348527,
+      "learning_rate": 1.978724931254178e-06,
+      "loss": 0.6901,
+      "step": 3145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3630641854819934,
+      "learning_rate": 1.9787051049873678e-06,
+      "loss": 0.6823,
+      "step": 3146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.782754133877477,
+      "learning_rate": 1.978685269586216e-06,
+      "loss": 0.6901,
+      "step": 3147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.790946497331579,
+      "learning_rate": 1.9786654250509076e-06,
+      "loss": 0.6693,
+      "step": 3148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.2907092523086945,
+      "learning_rate": 1.978645571381629e-06,
+      "loss": 0.6797,
+      "step": 3149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.312085215335766,
+      "learning_rate": 1.9786257085785642e-06,
+      "loss": 0.6921,
+      "step": 3150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3546338580277455,
+      "learning_rate": 1.9786058366418993e-06,
+      "loss": 0.6836,
+      "step": 3151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7824321650539378,
+      "learning_rate": 1.9785859555718192e-06,
+      "loss": 0.6771,
+      "step": 3152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.738421438018369,
+      "learning_rate": 1.97856606536851e-06,
+      "loss": 0.6947,
+      "step": 3153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6922949229824384,
+      "learning_rate": 1.9785461660321578e-06,
+      "loss": 0.679,
+      "step": 3154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1054509615341181,
+      "learning_rate": 1.9785262575629475e-06,
+      "loss": 0.6914,
+      "step": 3155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4513629334856608,
+      "learning_rate": 1.9785063399610648e-06,
+      "loss": 0.6849,
+      "step": 3156
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.839868274767603,
+      "learning_rate": 1.9784864132266956e-06,
+      "loss": 0.6882,
+      "step": 3157
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.110496693433575,
+      "learning_rate": 1.9784664773600267e-06,
+      "loss": 0.6836,
+      "step": 3158
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.125679102509735,
+      "learning_rate": 1.9784465323612433e-06,
+      "loss": 0.6908,
+      "step": 3159
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9634466873050622,
+      "learning_rate": 1.978426578230532e-06,
+      "loss": 0.6816,
+      "step": 3160
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.46791446551438,
+      "learning_rate": 1.978406614968079e-06,
+      "loss": 0.6966,
+      "step": 3161
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.211263389798821,
+      "learning_rate": 1.9783866425740707e-06,
+      "loss": 0.668,
+      "step": 3162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6875625531498124,
+      "learning_rate": 1.978366661048693e-06,
+      "loss": 0.681,
+      "step": 3163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.514977396087717,
+      "learning_rate": 1.978346670392133e-06,
+      "loss": 0.6888,
+      "step": 3164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1603785741067136,
+      "learning_rate": 1.9783266706045767e-06,
+      "loss": 0.6803,
+      "step": 3165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.98944050582575,
+      "learning_rate": 1.9783066616862116e-06,
+      "loss": 0.6862,
+      "step": 3166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.554000629230342,
+      "learning_rate": 1.978286643637223e-06,
+      "loss": 0.696,
+      "step": 3167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3892983198873132,
+      "learning_rate": 1.9782666164577995e-06,
+      "loss": 0.7161,
+      "step": 3168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1537967071489592,
+      "learning_rate": 1.9782465801481265e-06,
+      "loss": 0.6803,
+      "step": 3169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3745157566439712,
+      "learning_rate": 1.9782265347083926e-06,
+      "loss": 0.6888,
+      "step": 3170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.541854149005662,
+      "learning_rate": 1.978206480138783e-06,
+      "loss": 0.6842,
+      "step": 3171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.225189474643609,
+      "learning_rate": 1.9781864164394863e-06,
+      "loss": 0.6927,
+      "step": 3172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1019482922089705,
+      "learning_rate": 1.978166343610689e-06,
+      "loss": 0.6758,
+      "step": 3173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.23974543670924,
+      "learning_rate": 1.978146261652579e-06,
+      "loss": 0.6784,
+      "step": 3174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6310076828406525,
+      "learning_rate": 1.9781261705653433e-06,
+      "loss": 0.6888,
+      "step": 3175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.856283569827794,
+      "learning_rate": 1.9781060703491694e-06,
+      "loss": 0.7018,
+      "step": 3176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.1240757057175668,
+      "learning_rate": 1.978085961004245e-06,
+      "loss": 0.6888,
+      "step": 3177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4943456458351791,
+      "learning_rate": 1.978065842530758e-06,
+      "loss": 0.6875,
+      "step": 3178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6205423297282175,
+      "learning_rate": 1.978045714928896e-06,
+      "loss": 0.6888,
+      "step": 3179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4113996770183737,
+      "learning_rate": 1.978025578198847e-06,
+      "loss": 0.6901,
+      "step": 3180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8342517494952572,
+      "learning_rate": 1.978005432340799e-06,
+      "loss": 0.6745,
+      "step": 3181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.499637199839961,
+      "learning_rate": 1.9779852773549392e-06,
+      "loss": 0.696,
+      "step": 3182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0795354837540567,
+      "learning_rate": 1.9779651132414564e-06,
+      "loss": 0.6875,
+      "step": 3183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4728151444845357,
+      "learning_rate": 1.977944940000539e-06,
+      "loss": 0.6855,
+      "step": 3184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.055917290179741,
+      "learning_rate": 1.977924757632375e-06,
+      "loss": 0.6842,
+      "step": 3185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.873640771382161,
+      "learning_rate": 1.9779045661371524e-06,
+      "loss": 0.6953,
+      "step": 3186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.701080894688713,
+      "learning_rate": 1.97788436551506e-06,
+      "loss": 0.6855,
+      "step": 3187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.090749271043163,
+      "learning_rate": 1.977864155766287e-06,
+      "loss": 0.6745,
+      "step": 3188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8937717390396545,
+      "learning_rate": 1.9778439368910207e-06,
+      "loss": 0.6706,
+      "step": 3189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4791623139427181,
+      "learning_rate": 1.9778237088894506e-06,
+      "loss": 0.6771,
+      "step": 3190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5289294016794748,
+      "learning_rate": 1.9778034717617655e-06,
+      "loss": 0.6914,
+      "step": 3191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0485687699946835,
+      "learning_rate": 1.977783225508154e-06,
+      "loss": 0.6784,
+      "step": 3192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.0011782797541044,
+      "learning_rate": 1.977762970128805e-06,
+      "loss": 0.6895,
+      "step": 3193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6646457633774874,
+      "learning_rate": 1.9777427056239083e-06,
+      "loss": 0.6842,
+      "step": 3194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3299521923024087,
+      "learning_rate": 1.9777224319936517e-06,
+      "loss": 0.6771,
+      "step": 3195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.784086674291647,
+      "learning_rate": 1.9777021492382257e-06,
+      "loss": 0.6823,
+      "step": 3196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8683339436940476,
+      "learning_rate": 1.9776818573578188e-06,
+      "loss": 0.7109,
+      "step": 3197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.284697093993913,
+      "learning_rate": 1.977661556352621e-06,
+      "loss": 0.6862,
+      "step": 3198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.715911547702389,
+      "learning_rate": 1.977641246222821e-06,
+      "loss": 0.6921,
+      "step": 3199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9846319673194248,
+      "learning_rate": 1.977620926968609e-06,
+      "loss": 0.6732,
+      "step": 3200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.237267171582451,
+      "learning_rate": 1.977600598590175e-06,
+      "loss": 0.696,
+      "step": 3201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.305396875139353,
+      "learning_rate": 1.9775802610877075e-06,
+      "loss": 0.6829,
+      "step": 3202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2812228240558374,
+      "learning_rate": 1.977559914461397e-06,
+      "loss": 0.6888,
+      "step": 3203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9567158316038523,
+      "learning_rate": 1.977539558711434e-06,
+      "loss": 0.6947,
+      "step": 3204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.574192649093933,
+      "learning_rate": 1.9775191938380067e-06,
+      "loss": 0.6882,
+      "step": 3205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5283333885822636,
+      "learning_rate": 1.977498819841307e-06,
+      "loss": 0.6829,
+      "step": 3206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.751969002772794,
+      "learning_rate": 1.9774784367215245e-06,
+      "loss": 0.6914,
+      "step": 3207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.869140748538175,
+      "learning_rate": 1.9774580444788493e-06,
+      "loss": 0.6934,
+      "step": 3208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.741184888896479,
+      "learning_rate": 1.9774376431134713e-06,
+      "loss": 0.6947,
+      "step": 3209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.564794990079831,
+      "learning_rate": 1.977417232625582e-06,
+      "loss": 0.7096,
+      "step": 3210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.434101042827853,
+      "learning_rate": 1.9773968130153708e-06,
+      "loss": 0.6986,
+      "step": 3211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4724457007738105,
+      "learning_rate": 1.9773763842830286e-06,
+      "loss": 0.694,
+      "step": 3212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.701568572469744,
+      "learning_rate": 1.977355946428746e-06,
+      "loss": 0.6927,
+      "step": 3213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.121051720184709,
+      "learning_rate": 1.9773354994527144e-06,
+      "loss": 0.681,
+      "step": 3214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.95548868412497,
+      "learning_rate": 1.977315043355124e-06,
+      "loss": 0.6882,
+      "step": 3215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.713676629895356,
+      "learning_rate": 1.9772945781361658e-06,
+      "loss": 0.6979,
+      "step": 3216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9124886252456418,
+      "learning_rate": 1.977274103796031e-06,
+      "loss": 0.6836,
+      "step": 3217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.489435696740453,
+      "learning_rate": 1.9772536203349105e-06,
+      "loss": 0.6849,
+      "step": 3218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4752922358659086,
+      "learning_rate": 1.9772331277529957e-06,
+      "loss": 0.6816,
+      "step": 3219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.3456563486886495,
+      "learning_rate": 1.9772126260504774e-06,
+      "loss": 0.6868,
+      "step": 3220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.070441279154482,
+      "learning_rate": 1.9771921152275475e-06,
+      "loss": 0.6784,
+      "step": 3221
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8884287837042225,
+      "learning_rate": 1.9771715952843967e-06,
+      "loss": 0.6771,
+      "step": 3222
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3922062132908177,
+      "learning_rate": 1.9771510662212175e-06,
+      "loss": 0.6719,
+      "step": 3223
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.830760557042786,
+      "learning_rate": 1.977130528038201e-06,
+      "loss": 0.679,
+      "step": 3224
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.619705467286738,
+      "learning_rate": 1.9771099807355383e-06,
+      "loss": 0.6953,
+      "step": 3225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3140175122388698,
+      "learning_rate": 1.9770894243134224e-06,
+      "loss": 0.6823,
+      "step": 3226
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6185197424751987,
+      "learning_rate": 1.9770688587720443e-06,
+      "loss": 0.6895,
+      "step": 3227
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.439482805233935,
+      "learning_rate": 1.977048284111596e-06,
+      "loss": 0.6882,
+      "step": 3228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.947458979317542,
+      "learning_rate": 1.9770277003322694e-06,
+      "loss": 0.6777,
+      "step": 3229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.1090149011045956,
+      "learning_rate": 1.977007107434257e-06,
+      "loss": 0.6862,
+      "step": 3230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.450512685268132,
+      "learning_rate": 1.976986505417751e-06,
+      "loss": 0.7018,
+      "step": 3231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.029601460282169,
+      "learning_rate": 1.9769658942829438e-06,
+      "loss": 0.6901,
+      "step": 3232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.410877023442343,
+      "learning_rate": 1.976945274030027e-06,
+      "loss": 0.6797,
+      "step": 3233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.5418604978458195,
+      "learning_rate": 1.976924644659194e-06,
+      "loss": 0.6797,
+      "step": 3234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7875968193007177,
+      "learning_rate": 1.976904006170637e-06,
+      "loss": 0.681,
+      "step": 3235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6777601504351839,
+      "learning_rate": 1.976883358564548e-06,
+      "loss": 0.6868,
+      "step": 3236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.591694628528006,
+      "learning_rate": 1.9768627018411204e-06,
+      "loss": 0.7168,
+      "step": 3237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.917307346225396,
+      "learning_rate": 1.976842036000547e-06,
+      "loss": 0.668,
+      "step": 3238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.6262652171877425,
+      "learning_rate": 1.9768213610430205e-06,
+      "loss": 0.6823,
+      "step": 3239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.2222881570338,
+      "learning_rate": 1.9768006769687336e-06,
+      "loss": 0.7038,
+      "step": 3240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.815212508878316,
+      "learning_rate": 1.97677998377788e-06,
+      "loss": 0.6895,
+      "step": 3241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7839076563718684,
+      "learning_rate": 1.976759281470652e-06,
+      "loss": 0.6921,
+      "step": 3242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.806298260905028,
+      "learning_rate": 1.9767385700472433e-06,
+      "loss": 0.6979,
+      "step": 3243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.829183429487906,
+      "learning_rate": 1.9767178495078474e-06,
+      "loss": 0.6868,
+      "step": 3244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.026530236427871,
+      "learning_rate": 1.976697119852657e-06,
+      "loss": 0.6862,
+      "step": 3245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.200221609644321,
+      "learning_rate": 1.976676381081867e-06,
+      "loss": 0.6888,
+      "step": 3246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1458053295661967,
+      "learning_rate": 1.976655633195669e-06,
+      "loss": 0.6758,
+      "step": 3247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.205258485067019,
+      "learning_rate": 1.976634876194258e-06,
+      "loss": 0.707,
+      "step": 3248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.091915619212243,
+      "learning_rate": 1.9766141100778267e-06,
+      "loss": 0.6895,
+      "step": 3249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.31549808298204,
+      "learning_rate": 1.97659333484657e-06,
+      "loss": 0.6927,
+      "step": 3250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.124659388040409,
+      "learning_rate": 1.9765725505006816e-06,
+      "loss": 0.6934,
+      "step": 3251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.797567308433581,
+      "learning_rate": 1.976551757040355e-06,
+      "loss": 0.6973,
+      "step": 3252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9509824216691043,
+      "learning_rate": 1.9765309544657846e-06,
+      "loss": 0.6914,
+      "step": 3253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.99546773219133,
+      "learning_rate": 1.9765101427771644e-06,
+      "loss": 0.679,
+      "step": 3254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.962684914037917,
+      "learning_rate": 1.9764893219746886e-06,
+      "loss": 0.6712,
+      "step": 3255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.029995132521225,
+      "learning_rate": 1.9764684920585517e-06,
+      "loss": 0.6816,
+      "step": 3256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1625626853613817,
+      "learning_rate": 1.9764476530289476e-06,
+      "loss": 0.6654,
+      "step": 3257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.52859377110642,
+      "learning_rate": 1.9764268048860715e-06,
+      "loss": 0.7005,
+      "step": 3258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.650758810665243,
+      "learning_rate": 1.976405947630118e-06,
+      "loss": 0.6829,
+      "step": 3259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5638487598684265,
+      "learning_rate": 1.976385081261281e-06,
+      "loss": 0.6784,
+      "step": 3260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.382418692174264,
+      "learning_rate": 1.976364205779756e-06,
+      "loss": 0.7038,
+      "step": 3261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6718842473538005,
+      "learning_rate": 1.9763433211857373e-06,
+      "loss": 0.6979,
+      "step": 3262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.545840614315828,
+      "learning_rate": 1.97632242747942e-06,
+      "loss": 0.6999,
+      "step": 3263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.855623247258206,
+      "learning_rate": 1.976301524660999e-06,
+      "loss": 0.6797,
+      "step": 3264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.145725021894822,
+      "learning_rate": 1.97628061273067e-06,
+      "loss": 0.6895,
+      "step": 3265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.037232270813643,
+      "learning_rate": 1.9762596916886272e-06,
+      "loss": 0.6908,
+      "step": 3266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8814046269320985,
+      "learning_rate": 1.976238761535067e-06,
+      "loss": 0.6849,
+      "step": 3267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.669737067370233,
+      "learning_rate": 1.9762178222701835e-06,
+      "loss": 0.6953,
+      "step": 3268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.520963169935468,
+      "learning_rate": 1.976196873894173e-06,
+      "loss": 0.6882,
+      "step": 3269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7703623664838655,
+      "learning_rate": 1.976175916407231e-06,
+      "loss": 0.6855,
+      "step": 3270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7264869124374673,
+      "learning_rate": 1.9761549498095524e-06,
+      "loss": 0.7005,
+      "step": 3271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.176167628930603,
+      "learning_rate": 1.9761339741013337e-06,
+      "loss": 0.6608,
+      "step": 3272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7902274772194886,
+      "learning_rate": 1.97611298928277e-06,
+      "loss": 0.6855,
+      "step": 3273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3939808733735704,
+      "learning_rate": 1.9760919953540575e-06,
+      "loss": 0.6842,
+      "step": 3274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.263406925910313,
+      "learning_rate": 1.976070992315392e-06,
+      "loss": 0.6829,
+      "step": 3275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.227676742273933,
+      "learning_rate": 1.97604998016697e-06,
+      "loss": 0.6934,
+      "step": 3276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.083990229748188,
+      "learning_rate": 1.9760289589089867e-06,
+      "loss": 0.6771,
+      "step": 3277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1855306305078672,
+      "learning_rate": 1.9760079285416392e-06,
+      "loss": 0.679,
+      "step": 3278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.109721992684601,
+      "learning_rate": 1.975986889065123e-06,
+      "loss": 0.6751,
+      "step": 3279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7736510804056428,
+      "learning_rate": 1.975965840479635e-06,
+      "loss": 0.6992,
+      "step": 3280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.1705813823116469,
+      "learning_rate": 1.9759447827853715e-06,
+      "loss": 0.6999,
+      "step": 3281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6512400038837048,
+      "learning_rate": 1.9759237159825296e-06,
+      "loss": 0.6868,
+      "step": 3282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9779481182638695,
+      "learning_rate": 1.9759026400713048e-06,
+      "loss": 0.6855,
+      "step": 3283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.726258935864044,
+      "learning_rate": 1.9758815550518944e-06,
+      "loss": 0.6888,
+      "step": 3284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.97527462791273,
+      "learning_rate": 1.975860460924495e-06,
+      "loss": 0.6986,
+      "step": 3285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.295662860823034,
+      "learning_rate": 1.975839357689304e-06,
+      "loss": 0.6777,
+      "step": 3286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.876417632005305,
+      "learning_rate": 1.9758182453465176e-06,
+      "loss": 0.694,
+      "step": 3287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.364861267238436,
+      "learning_rate": 1.9757971238963337e-06,
+      "loss": 0.6764,
+      "step": 3288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.75848089779755,
+      "learning_rate": 1.9757759933389485e-06,
+      "loss": 0.6901,
+      "step": 3289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6878947564477006,
+      "learning_rate": 1.9757548536745595e-06,
+      "loss": 0.6908,
+      "step": 3290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.155391841072309,
+      "learning_rate": 1.9757337049033647e-06,
+      "loss": 0.6901,
+      "step": 3291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.629703296849899,
+      "learning_rate": 1.9757125470255608e-06,
+      "loss": 0.6764,
+      "step": 3292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.888671761701647,
+      "learning_rate": 1.9756913800413454e-06,
+      "loss": 0.6797,
+      "step": 3293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.240432642882757,
+      "learning_rate": 1.975670203950916e-06,
+      "loss": 0.6947,
+      "step": 3294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.894565450735684,
+      "learning_rate": 1.97564901875447e-06,
+      "loss": 0.6823,
+      "step": 3295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.266176515557385,
+      "learning_rate": 1.975627824452206e-06,
+      "loss": 0.6868,
+      "step": 3296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.177237161868175,
+      "learning_rate": 1.9756066210443205e-06,
+      "loss": 0.6764,
+      "step": 3297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5380795626753176,
+      "learning_rate": 1.9755854085310128e-06,
+      "loss": 0.6921,
+      "step": 3298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.639159906494914,
+      "learning_rate": 1.97556418691248e-06,
+      "loss": 0.6914,
+      "step": 3299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.674913730153612,
+      "learning_rate": 1.9755429561889205e-06,
+      "loss": 0.6758,
+      "step": 3300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.258477485617365,
+      "learning_rate": 1.9755217163605323e-06,
+      "loss": 0.7057,
+      "step": 3301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8488720246807766,
+      "learning_rate": 1.9755004674275133e-06,
+      "loss": 0.7018,
+      "step": 3302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0653352010691512,
+      "learning_rate": 1.9754792093900623e-06,
+      "loss": 0.6803,
+      "step": 3303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8957337320031256,
+      "learning_rate": 1.9754579422483777e-06,
+      "loss": 0.6895,
+      "step": 3304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6677429485528539,
+      "learning_rate": 1.975436666002658e-06,
+      "loss": 0.6634,
+      "step": 3305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.053992740815659,
+      "learning_rate": 1.9754153806531014e-06,
+      "loss": 0.6784,
+      "step": 3306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.897608083293544,
+      "learning_rate": 1.9753940861999067e-06,
+      "loss": 0.6901,
+      "step": 3307
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.950826675011799,
+      "learning_rate": 1.9753727826432727e-06,
+      "loss": 0.6829,
+      "step": 3308
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.596036642857945,
+      "learning_rate": 1.9753514699833987e-06,
+      "loss": 0.6914,
+      "step": 3309
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6179665552469439,
+      "learning_rate": 1.9753301482204827e-06,
+      "loss": 0.6895,
+      "step": 3310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.411225204944122,
+      "learning_rate": 1.9753088173547243e-06,
+      "loss": 0.6862,
+      "step": 3311
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.387351195051,
+      "learning_rate": 1.975287477386323e-06,
+      "loss": 0.6764,
+      "step": 3312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.157387857010507,
+      "learning_rate": 1.975266128315477e-06,
+      "loss": 0.6673,
+      "step": 3313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6578078581880082,
+      "learning_rate": 1.975244770142386e-06,
+      "loss": 0.6908,
+      "step": 3314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8998208076332337,
+      "learning_rate": 1.975223402867249e-06,
+      "loss": 0.6842,
+      "step": 3315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.548777635914467,
+      "learning_rate": 1.975202026490266e-06,
+      "loss": 0.6836,
+      "step": 3316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.325521154278511,
+      "learning_rate": 1.9751806410116363e-06,
+      "loss": 0.6875,
+      "step": 3317
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.512749568100071,
+      "learning_rate": 1.975159246431559e-06,
+      "loss": 0.6908,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6883724593412025,
+      "learning_rate": 1.975137842750235e-06,
+      "loss": 0.6706,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8655522182117052,
+      "learning_rate": 1.9751164299678628e-06,
+      "loss": 0.6673,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.53756834925391,
+      "learning_rate": 1.9750950080846428e-06,
+      "loss": 0.6927,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7862594726318175,
+      "learning_rate": 1.9750735771007746e-06,
+      "loss": 0.6777,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8566338713907861,
+      "learning_rate": 1.975052137016459e-06,
+      "loss": 0.6901,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.567276370376521,
+      "learning_rate": 1.9750306878318954e-06,
+      "loss": 0.666,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6966696367296312,
+      "learning_rate": 1.9750092295472837e-06,
+      "loss": 0.6725,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.195573655291586,
+      "learning_rate": 1.974987762162825e-06,
+      "loss": 0.6901,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.623480590120584,
+      "learning_rate": 1.9749662856787196e-06,
+      "loss": 0.6999,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.808847223700556,
+      "learning_rate": 1.9749448000951674e-06,
+      "loss": 0.7005,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.767234438183642,
+      "learning_rate": 1.9749233054123688e-06,
+      "loss": 0.6725,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.105522415312454,
+      "learning_rate": 1.974901801630525e-06,
+      "loss": 0.7064,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.571682383524815,
+      "learning_rate": 1.9748802887498368e-06,
+      "loss": 0.7272,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3803970632002722,
+      "learning_rate": 1.974858766770504e-06,
+      "loss": 0.6842,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2087318687352293,
+      "learning_rate": 1.9748372356927287e-06,
+      "loss": 0.6901,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0998794070113997,
+      "learning_rate": 1.974815695516711e-06,
+      "loss": 0.6862,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6727383526237685,
+      "learning_rate": 1.9747941462426524e-06,
+      "loss": 0.6751,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49524900299519936,
+      "learning_rate": 1.9747725878707536e-06,
+      "loss": 0.6875,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9258425313195144,
+      "learning_rate": 1.974751020401216e-06,
+      "loss": 0.6829,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.155417326644257,
+      "learning_rate": 1.974729443834241e-06,
+      "loss": 0.6816,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.620606880131691,
+      "learning_rate": 1.9747078581700303e-06,
+      "loss": 0.681,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.909173677120375,
+      "learning_rate": 1.9746862634087846e-06,
+      "loss": 0.6901,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.527612516959471,
+      "learning_rate": 1.974664659550706e-06,
+      "loss": 0.6823,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.174780775016768,
+      "learning_rate": 1.9746430465959955e-06,
+      "loss": 0.6868,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6190434973703973,
+      "learning_rate": 1.9746214245448555e-06,
+      "loss": 0.6895,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5660529815352286,
+      "learning_rate": 1.9745997933974877e-06,
+      "loss": 0.6882,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4855282996370439,
+      "learning_rate": 1.9745781531540935e-06,
+      "loss": 0.679,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.67928680699654,
+      "learning_rate": 1.9745565038148753e-06,
+      "loss": 0.6745,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4977537921944974,
+      "learning_rate": 1.9745348453800353e-06,
+      "loss": 0.6875,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.903304482088259,
+      "learning_rate": 1.9745131778497754e-06,
+      "loss": 0.696,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.968913842109329,
+      "learning_rate": 1.9744915012242976e-06,
+      "loss": 0.6882,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6851488706112824,
+      "learning_rate": 1.9744698155038044e-06,
+      "loss": 0.6667,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.365047811938867,
+      "learning_rate": 1.9744481206884986e-06,
+      "loss": 0.6816,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0738252816291767,
+      "learning_rate": 1.974426416778582e-06,
+      "loss": 0.7005,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6735569613085954,
+      "learning_rate": 1.974404703774258e-06,
+      "loss": 0.679,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2708622181928604,
+      "learning_rate": 1.974382981675728e-06,
+      "loss": 0.6934,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4703755619621663,
+      "learning_rate": 1.974361250483196e-06,
+      "loss": 0.6875,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.370989447158303,
+      "learning_rate": 1.974339510196864e-06,
+      "loss": 0.6751,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6357387647437704,
+      "learning_rate": 1.974317760816935e-06,
+      "loss": 0.6764,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.654476052073029,
+      "learning_rate": 1.9742960023436125e-06,
+      "loss": 0.7064,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8265202147657396,
+      "learning_rate": 1.9742742347770987e-06,
+      "loss": 0.6921,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7733901390983804,
+      "learning_rate": 1.974252458117598e-06,
+      "loss": 0.7129,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6108400596358946,
+      "learning_rate": 1.9742306723653126e-06,
+      "loss": 0.6758,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0373548157480106,
+      "learning_rate": 1.9742088775204463e-06,
+      "loss": 0.6777,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.603030146372593,
+      "learning_rate": 1.9741870735832026e-06,
+      "loss": 0.7122,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7803452407166143,
+      "learning_rate": 1.9741652605537845e-06,
+      "loss": 0.6634,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.260100907694952,
+      "learning_rate": 1.974143438432396e-06,
+      "loss": 0.7018,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.013374249200411,
+      "learning_rate": 1.9741216072192405e-06,
+      "loss": 0.6921,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3688172864286163,
+      "learning_rate": 1.9740997669145216e-06,
+      "loss": 0.6751,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4821060446714345,
+      "learning_rate": 1.974077917518444e-06,
+      "loss": 0.6777,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3904326133410947,
+      "learning_rate": 1.974056059031211e-06,
+      "loss": 0.7031,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40134713973787095,
+      "learning_rate": 1.974034191453027e-06,
+      "loss": 0.6921,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.509672299770044,
+      "learning_rate": 1.9740123147840946e-06,
+      "loss": 0.709,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4500144497196439,
+      "learning_rate": 1.9739904290246203e-06,
+      "loss": 0.6973,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9298564443761266,
+      "learning_rate": 1.9739685341748065e-06,
+      "loss": 0.6868,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.7345621254960975,
+      "learning_rate": 1.9739466302348583e-06,
+      "loss": 0.6888,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6911850873565273,
+      "learning_rate": 1.9739247172049803e-06,
+      "loss": 0.6725,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.416030717942761,
+      "learning_rate": 1.9739027950853766e-06,
+      "loss": 0.6908,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.336010569297612,
+      "learning_rate": 1.973880863876252e-06,
+      "loss": 0.6862,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.063957410018514,
+      "learning_rate": 1.973858923577811e-06,
+      "loss": 0.6868,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4177671234467364,
+      "learning_rate": 1.973836974190259e-06,
+      "loss": 0.6934,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8060641778416959,
+      "learning_rate": 1.9738150157138e-06,
+      "loss": 0.6816,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.962816767343177,
+      "learning_rate": 1.9737930481486397e-06,
+      "loss": 0.679,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4004184491637006,
+      "learning_rate": 1.9737710714949826e-06,
+      "loss": 0.6849,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8550269831682967,
+      "learning_rate": 1.9737490857530342e-06,
+      "loss": 0.6829,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7576637563188244,
+      "learning_rate": 1.973727090922999e-06,
+      "loss": 0.6771,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6499516543530457,
+      "learning_rate": 1.9737050870050832e-06,
+      "loss": 0.6712,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.01776548098043,
+      "learning_rate": 1.973683073999491e-06,
+      "loss": 0.6673,
+      "step": 3386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.210598318876258,
+      "learning_rate": 1.9736610519064296e-06,
+      "loss": 0.7012,
+      "step": 3387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.42208949176125,
+      "learning_rate": 1.973639020726103e-06,
+      "loss": 0.694,
+      "step": 3388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4191383287396413,
+      "learning_rate": 1.973616980458717e-06,
+      "loss": 0.6758,
+      "step": 3389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.340346460109826,
+      "learning_rate": 1.9735949311044777e-06,
+      "loss": 0.7038,
+      "step": 3390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.179790137653763,
+      "learning_rate": 1.9735728726635913e-06,
+      "loss": 0.7005,
+      "step": 3391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.5593125403531,
+      "learning_rate": 1.973550805136263e-06,
+      "loss": 0.6758,
+      "step": 3392
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.639485195211408,
+      "learning_rate": 1.9735287285226984e-06,
+      "loss": 0.6816,
+      "step": 3393
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.163846421354855,
+      "learning_rate": 1.9735066428231047e-06,
+      "loss": 0.6673,
+      "step": 3394
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.119726741517224,
+      "learning_rate": 1.973484548037687e-06,
+      "loss": 0.6823,
+      "step": 3395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6744236532698147,
+      "learning_rate": 1.9734624441666524e-06,
+      "loss": 0.6986,
+      "step": 3396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4433849237001746,
+      "learning_rate": 1.9734403312102064e-06,
+      "loss": 0.7077,
+      "step": 3397
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.898067417143446,
+      "learning_rate": 1.9734182091685557e-06,
+      "loss": 0.7057,
+      "step": 3398
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.276185537706004,
+      "learning_rate": 1.973396078041907e-06,
+      "loss": 0.7266,
+      "step": 3399
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5939320468611955,
+      "learning_rate": 1.973373937830466e-06,
+      "loss": 0.6921,
+      "step": 3400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1719486102175383,
+      "learning_rate": 1.9733517885344406e-06,
+      "loss": 0.6725,
+      "step": 3401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.599858942357769,
+      "learning_rate": 1.9733296301540367e-06,
+      "loss": 0.6758,
+      "step": 3402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.491464837209866,
+      "learning_rate": 1.973307462689461e-06,
+      "loss": 0.6777,
+      "step": 3403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.259275704803548,
+      "learning_rate": 1.9732852861409213e-06,
+      "loss": 0.7064,
+      "step": 3404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4881136313926255,
+      "learning_rate": 1.973263100508623e-06,
+      "loss": 0.6901,
+      "step": 3405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5672711098791419,
+      "learning_rate": 1.973240905792775e-06,
+      "loss": 0.6979,
+      "step": 3406
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7358751663539989,
+      "learning_rate": 1.973218701993583e-06,
+      "loss": 0.6615,
+      "step": 3407
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.176678602909893,
+      "learning_rate": 1.9731964891112555e-06,
+      "loss": 0.6882,
+      "step": 3408
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3711094134759825,
+      "learning_rate": 1.9731742671459984e-06,
+      "loss": 0.6875,
+      "step": 3409
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.975793610190349,
+      "learning_rate": 1.9731520360980204e-06,
+      "loss": 0.6849,
+      "step": 3410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5650723558685778,
+      "learning_rate": 1.973129795967528e-06,
+      "loss": 0.6699,
+      "step": 3411
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6757287760872324,
+      "learning_rate": 1.973107546754729e-06,
+      "loss": 0.6927,
+      "step": 3412
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.765419865062475,
+      "learning_rate": 1.973085288459831e-06,
+      "loss": 0.6855,
+      "step": 3413
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0286961882341252,
+      "learning_rate": 1.973063021083043e-06,
+      "loss": 0.6777,
+      "step": 3414
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7802510067681516,
+      "learning_rate": 1.9730407446245707e-06,
+      "loss": 0.6901,
+      "step": 3415
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.461290771343401,
+      "learning_rate": 1.973018459084624e-06,
+      "loss": 0.6979,
+      "step": 3416
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.627316982960048,
+      "learning_rate": 1.9729961644634092e-06,
+      "loss": 0.6901,
+      "step": 3417
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4872305703821873,
+      "learning_rate": 1.9729738607611354e-06,
+      "loss": 0.6823,
+      "step": 3418
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.432848250033093,
+      "learning_rate": 1.9729515479780106e-06,
+      "loss": 0.6973,
+      "step": 3419
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.673441857016266,
+      "learning_rate": 1.972929226114243e-06,
+      "loss": 0.6751,
+      "step": 3420
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.6723192945413174,
+      "learning_rate": 1.9729068951700405e-06,
+      "loss": 0.7005,
+      "step": 3421
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.84766082380395,
+      "learning_rate": 1.9728845551456123e-06,
+      "loss": 0.6745,
+      "step": 3422
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6209749611232093,
+      "learning_rate": 1.9728622060411663e-06,
+      "loss": 0.6693,
+      "step": 3423
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2475647034978725,
+      "learning_rate": 1.9728398478569113e-06,
+      "loss": 0.6797,
+      "step": 3424
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.373444561172873,
+      "learning_rate": 1.972817480593056e-06,
+      "loss": 0.6784,
+      "step": 3425
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.059216054167099,
+      "learning_rate": 1.972795104249809e-06,
+      "loss": 0.6797,
+      "step": 3426
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9047732841123061,
+      "learning_rate": 1.9727727188273796e-06,
+      "loss": 0.6745,
+      "step": 3427
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5668799930909005,
+      "learning_rate": 1.972750324325976e-06,
+      "loss": 0.7031,
+      "step": 3428
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7197317793902867,
+      "learning_rate": 1.972727920745808e-06,
+      "loss": 0.6621,
+      "step": 3429
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.680379040128478,
+      "learning_rate": 1.972705508087084e-06,
+      "loss": 0.7005,
+      "step": 3430
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.412655686590141,
+      "learning_rate": 1.972683086350014e-06,
+      "loss": 0.6706,
+      "step": 3431
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.055456646537599,
+      "learning_rate": 1.9726606555348063e-06,
+      "loss": 0.6667,
+      "step": 3432
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6866615264787044,
+      "learning_rate": 1.9726382156416704e-06,
+      "loss": 0.7018,
+      "step": 3433
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.931593070633431,
+      "learning_rate": 1.9726157666708164e-06,
+      "loss": 0.6999,
+      "step": 3434
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.182346886249022,
+      "learning_rate": 1.9725933086224534e-06,
+      "loss": 0.6986,
+      "step": 3435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.348344024201573,
+      "learning_rate": 1.972570841496791e-06,
+      "loss": 0.6686,
+      "step": 3436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5472192061395664,
+      "learning_rate": 1.972548365294039e-06,
+      "loss": 0.679,
+      "step": 3437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.151894273438815,
+      "learning_rate": 1.9725258800144072e-06,
+      "loss": 0.6862,
+      "step": 3438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.154757413874521,
+      "learning_rate": 1.972503385658105e-06,
+      "loss": 0.6693,
+      "step": 3439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.3339650556923335,
+      "learning_rate": 1.972480882225343e-06,
+      "loss": 0.6979,
+      "step": 3440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.352015649922371,
+      "learning_rate": 1.972458369716331e-06,
+      "loss": 0.6999,
+      "step": 3441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.683042256838884,
+      "learning_rate": 1.9724358481312782e-06,
+      "loss": 0.6595,
+      "step": 3442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.146119138132416,
+      "learning_rate": 1.9724133174703967e-06,
+      "loss": 0.6673,
+      "step": 3443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.583989486930955,
+      "learning_rate": 1.9723907777338955e-06,
+      "loss": 0.653,
+      "step": 3444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.972556836634812,
+      "learning_rate": 1.9723682289219847e-06,
+      "loss": 0.6673,
+      "step": 3445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.173897157272226,
+      "learning_rate": 1.9723456710348752e-06,
+      "loss": 0.681,
+      "step": 3446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0394465586600914,
+      "learning_rate": 1.972323104072778e-06,
+      "loss": 0.6745,
+      "step": 3447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.036557524533068,
+      "learning_rate": 1.972300528035903e-06,
+      "loss": 0.6986,
+      "step": 3448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.174873144487973,
+      "learning_rate": 1.9722779429244615e-06,
+      "loss": 0.6823,
+      "step": 3449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.15295608644208,
+      "learning_rate": 1.9722553487386635e-06,
+      "loss": 0.7044,
+      "step": 3450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2314183448091325,
+      "learning_rate": 1.972232745478721e-06,
+      "loss": 0.6927,
+      "step": 3451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.9524156776247406,
+      "learning_rate": 1.972210133144844e-06,
+      "loss": 0.6803,
+      "step": 3452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.338643200112965,
+      "learning_rate": 1.9721875117372436e-06,
+      "loss": 0.7044,
+      "step": 3453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.253169578899131,
+      "learning_rate": 1.9721648812561312e-06,
+      "loss": 0.6686,
+      "step": 3454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.570495594635108,
+      "learning_rate": 1.9721422417017185e-06,
+      "loss": 0.6875,
+      "step": 3455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.950044866577773,
+      "learning_rate": 1.972119593074216e-06,
+      "loss": 0.7194,
+      "step": 3456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.809143305014506,
+      "learning_rate": 1.972096935373835e-06,
+      "loss": 0.7012,
+      "step": 3457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1935977538267588,
+      "learning_rate": 1.9720742686007877e-06,
+      "loss": 0.6628,
+      "step": 3458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9984507083670017,
+      "learning_rate": 1.9720515927552856e-06,
+      "loss": 0.6895,
+      "step": 3459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.64769386355576,
+      "learning_rate": 1.9720289078375397e-06,
+      "loss": 0.7103,
+      "step": 3460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.96266057180559,
+      "learning_rate": 1.9720062138477617e-06,
+      "loss": 0.7031,
+      "step": 3461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4929795022950665,
+      "learning_rate": 1.9719835107861643e-06,
+      "loss": 0.6823,
+      "step": 3462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.082481383542202,
+      "learning_rate": 1.9719607986529584e-06,
+      "loss": 0.7038,
+      "step": 3463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8796348955787175,
+      "learning_rate": 1.9719380774483566e-06,
+      "loss": 0.6947,
+      "step": 3464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.237659606778615,
+      "learning_rate": 1.9719153471725707e-06,
+      "loss": 0.6849,
+      "step": 3465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.630863879332325,
+      "learning_rate": 1.971892607825813e-06,
+      "loss": 0.6641,
+      "step": 3466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.031999373858843,
+      "learning_rate": 1.9718698594082955e-06,
+      "loss": 0.6836,
+      "step": 3467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5623814697033072,
+      "learning_rate": 1.971847101920231e-06,
+      "loss": 0.6862,
+      "step": 3468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41463565902898497,
+      "learning_rate": 1.9718243353618315e-06,
+      "loss": 0.6751,
+      "step": 3469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.47233601594434366,
+      "learning_rate": 1.971801559733309e-06,
+      "loss": 0.6823,
+      "step": 3470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5154637996464804,
+      "learning_rate": 1.971778775034877e-06,
+      "loss": 0.6816,
+      "step": 3471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4323571288888672,
+      "learning_rate": 1.971755981266748e-06,
+      "loss": 0.694,
+      "step": 3472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.504332895306242,
+      "learning_rate": 1.971733178429134e-06,
+      "loss": 0.709,
+      "step": 3473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.302195392780128,
+      "learning_rate": 1.9717103665222486e-06,
+      "loss": 0.6947,
+      "step": 3474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.737602488665911,
+      "learning_rate": 1.971687545546304e-06,
+      "loss": 0.6908,
+      "step": 3475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.942699006562348,
+      "learning_rate": 1.971664715501514e-06,
+      "loss": 0.6855,
+      "step": 3476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.020594557264396,
+      "learning_rate": 1.971641876388091e-06,
+      "loss": 0.6836,
+      "step": 3477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9746718402464425,
+      "learning_rate": 1.971619028206249e-06,
+      "loss": 0.6927,
+      "step": 3478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.37789546515368,
+      "learning_rate": 1.9715961709562004e-06,
+      "loss": 0.6777,
+      "step": 3479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5697077598547327,
+      "learning_rate": 1.971573304638159e-06,
+      "loss": 0.679,
+      "step": 3480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4307041771557598,
+      "learning_rate": 1.9715504292523374e-06,
+      "loss": 0.6882,
+      "step": 3481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.651042718783252,
+      "learning_rate": 1.97152754479895e-06,
+      "loss": 0.6732,
+      "step": 3482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.950060242976906,
+      "learning_rate": 1.9715046512782102e-06,
+      "loss": 0.6842,
+      "step": 3483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.909974442637807,
+      "learning_rate": 1.971481748690332e-06,
+      "loss": 0.6797,
+      "step": 3484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.12926832986189,
+      "learning_rate": 1.971458837035528e-06,
+      "loss": 0.6895,
+      "step": 3485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.3182424592967354,
+      "learning_rate": 1.971435916314013e-06,
+      "loss": 0.6842,
+      "step": 3486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.371839102953884,
+      "learning_rate": 1.971412986526001e-06,
+      "loss": 0.6986,
+      "step": 3487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7151416399873288,
+      "learning_rate": 1.9713900476717055e-06,
+      "loss": 0.679,
+      "step": 3488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4126880930906602,
+      "learning_rate": 1.9713670997513403e-06,
+      "loss": 0.6901,
+      "step": 3489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.874493086521593,
+      "learning_rate": 1.9713441427651205e-06,
+      "loss": 0.681,
+      "step": 3490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.013089565084478,
+      "learning_rate": 1.97132117671326e-06,
+      "loss": 0.6803,
+      "step": 3491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9811021932495185,
+      "learning_rate": 1.971298201595973e-06,
+      "loss": 0.6803,
+      "step": 3492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4922514240319815,
+      "learning_rate": 1.9712752174134743e-06,
+      "loss": 0.6895,
+      "step": 3493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.853107151547969,
+      "learning_rate": 1.9712522241659774e-06,
+      "loss": 0.6803,
+      "step": 3494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.295170432881223,
+      "learning_rate": 1.971229221853698e-06,
+      "loss": 0.6803,
+      "step": 3495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7223700720786268,
+      "learning_rate": 1.9712062104768505e-06,
+      "loss": 0.6855,
+      "step": 3496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.277492689767486,
+      "learning_rate": 1.971183190035649e-06,
+      "loss": 0.6882,
+      "step": 3497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6410049247753724,
+      "learning_rate": 1.9711601605303093e-06,
+      "loss": 0.681,
+      "step": 3498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.797761189828696,
+      "learning_rate": 1.971137121961046e-06,
+      "loss": 0.6693,
+      "step": 3499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.883366757211563,
+      "learning_rate": 1.9711140743280737e-06,
+      "loss": 0.6842,
+      "step": 3500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8834279009480444,
+      "learning_rate": 1.9710910176316083e-06,
+      "loss": 0.6979,
+      "step": 3501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1539081807068086,
+      "learning_rate": 1.9710679518718645e-06,
+      "loss": 0.6855,
+      "step": 3502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1147280753314195,
+      "learning_rate": 1.9710448770490575e-06,
+      "loss": 0.6921,
+      "step": 3503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.362755197060164,
+      "learning_rate": 1.971021793163403e-06,
+      "loss": 0.6921,
+      "step": 3504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.33803268305054,
+      "learning_rate": 1.9709987002151157e-06,
+      "loss": 0.6829,
+      "step": 3505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7872107027454743,
+      "learning_rate": 1.9709755982044124e-06,
+      "loss": 0.6823,
+      "step": 3506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.63260940608839,
+      "learning_rate": 1.9709524871315073e-06,
+      "loss": 0.7122,
+      "step": 3507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.8551426872696535,
+      "learning_rate": 1.970929366996617e-06,
+      "loss": 0.681,
+      "step": 3508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.091326443572131,
+      "learning_rate": 1.970906237799957e-06,
+      "loss": 0.6745,
+      "step": 3509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.0139197541761105,
+      "learning_rate": 1.9708830995417435e-06,
+      "loss": 0.6875,
+      "step": 3510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0493741642247048,
+      "learning_rate": 1.970859952222192e-06,
+      "loss": 0.6771,
+      "step": 3511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9615595052686716,
+      "learning_rate": 1.9708367958415188e-06,
+      "loss": 0.6816,
+      "step": 3512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.505168203879857,
+      "learning_rate": 1.9708136303999392e-06,
+      "loss": 0.6829,
+      "step": 3513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9920653945079048,
+      "learning_rate": 1.970790455897671e-06,
+      "loss": 0.6777,
+      "step": 3514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.698138205763462,
+      "learning_rate": 1.9707672723349297e-06,
+      "loss": 0.6816,
+      "step": 3515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.342773871708384,
+      "learning_rate": 1.970744079711931e-06,
+      "loss": 0.6829,
+      "step": 3516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.335851762071092,
+      "learning_rate": 1.970720878028892e-06,
+      "loss": 0.6842,
+      "step": 3517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2425745466147653,
+      "learning_rate": 1.9706976672860296e-06,
+      "loss": 0.694,
+      "step": 3518
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.665765807376907,
+      "learning_rate": 1.97067444748356e-06,
+      "loss": 0.6732,
+      "step": 3519
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.234647784529134,
+      "learning_rate": 1.9706512186216996e-06,
+      "loss": 0.6947,
+      "step": 3520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.219799714259669,
+      "learning_rate": 1.9706279807006657e-06,
+      "loss": 0.6829,
+      "step": 3521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6340866786377587,
+      "learning_rate": 1.970604733720675e-06,
+      "loss": 0.6966,
+      "step": 3522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4183458307990087,
+      "learning_rate": 1.970581477681945e-06,
+      "loss": 0.6816,
+      "step": 3523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.758255444165924,
+      "learning_rate": 1.9705582125846914e-06,
+      "loss": 0.6725,
+      "step": 3524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.918031474809785,
+      "learning_rate": 1.9705349384291326e-06,
+      "loss": 0.6914,
+      "step": 3525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.783067578654024,
+      "learning_rate": 1.9705116552154857e-06,
+      "loss": 0.6966,
+      "step": 3526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.433904413299071,
+      "learning_rate": 1.970488362943967e-06,
+      "loss": 0.6882,
+      "step": 3527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7680183571638314,
+      "learning_rate": 1.970465061614795e-06,
+      "loss": 0.6673,
+      "step": 3528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.896723069816814,
+      "learning_rate": 1.970441751228186e-06,
+      "loss": 0.6947,
+      "step": 3529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.427634833547123,
+      "learning_rate": 1.9704184317843596e-06,
+      "loss": 0.6797,
+      "step": 3530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.754002834436968,
+      "learning_rate": 1.9703951032835315e-06,
+      "loss": 0.7031,
+      "step": 3531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.8718522954910535,
+      "learning_rate": 1.97037176572592e-06,
+      "loss": 0.6732,
+      "step": 3532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6399882499125903,
+      "learning_rate": 1.970348419111743e-06,
+      "loss": 0.7083,
+      "step": 3533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8800335070831883,
+      "learning_rate": 1.9703250634412186e-06,
+      "loss": 0.6699,
+      "step": 3534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.5370472614907635,
+      "learning_rate": 1.9703016987145642e-06,
+      "loss": 0.6966,
+      "step": 3535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.635494803024828,
+      "learning_rate": 1.9702783249319986e-06,
+      "loss": 0.7005,
+      "step": 3536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5219030681291437,
+      "learning_rate": 1.9702549420937397e-06,
+      "loss": 0.7018,
+      "step": 3537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.993109168566994,
+      "learning_rate": 1.9702315502000053e-06,
+      "loss": 0.6992,
+      "step": 3538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.451277227844012,
+      "learning_rate": 1.970208149251014e-06,
+      "loss": 0.6628,
+      "step": 3539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6951726991547227,
+      "learning_rate": 1.9701847392469847e-06,
+      "loss": 0.6829,
+      "step": 3540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7996316804926198,
+      "learning_rate": 1.970161320188135e-06,
+      "loss": 0.6908,
+      "step": 3541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5732038826822525,
+      "learning_rate": 1.970137892074684e-06,
+      "loss": 0.6745,
+      "step": 3542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9587112298861995,
+      "learning_rate": 1.9701144549068505e-06,
+      "loss": 0.6784,
+      "step": 3543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5288555034870566,
+      "learning_rate": 1.970091008684853e-06,
+      "loss": 0.7038,
+      "step": 3544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.350544149744056,
+      "learning_rate": 1.9700675534089097e-06,
+      "loss": 0.6673,
+      "step": 3545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.9332381898261914,
+      "learning_rate": 1.9700440890792403e-06,
+      "loss": 0.6875,
+      "step": 3546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.187464901320088,
+      "learning_rate": 1.9700206156960643e-06,
+      "loss": 0.694,
+      "step": 3547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.656791502303474,
+      "learning_rate": 1.9699971332595994e-06,
+      "loss": 0.6699,
+      "step": 3548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7756807291794633,
+      "learning_rate": 1.969973641770066e-06,
+      "loss": 0.6888,
+      "step": 3549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.496462882828012,
+      "learning_rate": 1.9699501412276825e-06,
+      "loss": 0.6927,
+      "step": 3550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.666027201886202,
+      "learning_rate": 1.9699266316326685e-06,
+      "loss": 0.6732,
+      "step": 3551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.918367989190711,
+      "learning_rate": 1.9699031129852436e-06,
+      "loss": 0.707,
+      "step": 3552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5588843518984536,
+      "learning_rate": 1.969879585285627e-06,
+      "loss": 0.6829,
+      "step": 3553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1195752522419609,
+      "learning_rate": 1.969856048534039e-06,
+      "loss": 0.6816,
+      "step": 3554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.621024908198184,
+      "learning_rate": 1.969832502730698e-06,
+      "loss": 0.6862,
+      "step": 3555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9564199039169945,
+      "learning_rate": 1.969808947875825e-06,
+      "loss": 0.707,
+      "step": 3556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.151435337942184,
+      "learning_rate": 1.9697853839696395e-06,
+      "loss": 0.6849,
+      "step": 3557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3763400516511493,
+      "learning_rate": 1.969761811012361e-06,
+      "loss": 0.679,
+      "step": 3558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.963512337798386,
+      "learning_rate": 1.9697382290042094e-06,
+      "loss": 0.6797,
+      "step": 3559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1302384136339416,
+      "learning_rate": 1.9697146379454057e-06,
+      "loss": 0.6712,
+      "step": 3560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3243947225862134,
+      "learning_rate": 1.969691037836169e-06,
+      "loss": 0.6875,
+      "step": 3561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3051323047794186,
+      "learning_rate": 1.9696674286767204e-06,
+      "loss": 0.681,
+      "step": 3562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9891936539987681,
+      "learning_rate": 1.9696438104672798e-06,
+      "loss": 0.694,
+      "step": 3563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9505473676023859,
+      "learning_rate": 1.969620183208068e-06,
+      "loss": 0.6673,
+      "step": 3564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.810352792794082,
+      "learning_rate": 1.969596546899305e-06,
+      "loss": 0.6888,
+      "step": 3565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.672511325520642,
+      "learning_rate": 1.9695729015412117e-06,
+      "loss": 0.6966,
+      "step": 3566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8768071903047288,
+      "learning_rate": 1.969549247134009e-06,
+      "loss": 0.6836,
+      "step": 3567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5330383739598616,
+      "learning_rate": 1.9695255836779174e-06,
+      "loss": 0.6764,
+      "step": 3568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1392415679453876,
+      "learning_rate": 1.9695019111731575e-06,
+      "loss": 0.7064,
+      "step": 3569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4732741409610273,
+      "learning_rate": 1.9694782296199507e-06,
+      "loss": 0.7025,
+      "step": 3570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.463378462623523,
+      "learning_rate": 1.969454539018518e-06,
+      "loss": 0.6719,
+      "step": 3571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4473621218775345,
+      "learning_rate": 1.96943083936908e-06,
+      "loss": 0.6738,
+      "step": 3572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.462615171688518,
+      "learning_rate": 1.9694071306718585e-06,
+      "loss": 0.679,
+      "step": 3573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.916905013888394,
+      "learning_rate": 1.9693834129270747e-06,
+      "loss": 0.6882,
+      "step": 3574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.0835019955868175,
+      "learning_rate": 1.9693596861349495e-06,
+      "loss": 0.6576,
+      "step": 3575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.82407869543537,
+      "learning_rate": 1.969335950295705e-06,
+      "loss": 0.694,
+      "step": 3576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.4955395580249995,
+      "learning_rate": 1.9693122054095617e-06,
+      "loss": 0.6842,
+      "step": 3577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.711701973488923,
+      "learning_rate": 1.9692884514767424e-06,
+      "loss": 0.6641,
+      "step": 3578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0821108887887332,
+      "learning_rate": 1.9692646884974677e-06,
+      "loss": 0.679,
+      "step": 3579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.939240837516655,
+      "learning_rate": 1.9692409164719607e-06,
+      "loss": 0.6608,
+      "step": 3580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3810394695217574,
+      "learning_rate": 1.9692171354004417e-06,
+      "loss": 0.6953,
+      "step": 3581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8251247493628804,
+      "learning_rate": 1.9691933452831343e-06,
+      "loss": 0.6654,
+      "step": 3582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1184837110688253,
+      "learning_rate": 1.969169546120259e-06,
+      "loss": 0.6901,
+      "step": 3583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8933573181919225,
+      "learning_rate": 1.969145737912039e-06,
+      "loss": 0.6797,
+      "step": 3584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.5895607220243,
+      "learning_rate": 1.969121920658696e-06,
+      "loss": 0.6647,
+      "step": 3585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.73171948947845,
+      "learning_rate": 1.9690980943604523e-06,
+      "loss": 0.6855,
+      "step": 3586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.866092360914073,
+      "learning_rate": 1.9690742590175304e-06,
+      "loss": 0.6764,
+      "step": 3587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.597298374348386,
+      "learning_rate": 1.969050414630153e-06,
+      "loss": 0.707,
+      "step": 3588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9303640017597252,
+      "learning_rate": 1.969026561198542e-06,
+      "loss": 0.7188,
+      "step": 3589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.170024686038682,
+      "learning_rate": 1.969002698722921e-06,
+      "loss": 0.6738,
+      "step": 3590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2735795298093024,
+      "learning_rate": 1.9689788272035116e-06,
+      "loss": 0.6875,
+      "step": 3591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.961760363617031,
+      "learning_rate": 1.9689549466405373e-06,
+      "loss": 0.6868,
+      "step": 3592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5807653095794687,
+      "learning_rate": 1.9689310570342208e-06,
+      "loss": 0.6966,
+      "step": 3593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3708786255471144,
+      "learning_rate": 1.968907158384785e-06,
+      "loss": 0.6784,
+      "step": 3594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.167946159089024,
+      "learning_rate": 1.9688832506924527e-06,
+      "loss": 0.6842,
+      "step": 3595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7934129563619194,
+      "learning_rate": 1.968859333957448e-06,
+      "loss": 0.6829,
+      "step": 3596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.839411804374205,
+      "learning_rate": 1.968835408179993e-06,
+      "loss": 0.6895,
+      "step": 3597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.347911374383133,
+      "learning_rate": 1.968811473360312e-06,
+      "loss": 0.6823,
+      "step": 3598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.079747059220617,
+      "learning_rate": 1.968787529498627e-06,
+      "loss": 0.6849,
+      "step": 3599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6836367966392416,
+      "learning_rate": 1.9687635765951627e-06,
+      "loss": 0.6673,
+      "step": 3600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7495840620639854,
+      "learning_rate": 1.9687396146501425e-06,
+      "loss": 0.6745,
+      "step": 3601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7337724534988834,
+      "learning_rate": 1.9687156436637896e-06,
+      "loss": 0.6979,
+      "step": 3602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.43907575434639706,
+      "learning_rate": 1.968691663636328e-06,
+      "loss": 0.6719,
+      "step": 3603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.696253727765155,
+      "learning_rate": 1.9686676745679815e-06,
+      "loss": 0.6901,
+      "step": 3604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7366895016029822,
+      "learning_rate": 1.9686436764589734e-06,
+      "loss": 0.6732,
+      "step": 3605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.034748150432366,
+      "learning_rate": 1.968619669309529e-06,
+      "loss": 0.6882,
+      "step": 3606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.664784857869408,
+      "learning_rate": 1.9685956531198707e-06,
+      "loss": 0.6875,
+      "step": 3607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5213461740754184,
+      "learning_rate": 1.968571627890224e-06,
+      "loss": 0.6732,
+      "step": 3608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4857547658388524,
+      "learning_rate": 1.968547593620813e-06,
+      "loss": 0.6882,
+      "step": 3609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5288782611238645,
+      "learning_rate": 1.968523550311861e-06,
+      "loss": 0.696,
+      "step": 3610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6778919992781658,
+      "learning_rate": 1.968499497963593e-06,
+      "loss": 0.6719,
+      "step": 3611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.144118482638643,
+      "learning_rate": 1.9684754365762345e-06,
+      "loss": 0.6732,
+      "step": 3612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8914479605053919,
+      "learning_rate": 1.968451366150008e-06,
+      "loss": 0.6875,
+      "step": 3613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3866457955456535,
+      "learning_rate": 1.9684272866851397e-06,
+      "loss": 0.6979,
+      "step": 3614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6757865785873396,
+      "learning_rate": 1.968403198181854e-06,
+      "loss": 0.6738,
+      "step": 3615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.472663236032101,
+      "learning_rate": 1.9683791006403756e-06,
+      "loss": 0.6745,
+      "step": 3616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8884580043697836,
+      "learning_rate": 1.9683549940609293e-06,
+      "loss": 0.6732,
+      "step": 3617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.184649030279349,
+      "learning_rate": 1.96833087844374e-06,
+      "loss": 0.6758,
+      "step": 3618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3210561339896136,
+      "learning_rate": 1.9683067537890333e-06,
+      "loss": 0.6882,
+      "step": 3619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8510506182891955,
+      "learning_rate": 1.9682826200970343e-06,
+      "loss": 0.6784,
+      "step": 3620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.138523162273659,
+      "learning_rate": 1.9682584773679675e-06,
+      "loss": 0.7012,
+      "step": 3621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4625007672276469,
+      "learning_rate": 1.9682343256020587e-06,
+      "loss": 0.6868,
+      "step": 3622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9637723972345444,
+      "learning_rate": 1.9682101647995335e-06,
+      "loss": 0.6966,
+      "step": 3623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1567341033791125,
+      "learning_rate": 1.9681859949606174e-06,
+      "loss": 0.6621,
+      "step": 3624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3594543712297003,
+      "learning_rate": 1.968161816085535e-06,
+      "loss": 0.7005,
+      "step": 3625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3398909804120573,
+      "learning_rate": 1.9681376281745137e-06,
+      "loss": 0.6927,
+      "step": 3626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7974214204358928,
+      "learning_rate": 1.9681134312277777e-06,
+      "loss": 0.6732,
+      "step": 3627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.353501886564669,
+      "learning_rate": 1.9680892252455536e-06,
+      "loss": 0.6849,
+      "step": 3628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3803021995330034,
+      "learning_rate": 1.968065010228067e-06,
+      "loss": 0.6895,
+      "step": 3629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.8477894712186345,
+      "learning_rate": 1.9680407861755446e-06,
+      "loss": 0.6901,
+      "step": 3630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4670733729152745,
+      "learning_rate": 1.9680165530882116e-06,
+      "loss": 0.6745,
+      "step": 3631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.2319592792209315,
+      "learning_rate": 1.9679923109662942e-06,
+      "loss": 0.6947,
+      "step": 3632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.790642825941291,
+      "learning_rate": 1.9679680598100196e-06,
+      "loss": 0.6927,
+      "step": 3633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.179558056193482,
+      "learning_rate": 1.967943799619613e-06,
+      "loss": 0.6875,
+      "step": 3634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5740789574946907,
+      "learning_rate": 1.967919530395301e-06,
+      "loss": 0.6947,
+      "step": 3635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5191899145524347,
+      "learning_rate": 1.967895252137311e-06,
+      "loss": 0.6862,
+      "step": 3636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.468106084507407,
+      "learning_rate": 1.967870964845869e-06,
+      "loss": 0.6758,
+      "step": 3637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.94370922670345,
+      "learning_rate": 1.9678466685212017e-06,
+      "loss": 0.6797,
+      "step": 3638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.0996475215432615,
+      "learning_rate": 1.9678223631635357e-06,
+      "loss": 0.6764,
+      "step": 3639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.298703800001254,
+      "learning_rate": 1.9677980487730977e-06,
+      "loss": 0.6901,
+      "step": 3640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.474378098522654,
+      "learning_rate": 1.967773725350115e-06,
+      "loss": 0.6921,
+      "step": 3641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8032390544675003,
+      "learning_rate": 1.967749392894815e-06,
+      "loss": 0.6855,
+      "step": 3642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.386196891552804,
+      "learning_rate": 1.9677250514074243e-06,
+      "loss": 0.6908,
+      "step": 3643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.737834247611416,
+      "learning_rate": 1.9677007008881698e-06,
+      "loss": 0.6641,
+      "step": 3644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2124108183276345,
+      "learning_rate": 1.967676341337279e-06,
+      "loss": 0.6921,
+      "step": 3645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0414629178265122,
+      "learning_rate": 1.9676519727549796e-06,
+      "loss": 0.6758,
+      "step": 3646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.236631700128787,
+      "learning_rate": 1.967627595141499e-06,
+      "loss": 0.6979,
+      "step": 3647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.119853535283669,
+      "learning_rate": 1.967603208497064e-06,
+      "loss": 0.6953,
+      "step": 3648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9116912866995548,
+      "learning_rate": 1.9675788128219026e-06,
+      "loss": 0.6764,
+      "step": 3649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2768312782923426,
+      "learning_rate": 1.967554408116243e-06,
+      "loss": 0.6719,
+      "step": 3650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5868529843443566,
+      "learning_rate": 1.9675299943803123e-06,
+      "loss": 0.6882,
+      "step": 3651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.176932832604509,
+      "learning_rate": 1.9675055716143384e-06,
+      "loss": 0.6849,
+      "step": 3652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.610695680340132,
+      "learning_rate": 1.96748113981855e-06,
+      "loss": 0.6777,
+      "step": 3653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6709460427609155,
+      "learning_rate": 1.967456698993174e-06,
+      "loss": 0.6725,
+      "step": 3654
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.206862379209847,
+      "learning_rate": 1.9674322491384395e-06,
+      "loss": 0.6686,
+      "step": 3655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7974632559499063,
+      "learning_rate": 1.967407790254574e-06,
+      "loss": 0.6751,
+      "step": 3656
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.722734748150511,
+      "learning_rate": 1.967383322341806e-06,
+      "loss": 0.6888,
+      "step": 3657
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.93457476926023,
+      "learning_rate": 1.967358845400364e-06,
+      "loss": 0.6855,
+      "step": 3658
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0784467740433386,
+      "learning_rate": 1.9673343594304766e-06,
+      "loss": 0.7044,
+      "step": 3659
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3538162933011684,
+      "learning_rate": 1.9673098644323718e-06,
+      "loss": 0.7077,
+      "step": 3660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.045816313115465,
+      "learning_rate": 1.9672853604062785e-06,
+      "loss": 0.668,
+      "step": 3661
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.755817336486679,
+      "learning_rate": 1.9672608473524256e-06,
+      "loss": 0.6797,
+      "step": 3662
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.987639788974708,
+      "learning_rate": 1.967236325271041e-06,
+      "loss": 0.6719,
+      "step": 3663
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4914882600162653,
+      "learning_rate": 1.9672117941623552e-06,
+      "loss": 0.6868,
+      "step": 3664
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6932894896623811,
+      "learning_rate": 1.9671872540265957e-06,
+      "loss": 0.6771,
+      "step": 3665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.45853153974924,
+      "learning_rate": 1.967162704863992e-06,
+      "loss": 0.679,
+      "step": 3666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2795099993686705,
+      "learning_rate": 1.9671381466747734e-06,
+      "loss": 0.6758,
+      "step": 3667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.208288087877923,
+      "learning_rate": 1.967113579459169e-06,
+      "loss": 0.6973,
+      "step": 3668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3029182384520326,
+      "learning_rate": 1.9670890032174077e-06,
+      "loss": 0.6803,
+      "step": 3669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.743093302535752,
+      "learning_rate": 1.9670644179497194e-06,
+      "loss": 0.696,
+      "step": 3670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.164164298831938,
+      "learning_rate": 1.9670398236563333e-06,
+      "loss": 0.6921,
+      "step": 3671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1586169021232084,
+      "learning_rate": 1.9670152203374792e-06,
+      "loss": 0.6973,
+      "step": 3672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.39785026097882,
+      "learning_rate": 1.966990607993386e-06,
+      "loss": 0.6849,
+      "step": 3673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2243715982332164,
+      "learning_rate": 1.9669659866242845e-06,
+      "loss": 0.6895,
+      "step": 3674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.0234968358374699,
+      "learning_rate": 1.966941356230404e-06,
+      "loss": 0.6693,
+      "step": 3675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8485748209311277,
+      "learning_rate": 1.966916716811974e-06,
+      "loss": 0.6634,
+      "step": 3676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.842783213630202,
+      "learning_rate": 1.966892068369225e-06,
+      "loss": 0.7168,
+      "step": 3677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.653873396743471,
+      "learning_rate": 1.9668674109023865e-06,
+      "loss": 0.7018,
+      "step": 3678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.271410374701863,
+      "learning_rate": 1.966842744411689e-06,
+      "loss": 0.6888,
+      "step": 3679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1183514469547402,
+      "learning_rate": 1.9668180688973627e-06,
+      "loss": 0.6784,
+      "step": 3680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1175565394092517,
+      "learning_rate": 1.966793384359638e-06,
+      "loss": 0.6842,
+      "step": 3681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.132639851412952,
+      "learning_rate": 1.9667686907987454e-06,
+      "loss": 0.6862,
+      "step": 3682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3075110608724112,
+      "learning_rate": 1.9667439882149148e-06,
+      "loss": 0.6953,
+      "step": 3683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9777636161047825,
+      "learning_rate": 1.966719276608377e-06,
+      "loss": 0.6602,
+      "step": 3684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.015347643454091,
+      "learning_rate": 1.9666945559793628e-06,
+      "loss": 0.6803,
+      "step": 3685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5503637940131926,
+      "learning_rate": 1.9666698263281026e-06,
+      "loss": 0.6966,
+      "step": 3686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2468316281595104,
+      "learning_rate": 1.9666450876548273e-06,
+      "loss": 0.6732,
+      "step": 3687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4925165240348916,
+      "learning_rate": 1.9666203399597685e-06,
+      "loss": 0.6836,
+      "step": 3688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8151180705138686,
+      "learning_rate": 1.966595583243156e-06,
+      "loss": 0.6888,
+      "step": 3689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.704402345831273,
+      "learning_rate": 1.966570817505222e-06,
+      "loss": 0.6992,
+      "step": 3690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4741413411692523,
+      "learning_rate": 1.9665460427461967e-06,
+      "loss": 0.6803,
+      "step": 3691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8118607788144223,
+      "learning_rate": 1.966521258966312e-06,
+      "loss": 0.6732,
+      "step": 3692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5080881957703163,
+      "learning_rate": 1.9664964661657985e-06,
+      "loss": 0.6973,
+      "step": 3693
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5791105844964657,
+      "learning_rate": 1.9664716643448884e-06,
+      "loss": 0.6875,
+      "step": 3694
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.345589788877535,
+      "learning_rate": 1.9664468535038126e-06,
+      "loss": 0.6921,
+      "step": 3695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5242076335566082,
+      "learning_rate": 1.966422033642803e-06,
+      "loss": 0.6855,
+      "step": 3696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.859772626761817,
+      "learning_rate": 1.9663972047620907e-06,
+      "loss": 0.6986,
+      "step": 3697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.451286746267935,
+      "learning_rate": 1.9663723668619084e-06,
+      "loss": 0.6738,
+      "step": 3698
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.582202455098912,
+      "learning_rate": 1.966347519942487e-06,
+      "loss": 0.6745,
+      "step": 3699
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6844207369904334,
+      "learning_rate": 1.9663226640040587e-06,
+      "loss": 0.6823,
+      "step": 3700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5988983407193347,
+      "learning_rate": 1.9662977990468556e-06,
+      "loss": 0.6934,
+      "step": 3701
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5606673237990367,
+      "learning_rate": 1.9662729250711098e-06,
+      "loss": 0.6875,
+      "step": 3702
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7383977010906838,
+      "learning_rate": 1.9662480420770532e-06,
+      "loss": 0.6784,
+      "step": 3703
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5577519233570305,
+      "learning_rate": 1.966223150064918e-06,
+      "loss": 0.6745,
+      "step": 3704
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.401823679669317,
+      "learning_rate": 1.966198249034937e-06,
+      "loss": 0.6738,
+      "step": 3705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.486324266732481,
+      "learning_rate": 1.9661733389873422e-06,
+      "loss": 0.6823,
+      "step": 3706
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.026627755623926,
+      "learning_rate": 1.966148419922366e-06,
+      "loss": 0.6745,
+      "step": 3707
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.780836996663193,
+      "learning_rate": 1.9661234918402413e-06,
+      "loss": 0.7005,
+      "step": 3708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.807484319477287,
+      "learning_rate": 1.9660985547412006e-06,
+      "loss": 0.6901,
+      "step": 3709
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3065071552048613,
+      "learning_rate": 1.9660736086254768e-06,
+      "loss": 0.6999,
+      "step": 3710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.831822834866347,
+      "learning_rate": 1.9660486534933023e-06,
+      "loss": 0.6855,
+      "step": 3711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.587260624263724,
+      "learning_rate": 1.9660236893449107e-06,
+      "loss": 0.6966,
+      "step": 3712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.74296962552928,
+      "learning_rate": 1.9659987161805344e-06,
+      "loss": 0.6901,
+      "step": 3713
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.728369726001187,
+      "learning_rate": 1.9659737340004067e-06,
+      "loss": 0.6667,
+      "step": 3714
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8050416409455917,
+      "learning_rate": 1.96594874280476e-06,
+      "loss": 0.6725,
+      "step": 3715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.091136714092812,
+      "learning_rate": 1.965923742593829e-06,
+      "loss": 0.7194,
+      "step": 3716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3675373669400974,
+      "learning_rate": 1.965898733367846e-06,
+      "loss": 0.6777,
+      "step": 3717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.559252503320548,
+      "learning_rate": 1.9658737151270454e-06,
+      "loss": 0.6973,
+      "step": 3718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.977029440042897,
+      "learning_rate": 1.965848687871659e-06,
+      "loss": 0.7096,
+      "step": 3719
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.1410566658459755,
+      "learning_rate": 1.965823651601922e-06,
+      "loss": 0.6829,
+      "step": 3720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.43638493301439035,
+      "learning_rate": 1.9657986063180673e-06,
+      "loss": 0.6895,
+      "step": 3721
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8091999605402336,
+      "learning_rate": 1.965773552020329e-06,
+      "loss": 0.6934,
+      "step": 3722
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4207734051158596,
+      "learning_rate": 1.96574848870894e-06,
+      "loss": 0.6953,
+      "step": 3723
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4149243265913563,
+      "learning_rate": 1.9657234163841357e-06,
+      "loss": 0.6914,
+      "step": 3724
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.303171284325851,
+      "learning_rate": 1.965698335046149e-06,
+      "loss": 0.6764,
+      "step": 3725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.531516925916454,
+      "learning_rate": 1.9656732446952146e-06,
+      "loss": 0.6764,
+      "step": 3726
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7864803366761004,
+      "learning_rate": 1.965648145331566e-06,
+      "loss": 0.6862,
+      "step": 3727
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.894009984196398,
+      "learning_rate": 1.965623036955438e-06,
+      "loss": 0.6868,
+      "step": 3728
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.049867504624438,
+      "learning_rate": 1.965597919567065e-06,
+      "loss": 0.6908,
+      "step": 3729
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.565187305079142,
+      "learning_rate": 1.965572793166681e-06,
+      "loss": 0.6888,
+      "step": 3730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.52069314700066,
+      "learning_rate": 1.965547657754521e-06,
+      "loss": 0.7051,
+      "step": 3731
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.360214296857064,
+      "learning_rate": 1.965522513330819e-06,
+      "loss": 0.6849,
+      "step": 3732
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.500339432686903,
+      "learning_rate": 1.9654973598958105e-06,
+      "loss": 0.694,
+      "step": 3733
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.894216184673713,
+      "learning_rate": 1.965472197449729e-06,
+      "loss": 0.6803,
+      "step": 3734
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.218971054787895,
+      "learning_rate": 1.9654470259928108e-06,
+      "loss": 0.6855,
+      "step": 3735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.8019126164047625,
+      "learning_rate": 1.96542184552529e-06,
+      "loss": 0.6764,
+      "step": 3736
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6287788868470976,
+      "learning_rate": 1.9653966560474012e-06,
+      "loss": 0.696,
+      "step": 3737
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.812260055488206,
+      "learning_rate": 1.9653714575593805e-06,
+      "loss": 0.696,
+      "step": 3738
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9485862299119361,
+      "learning_rate": 1.9653462500614627e-06,
+      "loss": 0.6875,
+      "step": 3739
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9881288306921268,
+      "learning_rate": 1.965321033553883e-06,
+      "loss": 0.7018,
+      "step": 3740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.4488239336601,
+      "learning_rate": 1.965295808036876e-06,
+      "loss": 0.6654,
+      "step": 3741
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.467402850939774,
+      "learning_rate": 1.9652705735106786e-06,
+      "loss": 0.6966,
+      "step": 3742
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.09520792189266,
+      "learning_rate": 1.9652453299755256e-06,
+      "loss": 0.6862,
+      "step": 3743
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.768131398368025,
+      "learning_rate": 1.965220077431652e-06,
+      "loss": 0.6732,
+      "step": 3744
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.007908546397788,
+      "learning_rate": 1.9651948158792945e-06,
+      "loss": 0.7077,
+      "step": 3745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.34080261520013,
+      "learning_rate": 1.9651695453186883e-06,
+      "loss": 0.6908,
+      "step": 3746
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5106959085884515,
+      "learning_rate": 1.9651442657500693e-06,
+      "loss": 0.6738,
+      "step": 3747
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6145137563879117,
+      "learning_rate": 1.9651189771736737e-06,
+      "loss": 0.7005,
+      "step": 3748
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.342788949678008,
+      "learning_rate": 1.9650936795897375e-06,
+      "loss": 0.7109,
+      "step": 3749
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7402105330432001,
+      "learning_rate": 1.965068372998496e-06,
+      "loss": 0.6784,
+      "step": 3750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9639859169529568,
+      "learning_rate": 1.9650430574001867e-06,
+      "loss": 0.6797,
+      "step": 3751
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6637639697435727,
+      "learning_rate": 1.9650177327950452e-06,
+      "loss": 0.679,
+      "step": 3752
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1894620700007232,
+      "learning_rate": 1.964992399183307e-06,
+      "loss": 0.6921,
+      "step": 3753
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.847707077095592,
+      "learning_rate": 1.9649670565652104e-06,
+      "loss": 0.668,
+      "step": 3754
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.587382270285597,
+      "learning_rate": 1.9649417049409906e-06,
+      "loss": 0.6973,
+      "step": 3755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9383416732049389,
+      "learning_rate": 1.9649163443108844e-06,
+      "loss": 0.6816,
+      "step": 3756
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.292110936324973,
+      "learning_rate": 1.9648909746751286e-06,
+      "loss": 0.6751,
+      "step": 3757
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6547609057392265,
+      "learning_rate": 1.96486559603396e-06,
+      "loss": 0.6875,
+      "step": 3758
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.874749821957016,
+      "learning_rate": 1.9648402083876154e-06,
+      "loss": 0.6829,
+      "step": 3759
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.1454685957970225,
+      "learning_rate": 1.964814811736332e-06,
+      "loss": 0.6764,
+      "step": 3760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9813448207539253,
+      "learning_rate": 1.9647894060803463e-06,
+      "loss": 0.6745,
+      "step": 3761
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6220997319214924,
+      "learning_rate": 1.9647639914198964e-06,
+      "loss": 0.6947,
+      "step": 3762
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.687799170564092,
+      "learning_rate": 1.9647385677552182e-06,
+      "loss": 0.6712,
+      "step": 3763
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.766074453562593,
+      "learning_rate": 1.96471313508655e-06,
+      "loss": 0.668,
+      "step": 3764
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0895104110279157,
+      "learning_rate": 1.964687693414129e-06,
+      "loss": 0.6927,
+      "step": 3765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8148210378090583,
+      "learning_rate": 1.964662242738192e-06,
+      "loss": 0.6895,
+      "step": 3766
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4260581386490567,
+      "learning_rate": 1.964636783058977e-06,
+      "loss": 0.6836,
+      "step": 3767
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3243533445575006,
+      "learning_rate": 1.9646113143767217e-06,
+      "loss": 0.6803,
+      "step": 3768
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.361578499152654,
+      "learning_rate": 1.964585836691664e-06,
+      "loss": 0.6816,
+      "step": 3769
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0160336827603258,
+      "learning_rate": 1.964560350004041e-06,
+      "loss": 0.6686,
+      "step": 3770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.002235608822072,
+      "learning_rate": 1.964534854314091e-06,
+      "loss": 0.7051,
+      "step": 3771
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.919036141585281,
+      "learning_rate": 1.9645093496220516e-06,
+      "loss": 0.6895,
+      "step": 3772
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7562778488848814,
+      "learning_rate": 1.9644838359281617e-06,
+      "loss": 0.6719,
+      "step": 3773
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.515015585848878,
+      "learning_rate": 1.9644583132326586e-06,
+      "loss": 0.6777,
+      "step": 3774
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.439102268246573,
+      "learning_rate": 1.964432781535781e-06,
+      "loss": 0.6934,
+      "step": 3775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.945146635713698,
+      "learning_rate": 1.9644072408377666e-06,
+      "loss": 0.681,
+      "step": 3776
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6824169195298266,
+      "learning_rate": 1.9643816911388545e-06,
+      "loss": 0.6888,
+      "step": 3777
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0421027320810763,
+      "learning_rate": 1.9643561324392826e-06,
+      "loss": 0.6966,
+      "step": 3778
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.358988525563421,
+      "learning_rate": 1.9643305647392894e-06,
+      "loss": 0.6673,
+      "step": 3779
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4682808180660423,
+      "learning_rate": 1.964304988039114e-06,
+      "loss": 0.6875,
+      "step": 3780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5684492397625667,
+      "learning_rate": 1.9642794023389953e-06,
+      "loss": 0.6875,
+      "step": 3781
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.1814185831359894,
+      "learning_rate": 1.964253807639171e-06,
+      "loss": 0.681,
+      "step": 3782
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6768923125984199,
+      "learning_rate": 1.964228203939881e-06,
+      "loss": 0.6764,
+      "step": 3783
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.610517694559695,
+      "learning_rate": 1.9642025912413637e-06,
+      "loss": 0.6901,
+      "step": 3784
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.778411729644477,
+      "learning_rate": 1.9641769695438582e-06,
+      "loss": 0.7135,
+      "step": 3785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6968882853839965,
+      "learning_rate": 1.9641513388476043e-06,
+      "loss": 0.6992,
+      "step": 3786
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6507938719020636,
+      "learning_rate": 1.9641256991528404e-06,
+      "loss": 0.666,
+      "step": 3787
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8006640764358124,
+      "learning_rate": 1.964100050459806e-06,
+      "loss": 0.679,
+      "step": 3788
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3698174584285154,
+      "learning_rate": 1.9640743927687407e-06,
+      "loss": 0.6966,
+      "step": 3789
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.408037566255943,
+      "learning_rate": 1.9640487260798838e-06,
+      "loss": 0.6777,
+      "step": 3790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.228237462574511,
+      "learning_rate": 1.964023050393475e-06,
+      "loss": 0.6725,
+      "step": 3791
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.967235259995864,
+      "learning_rate": 1.9639973657097533e-06,
+      "loss": 0.6882,
+      "step": 3792
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.634036786048584,
+      "learning_rate": 1.9639716720289595e-06,
+      "loss": 0.6953,
+      "step": 3793
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8944883823131704,
+      "learning_rate": 1.963945969351333e-06,
+      "loss": 0.6934,
+      "step": 3794
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.259666153234905,
+      "learning_rate": 1.9639202576771127e-06,
+      "loss": 0.6764,
+      "step": 3795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.529404331940159,
+      "learning_rate": 1.96389453700654e-06,
+      "loss": 0.6914,
+      "step": 3796
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8649649564085848,
+      "learning_rate": 1.963868807339854e-06,
+      "loss": 0.6803,
+      "step": 3797
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.445947854360228,
+      "learning_rate": 1.9638430686772957e-06,
+      "loss": 0.6855,
+      "step": 3798
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3558316152081895,
+      "learning_rate": 1.963817321019104e-06,
+      "loss": 0.6777,
+      "step": 3799
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8825141108701058,
+      "learning_rate": 1.963791564365521e-06,
+      "loss": 0.6797,
+      "step": 3800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0583886434641583,
+      "learning_rate": 1.9637657987167855e-06,
+      "loss": 0.6934,
+      "step": 3801
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1042933697449695,
+      "learning_rate": 1.9637400240731386e-06,
+      "loss": 0.681,
+      "step": 3802
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5831845611433404,
+      "learning_rate": 1.963714240434821e-06,
+      "loss": 0.6829,
+      "step": 3803
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.674935211522944,
+      "learning_rate": 1.963688447802073e-06,
+      "loss": 0.6868,
+      "step": 3804
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.74603858750652,
+      "learning_rate": 1.9636626461751352e-06,
+      "loss": 0.6895,
+      "step": 3805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1469455373624822,
+      "learning_rate": 1.9636368355542493e-06,
+      "loss": 0.694,
+      "step": 3806
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.887776179581965,
+      "learning_rate": 1.963611015939655e-06,
+      "loss": 0.6641,
+      "step": 3807
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8925983118316652,
+      "learning_rate": 1.9635851873315946e-06,
+      "loss": 0.6693,
+      "step": 3808
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5382339129168408,
+      "learning_rate": 1.963559349730308e-06,
+      "loss": 0.6875,
+      "step": 3809
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.403266530005583,
+      "learning_rate": 1.963533503136037e-06,
+      "loss": 0.6771,
+      "step": 3810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6534269853631822,
+      "learning_rate": 1.963507647549022e-06,
+      "loss": 0.6882,
+      "step": 3811
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9898416118320187,
+      "learning_rate": 1.9634817829695056e-06,
+      "loss": 0.6771,
+      "step": 3812
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.242468137685684,
+      "learning_rate": 1.9634559093977285e-06,
+      "loss": 0.6816,
+      "step": 3813
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.060421819538773,
+      "learning_rate": 1.9634300268339315e-06,
+      "loss": 0.6979,
+      "step": 3814
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2217601153456448,
+      "learning_rate": 1.9634041352783575e-06,
+      "loss": 0.6823,
+      "step": 3815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.668473946186744,
+      "learning_rate": 1.9633782347312475e-06,
+      "loss": 0.6699,
+      "step": 3816
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.475856252313151,
+      "learning_rate": 1.9633523251928426e-06,
+      "loss": 0.6621,
+      "step": 3817
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8736827739386337,
+      "learning_rate": 1.963326406663386e-06,
+      "loss": 0.6647,
+      "step": 3818
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.924302773097319,
+      "learning_rate": 1.9633004791431186e-06,
+      "loss": 0.7122,
+      "step": 3819
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.312038455022098,
+      "learning_rate": 1.9632745426322827e-06,
+      "loss": 0.6803,
+      "step": 3820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.119407166037436,
+      "learning_rate": 1.96324859713112e-06,
+      "loss": 0.6921,
+      "step": 3821
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.310626114666542,
+      "learning_rate": 1.963222642639873e-06,
+      "loss": 0.7454,
+      "step": 3822
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2868273519900852,
+      "learning_rate": 1.963196679158784e-06,
+      "loss": 0.6536,
+      "step": 3823
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7198580871750435,
+      "learning_rate": 1.963170706688095e-06,
+      "loss": 0.6771,
+      "step": 3824
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.32653098888482,
+      "learning_rate": 1.9631447252280495e-06,
+      "loss": 0.6771,
+      "step": 3825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.934977087537867,
+      "learning_rate": 1.9631187347788886e-06,
+      "loss": 0.6895,
+      "step": 3826
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9031189483375662,
+      "learning_rate": 1.9630927353408553e-06,
+      "loss": 0.6908,
+      "step": 3827
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.718749364216564,
+      "learning_rate": 1.963066726914192e-06,
+      "loss": 0.6758,
+      "step": 3828
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2509816412787074,
+      "learning_rate": 1.963040709499142e-06,
+      "loss": 0.6758,
+      "step": 3829
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3003813917126092,
+      "learning_rate": 1.963014683095948e-06,
+      "loss": 0.681,
+      "step": 3830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7828907468602248,
+      "learning_rate": 1.962988647704853e-06,
+      "loss": 0.6758,
+      "step": 3831
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.687745550747511,
+      "learning_rate": 1.9629626033260993e-06,
+      "loss": 0.6921,
+      "step": 3832
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.771063519583802,
+      "learning_rate": 1.962936549959931e-06,
+      "loss": 0.6921,
+      "step": 3833
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.9591017085094589,
+      "learning_rate": 1.9629104876065903e-06,
+      "loss": 0.6706,
+      "step": 3834
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.169516335284841,
+      "learning_rate": 1.962884416266321e-06,
+      "loss": 0.6999,
+      "step": 3835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.420780077259901,
+      "learning_rate": 1.9628583359393664e-06,
+      "loss": 0.668,
+      "step": 3836
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8307227312236624,
+      "learning_rate": 1.9628322466259696e-06,
+      "loss": 0.6816,
+      "step": 3837
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.602254495244231,
+      "learning_rate": 1.9628061483263744e-06,
+      "loss": 0.6647,
+      "step": 3838
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.625919157186785,
+      "learning_rate": 1.962780041040824e-06,
+      "loss": 0.6719,
+      "step": 3839
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7242649040625995,
+      "learning_rate": 1.9627539247695627e-06,
+      "loss": 0.6862,
+      "step": 3840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.828321940687699,
+      "learning_rate": 1.9627277995128336e-06,
+      "loss": 0.6732,
+      "step": 3841
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.606434434434044,
+      "learning_rate": 1.962701665270881e-06,
+      "loss": 0.6908,
+      "step": 3842
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2954440937993605,
+      "learning_rate": 1.9626755220439485e-06,
+      "loss": 0.6927,
+      "step": 3843
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8828609766103519,
+      "learning_rate": 1.9626493698322805e-06,
+      "loss": 0.6953,
+      "step": 3844
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.501729568661364,
+      "learning_rate": 1.9626232086361204e-06,
+      "loss": 0.6803,
+      "step": 3845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.422370457228134,
+      "learning_rate": 1.962597038455713e-06,
+      "loss": 0.679,
+      "step": 3846
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8012538895946655,
+      "learning_rate": 1.9625708592913026e-06,
+      "loss": 0.6738,
+      "step": 3847
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9125844000420422,
+      "learning_rate": 1.9625446711431327e-06,
+      "loss": 0.7025,
+      "step": 3848
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3605454755531463,
+      "learning_rate": 1.9625184740114482e-06,
+      "loss": 0.696,
+      "step": 3849
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.478191205448045,
+      "learning_rate": 1.962492267896494e-06,
+      "loss": 0.6875,
+      "step": 3850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6699388530126948,
+      "learning_rate": 1.9624660527985144e-06,
+      "loss": 0.6777,
+      "step": 3851
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5313534483540796,
+      "learning_rate": 1.962439828717754e-06,
+      "loss": 0.6797,
+      "step": 3852
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.858563632842911,
+      "learning_rate": 1.9624135956544574e-06,
+      "loss": 0.6855,
+      "step": 3853
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8496338630241174,
+      "learning_rate": 1.96238735360887e-06,
+      "loss": 0.6719,
+      "step": 3854
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.491726328350103,
+      "learning_rate": 1.9623611025812355e-06,
+      "loss": 0.6849,
+      "step": 3855
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.101991390128922,
+      "learning_rate": 1.9623348425718004e-06,
+      "loss": 0.6875,
+      "step": 3856
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.5565806986850985,
+      "learning_rate": 1.962308573580809e-06,
+      "loss": 0.6719,
+      "step": 3857
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3322255037096307,
+      "learning_rate": 1.9622822956085064e-06,
+      "loss": 0.6979,
+      "step": 3858
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9808085616265378,
+      "learning_rate": 1.9622560086551384e-06,
+      "loss": 0.6576,
+      "step": 3859
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4297151042146607,
+      "learning_rate": 1.9622297127209494e-06,
+      "loss": 0.6823,
+      "step": 3860
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.40022625642795,
+      "learning_rate": 1.962203407806186e-06,
+      "loss": 0.6784,
+      "step": 3861
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4582127421950792,
+      "learning_rate": 1.9621770939110926e-06,
+      "loss": 0.6953,
+      "step": 3862
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4860815685403375,
+      "learning_rate": 1.9621507710359153e-06,
+      "loss": 0.6992,
+      "step": 3863
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0733231293517567,
+      "learning_rate": 1.9621244391809e-06,
+      "loss": 0.6901,
+      "step": 3864
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.0293980443204485,
+      "learning_rate": 1.9620980983462924e-06,
+      "loss": 0.6706,
+      "step": 3865
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8379042224033117,
+      "learning_rate": 1.962071748532338e-06,
+      "loss": 0.6836,
+      "step": 3866
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5384935170930933,
+      "learning_rate": 1.962045389739283e-06,
+      "loss": 0.6745,
+      "step": 3867
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9751888672796776,
+      "learning_rate": 1.962019021967373e-06,
+      "loss": 0.6966,
+      "step": 3868
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.7809107706631995,
+      "learning_rate": 1.9619926452168543e-06,
+      "loss": 0.6758,
+      "step": 3869
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1502845582152235,
+      "learning_rate": 1.9619662594879737e-06,
+      "loss": 0.6719,
+      "step": 3870
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7175504573936635,
+      "learning_rate": 1.9619398647809762e-06,
+      "loss": 0.6803,
+      "step": 3871
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8297620886708922,
+      "learning_rate": 1.9619134610961097e-06,
+      "loss": 0.6719,
+      "step": 3872
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6243641411692787,
+      "learning_rate": 1.9618870484336193e-06,
+      "loss": 0.6764,
+      "step": 3873
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.117417899724257,
+      "learning_rate": 1.961860626793752e-06,
+      "loss": 0.6725,
+      "step": 3874
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.229875635156921,
+      "learning_rate": 1.9618341961767546e-06,
+      "loss": 0.681,
+      "step": 3875
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5917029141551127,
+      "learning_rate": 1.9618077565828736e-06,
+      "loss": 0.7005,
+      "step": 3876
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.518349856283,
+      "learning_rate": 1.9617813080123556e-06,
+      "loss": 0.7077,
+      "step": 3877
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6023187303669425,
+      "learning_rate": 1.9617548504654475e-06,
+      "loss": 0.6816,
+      "step": 3878
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8100896292642643,
+      "learning_rate": 1.9617283839423965e-06,
+      "loss": 0.6647,
+      "step": 3879
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.331182411383859,
+      "learning_rate": 1.961701908443449e-06,
+      "loss": 0.6712,
+      "step": 3880
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.634358883948567,
+      "learning_rate": 1.9616754239688533e-06,
+      "loss": 0.7077,
+      "step": 3881
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8345240397324116,
+      "learning_rate": 1.9616489305188554e-06,
+      "loss": 0.6927,
+      "step": 3882
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.052974541459646,
+      "learning_rate": 1.9616224280937035e-06,
+      "loss": 0.6927,
+      "step": 3883
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5922564288409395,
+      "learning_rate": 1.9615959166936437e-06,
+      "loss": 0.7077,
+      "step": 3884
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5842648740754557,
+      "learning_rate": 1.9615693963189248e-06,
+      "loss": 0.6999,
+      "step": 3885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6896679043815622,
+      "learning_rate": 1.9615428669697935e-06,
+      "loss": 0.6875,
+      "step": 3886
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.274744898127028,
+      "learning_rate": 1.961516328646497e-06,
+      "loss": 0.668,
+      "step": 3887
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.96287002777643,
+      "learning_rate": 1.961489781349284e-06,
+      "loss": 0.6868,
+      "step": 3888
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.819331479056289,
+      "learning_rate": 1.961463225078402e-06,
+      "loss": 0.7064,
+      "step": 3889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.367772802069378,
+      "learning_rate": 1.9614366598340984e-06,
+      "loss": 0.6999,
+      "step": 3890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.931555628527842,
+      "learning_rate": 1.961410085616622e-06,
+      "loss": 0.679,
+      "step": 3891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.511273571548001,
+      "learning_rate": 1.96138350242622e-06,
+      "loss": 0.6797,
+      "step": 3892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.582701370324537,
+      "learning_rate": 1.9613569102631403e-06,
+      "loss": 0.6921,
+      "step": 3893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2743651792970305,
+      "learning_rate": 1.9613303091276317e-06,
+      "loss": 0.6777,
+      "step": 3894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.668922470201758,
+      "learning_rate": 1.9613036990199423e-06,
+      "loss": 0.6999,
+      "step": 3895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.420067242913327,
+      "learning_rate": 1.961277079940321e-06,
+      "loss": 0.6999,
+      "step": 3896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.787314281498529,
+      "learning_rate": 1.961250451889015e-06,
+      "loss": 0.6784,
+      "step": 3897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3040130575266384,
+      "learning_rate": 1.961223814866274e-06,
+      "loss": 0.6836,
+      "step": 3898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7393538497158081,
+      "learning_rate": 1.961197168872346e-06,
+      "loss": 0.6862,
+      "step": 3899
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.174397950270958,
+      "learning_rate": 1.96117051390748e-06,
+      "loss": 0.6895,
+      "step": 3900
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.552105489514348,
+      "learning_rate": 1.961143849971924e-06,
+      "loss": 0.6855,
+      "step": 3901
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9620320726923441,
+      "learning_rate": 1.9611171770659276e-06,
+      "loss": 0.6764,
+      "step": 3902
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.450534173513307,
+      "learning_rate": 1.96109049518974e-06,
+      "loss": 0.6855,
+      "step": 3903
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.617473614216905,
+      "learning_rate": 1.9610638043436097e-06,
+      "loss": 0.6868,
+      "step": 3904
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5668238659891371,
+      "learning_rate": 1.9610371045277857e-06,
+      "loss": 0.6927,
+      "step": 3905
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7820542585544608,
+      "learning_rate": 1.9610103957425176e-06,
+      "loss": 0.6966,
+      "step": 3906
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3573145538801878,
+      "learning_rate": 1.9609836779880545e-06,
+      "loss": 0.6986,
+      "step": 3907
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5338790594619836,
+      "learning_rate": 1.9609569512646454e-06,
+      "loss": 0.6895,
+      "step": 3908
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7075086817340472,
+      "learning_rate": 1.9609302155725407e-06,
+      "loss": 0.7057,
+      "step": 3909
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.90564956342266,
+      "learning_rate": 1.9609034709119888e-06,
+      "loss": 0.6908,
+      "step": 3910
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8730321773919334,
+      "learning_rate": 1.96087671728324e-06,
+      "loss": 0.6719,
+      "step": 3911
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7458495144023884,
+      "learning_rate": 1.960849954686544e-06,
+      "loss": 0.6797,
+      "step": 3912
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.519976055124276,
+      "learning_rate": 1.96082318312215e-06,
+      "loss": 0.6849,
+      "step": 3913
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6314889539762295,
+      "learning_rate": 1.960796402590309e-06,
+      "loss": 0.6836,
+      "step": 3914
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.500973895720287,
+      "learning_rate": 1.9607696130912696e-06,
+      "loss": 0.6855,
+      "step": 3915
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.784727286440555,
+      "learning_rate": 1.9607428146252825e-06,
+      "loss": 0.679,
+      "step": 3916
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3693557616687644,
+      "learning_rate": 1.9607160071925978e-06,
+      "loss": 0.6771,
+      "step": 3917
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8853740135386574,
+      "learning_rate": 1.960689190793466e-06,
+      "loss": 0.6797,
+      "step": 3918
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.102566016979119,
+      "learning_rate": 1.960662365428136e-06,
+      "loss": 0.6875,
+      "step": 3919
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.341281006733201,
+      "learning_rate": 1.96063553109686e-06,
+      "loss": 0.6868,
+      "step": 3920
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6532126380887497,
+      "learning_rate": 1.9606086877998877e-06,
+      "loss": 0.6732,
+      "step": 3921
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2995527461786816,
+      "learning_rate": 1.9605818355374693e-06,
+      "loss": 0.6882,
+      "step": 3922
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.187022434909524,
+      "learning_rate": 1.960554974309856e-06,
+      "loss": 0.6615,
+      "step": 3923
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.405513982947554,
+      "learning_rate": 1.9605281041172975e-06,
+      "loss": 0.6732,
+      "step": 3924
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.175441911497109,
+      "learning_rate": 1.9605012249600457e-06,
+      "loss": 0.6595,
+      "step": 3925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.577844835985271,
+      "learning_rate": 1.960474336838351e-06,
+      "loss": 0.6797,
+      "step": 3926
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2809690172249064,
+      "learning_rate": 1.9604474397524644e-06,
+      "loss": 0.681,
+      "step": 3927
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1819926849794826,
+      "learning_rate": 1.960420533702637e-06,
+      "loss": 0.6589,
+      "step": 3928
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.994122820097765,
+      "learning_rate": 1.9603936186891196e-06,
+      "loss": 0.666,
+      "step": 3929
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0055661193507377,
+      "learning_rate": 1.9603666947121637e-06,
+      "loss": 0.6921,
+      "step": 3930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3750371687971645,
+      "learning_rate": 1.960339761772021e-06,
+      "loss": 0.6478,
+      "step": 3931
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9185301810686155,
+      "learning_rate": 1.9603128198689416e-06,
+      "loss": 0.6888,
+      "step": 3932
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8272131350141942,
+      "learning_rate": 1.960285869003178e-06,
+      "loss": 0.7207,
+      "step": 3933
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.892668916990297,
+      "learning_rate": 1.9602589091749814e-06,
+      "loss": 0.6738,
+      "step": 3934
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.997270861776444,
+      "learning_rate": 1.960231940384604e-06,
+      "loss": 0.7311,
+      "step": 3935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.366762084384216,
+      "learning_rate": 1.960204962632296e-06,
+      "loss": 0.6823,
+      "step": 3936
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2174896554451022,
+      "learning_rate": 1.960177975918311e-06,
+      "loss": 0.6914,
+      "step": 3937
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.326649322533905,
+      "learning_rate": 1.9601509802428998e-06,
+      "loss": 0.6634,
+      "step": 3938
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9605417845694567,
+      "learning_rate": 1.9601239756063146e-06,
+      "loss": 0.7109,
+      "step": 3939
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.682364880252619,
+      "learning_rate": 1.9600969620088073e-06,
+      "loss": 0.7135,
+      "step": 3940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3441056391409982,
+      "learning_rate": 1.96006993945063e-06,
+      "loss": 0.7051,
+      "step": 3941
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.744669484340444,
+      "learning_rate": 1.9600429079320354e-06,
+      "loss": 0.6836,
+      "step": 3942
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.1607581795770285,
+      "learning_rate": 1.9600158674532752e-06,
+      "loss": 0.6797,
+      "step": 3943
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9618435704119774,
+      "learning_rate": 1.959988818014602e-06,
+      "loss": 0.7181,
+      "step": 3944
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.451522302465227,
+      "learning_rate": 1.9599617596162687e-06,
+      "loss": 0.6868,
+      "step": 3945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.455942226359356,
+      "learning_rate": 1.959934692258527e-06,
+      "loss": 0.6699,
+      "step": 3946
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.37638994941778,
+      "learning_rate": 1.95990761594163e-06,
+      "loss": 0.7005,
+      "step": 3947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5476921122875082,
+      "learning_rate": 1.9598805306658305e-06,
+      "loss": 0.6862,
+      "step": 3948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4307311102620996,
+      "learning_rate": 1.959853436431381e-06,
+      "loss": 0.6947,
+      "step": 3949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7359321415145472,
+      "learning_rate": 1.959826333238534e-06,
+      "loss": 0.6751,
+      "step": 3950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.9723810924923737,
+      "learning_rate": 1.9597992210875437e-06,
+      "loss": 0.6823,
+      "step": 3951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.17342934726576,
+      "learning_rate": 1.9597720999786623e-06,
+      "loss": 0.6875,
+      "step": 3952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.028101116324916,
+      "learning_rate": 1.9597449699121427e-06,
+      "loss": 0.6784,
+      "step": 3953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8474883908780593,
+      "learning_rate": 1.9597178308882387e-06,
+      "loss": 0.6908,
+      "step": 3954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4615231514906086,
+      "learning_rate": 1.9596906829072027e-06,
+      "loss": 0.6829,
+      "step": 3955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.481818475765517,
+      "learning_rate": 1.9596635259692892e-06,
+      "loss": 0.6803,
+      "step": 3956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.095893466086206,
+      "learning_rate": 1.959636360074751e-06,
+      "loss": 0.6921,
+      "step": 3957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48145398992197186,
+      "learning_rate": 1.959609185223842e-06,
+      "loss": 0.6823,
+      "step": 3958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.699358222158069,
+      "learning_rate": 1.9595820014168157e-06,
+      "loss": 0.6829,
+      "step": 3959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.446082113708135,
+      "learning_rate": 1.9595548086539253e-06,
+      "loss": 0.6921,
+      "step": 3960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8202587940209765,
+      "learning_rate": 1.959527606935425e-06,
+      "loss": 0.696,
+      "step": 3961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.497576533677218,
+      "learning_rate": 1.959500396261569e-06,
+      "loss": 0.6803,
+      "step": 3962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1184523794685934,
+      "learning_rate": 1.959473176632611e-06,
+      "loss": 0.6803,
+      "step": 3963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.332206006103665,
+      "learning_rate": 1.959445948048805e-06,
+      "loss": 0.6862,
+      "step": 3964
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.68970349252278,
+      "learning_rate": 1.959418710510405e-06,
+      "loss": 0.694,
+      "step": 3965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8720707643990415,
+      "learning_rate": 1.959391464017665e-06,
+      "loss": 0.6875,
+      "step": 3966
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.5223639259689445,
+      "learning_rate": 1.95936420857084e-06,
+      "loss": 0.6927,
+      "step": 3967
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.210141192386821,
+      "learning_rate": 1.9593369441701835e-06,
+      "loss": 0.6973,
+      "step": 3968
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8251436365953215,
+      "learning_rate": 1.9593096708159512e-06,
+      "loss": 0.6901,
+      "step": 3969
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.952787767280889,
+      "learning_rate": 1.9592823885083965e-06,
+      "loss": 0.6849,
+      "step": 3970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47106175284083895,
+      "learning_rate": 1.959255097247774e-06,
+      "loss": 0.6855,
+      "step": 3971
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5801544550086577,
+      "learning_rate": 1.9592277970343397e-06,
+      "loss": 0.681,
+      "step": 3972
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.0360853326189705,
+      "learning_rate": 1.9592004878683477e-06,
+      "loss": 0.6771,
+      "step": 3973
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0331613824577497,
+      "learning_rate": 1.9591731697500517e-06,
+      "loss": 0.6706,
+      "step": 3974
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0880831188245446,
+      "learning_rate": 1.9591458426797084e-06,
+      "loss": 0.6908,
+      "step": 3975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6235330072136963,
+      "learning_rate": 1.959118506657572e-06,
+      "loss": 0.6934,
+      "step": 3976
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.50897127400615,
+      "learning_rate": 1.9590911616838974e-06,
+      "loss": 0.6901,
+      "step": 3977
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2132503647731894,
+      "learning_rate": 1.9590638077589404e-06,
+      "loss": 0.6816,
+      "step": 3978
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5797205435082124,
+      "learning_rate": 1.9590364448829563e-06,
+      "loss": 0.6842,
+      "step": 3979
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5154246648316986,
+      "learning_rate": 1.9590090730562e-06,
+      "loss": 0.6842,
+      "step": 3980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.696614340214914,
+      "learning_rate": 1.9589816922789274e-06,
+      "loss": 0.6921,
+      "step": 3981
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8819309561142313,
+      "learning_rate": 1.9589543025513933e-06,
+      "loss": 0.6738,
+      "step": 3982
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.8373643823969354,
+      "learning_rate": 1.9589269038738545e-06,
+      "loss": 0.6706,
+      "step": 3983
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.246328179295314,
+      "learning_rate": 1.958899496246566e-06,
+      "loss": 0.6868,
+      "step": 3984
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.026683009599949,
+      "learning_rate": 1.9588720796697837e-06,
+      "loss": 0.6712,
+      "step": 3985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.787480409276666,
+      "learning_rate": 1.9588446541437633e-06,
+      "loss": 0.679,
+      "step": 3986
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8466456183937954,
+      "learning_rate": 1.958817219668761e-06,
+      "loss": 0.6764,
+      "step": 3987
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9838579890845673,
+      "learning_rate": 1.958789776245033e-06,
+      "loss": 0.6738,
+      "step": 3988
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.935653188206649,
+      "learning_rate": 1.958762323872835e-06,
+      "loss": 0.6934,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.460340974416339,
+      "learning_rate": 1.9587348625524235e-06,
+      "loss": 0.7012,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.56482380113766,
+      "learning_rate": 1.9587073922840547e-06,
+      "loss": 0.7038,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.786009789012627,
+      "learning_rate": 1.9586799130679853e-06,
+      "loss": 0.6979,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6839484663140704,
+      "learning_rate": 1.958652424904471e-06,
+      "loss": 0.6615,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5774055451904492,
+      "learning_rate": 1.958624927793769e-06,
+      "loss": 0.6875,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3932997062958101,
+      "learning_rate": 1.958597421736136e-06,
+      "loss": 0.6927,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.303530239188593,
+      "learning_rate": 1.9585699067318285e-06,
+      "loss": 0.6914,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1386108801105643,
+      "learning_rate": 1.958542382781103e-06,
+      "loss": 0.6699,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.066223885972834,
+      "learning_rate": 1.9585148498842168e-06,
+      "loss": 0.6602,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9795628169799635,
+      "learning_rate": 1.9584873080414268e-06,
+      "loss": 0.6973,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.156355283906168,
+      "learning_rate": 1.95845975725299e-06,
+      "loss": 0.6855,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.616608917664683,
+      "learning_rate": 1.9584321975191634e-06,
+      "loss": 0.7057,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3174820690563822,
+      "learning_rate": 1.9584046288402046e-06,
+      "loss": 0.6582,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.7891821846453935,
+      "learning_rate": 1.9583770512163704e-06,
+      "loss": 0.696,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.412268070830071,
+      "learning_rate": 1.958349464647918e-06,
+      "loss": 0.6693,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0452410112915203,
+      "learning_rate": 1.9583218691351064e-06,
+      "loss": 0.696,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.895086306830729,
+      "learning_rate": 1.9582942646781912e-06,
+      "loss": 0.6927,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7135245958281526,
+      "learning_rate": 1.958266651277431e-06,
+      "loss": 0.6751,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4472277938719698,
+      "learning_rate": 1.9582390289330837e-06,
+      "loss": 0.6738,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1797451293710517,
+      "learning_rate": 1.958211397645407e-06,
+      "loss": 0.6784,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.572481716305568,
+      "learning_rate": 1.9581837574146578e-06,
+      "loss": 0.6699,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.014340972075616,
+      "learning_rate": 1.958156108241095e-06,
+      "loss": 0.6947,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.464016429637632,
+      "learning_rate": 1.958128450124977e-06,
+      "loss": 0.6693,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3502021434163196,
+      "learning_rate": 1.958100783066561e-06,
+      "loss": 0.6914,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5099893046547015,
+      "learning_rate": 1.9580731070661057e-06,
+      "loss": 0.6947,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.83322625940944,
+      "learning_rate": 1.9580454221238696e-06,
+      "loss": 0.6823,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.192420526147238,
+      "learning_rate": 1.958017728240111e-06,
+      "loss": 0.6751,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.90321364124129,
+      "learning_rate": 1.957990025415088e-06,
+      "loss": 0.6882,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6265197757381697,
+      "learning_rate": 1.9579623136490594e-06,
+      "loss": 0.6934,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.761812727392289,
+      "learning_rate": 1.9579345929422835e-06,
+      "loss": 0.6953,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.52918903678139,
+      "learning_rate": 1.95790686329502e-06,
+      "loss": 0.6764,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.141868885031518,
+      "learning_rate": 1.9578791247075266e-06,
+      "loss": 0.6764,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.496821379513251,
+      "learning_rate": 1.9578513771800623e-06,
+      "loss": 0.6862,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1557683368975713,
+      "learning_rate": 1.957823620712887e-06,
+      "loss": 0.7025,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.142439758743948,
+      "learning_rate": 1.9577958553062588e-06,
+      "loss": 0.7018,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.163085933674491,
+      "learning_rate": 1.9577680809604375e-06,
+      "loss": 0.6562,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7145914203334987,
+      "learning_rate": 1.9577402976756814e-06,
+      "loss": 0.6868,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.895840964447812,
+      "learning_rate": 1.957712505452251e-06,
+      "loss": 0.6784,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1594590330718444,
+      "learning_rate": 1.9576847042904047e-06,
+      "loss": 0.6667,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0794737136176846,
+      "learning_rate": 1.9576568941904024e-06,
+      "loss": 0.6842,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.004487958916995,
+      "learning_rate": 1.957629075152504e-06,
+      "loss": 0.7148,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.3398644028515125,
+      "learning_rate": 1.9576012471769686e-06,
+      "loss": 0.7012,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.4921513027963345,
+      "learning_rate": 1.9575734102640558e-06,
+      "loss": 0.6829,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.404839557672954,
+      "learning_rate": 1.957545564414026e-06,
+      "loss": 0.7064,
+      "step": 4033
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5573902984541856,
+      "learning_rate": 1.9575177096271386e-06,
+      "loss": 0.6999,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.938654157431562,
+      "learning_rate": 1.9574898459036535e-06,
+      "loss": 0.6849,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9264968961721327,
+      "learning_rate": 1.9574619732438315e-06,
+      "loss": 0.6699,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.9247590474089495,
+      "learning_rate": 1.957434091647932e-06,
+      "loss": 0.6823,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1660168307275596,
+      "learning_rate": 1.9574062011162155e-06,
+      "loss": 0.6947,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4541799612820976,
+      "learning_rate": 1.957378301648942e-06,
+      "loss": 0.6803,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6280357913784317,
+      "learning_rate": 1.9573503932463723e-06,
+      "loss": 0.694,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.599817030930131,
+      "learning_rate": 1.9573224759087666e-06,
+      "loss": 0.6901,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8656981710514706,
+      "learning_rate": 1.957294549636386e-06,
+      "loss": 0.6823,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.65739125697555,
+      "learning_rate": 1.95726661442949e-06,
+      "loss": 0.6823,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.615779307213479,
+      "learning_rate": 1.9572386702883406e-06,
+      "loss": 0.6921,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1292103854380584,
+      "learning_rate": 1.957210717213198e-06,
+      "loss": 0.7057,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9394376220672672,
+      "learning_rate": 1.957182755204323e-06,
+      "loss": 0.6836,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8294437181966376,
+      "learning_rate": 1.9571547842619767e-06,
+      "loss": 0.6667,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6863182660208595,
+      "learning_rate": 1.9571268043864204e-06,
+      "loss": 0.6829,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.1507089730724935,
+      "learning_rate": 1.9570988155779144e-06,
+      "loss": 0.6921,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.993811303083122,
+      "learning_rate": 1.957070817836721e-06,
+      "loss": 0.6816,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6062388042739855,
+      "learning_rate": 1.957042811163101e-06,
+      "loss": 0.6738,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.909260849054104,
+      "learning_rate": 1.9570147955573155e-06,
+      "loss": 0.6934,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.647787474408874,
+      "learning_rate": 1.9569867710196262e-06,
+      "loss": 0.6836,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6769760047530439,
+      "learning_rate": 1.956958737550295e-06,
+      "loss": 0.6842,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.445888198285777,
+      "learning_rate": 1.9569306951495834e-06,
+      "loss": 0.6797,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1654036358245947,
+      "learning_rate": 1.956902643817753e-06,
+      "loss": 0.696,
+      "step": 4056
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.176039156737428,
+      "learning_rate": 1.9568745835550654e-06,
+      "loss": 0.6855,
+      "step": 4057
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8835827114953932,
+      "learning_rate": 1.956846514361783e-06,
+      "loss": 0.6803,
+      "step": 4058
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.545362533324015,
+      "learning_rate": 1.956818436238167e-06,
+      "loss": 0.6914,
+      "step": 4059
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.707571300534269,
+      "learning_rate": 1.95679034918448e-06,
+      "loss": 0.6882,
+      "step": 4060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.023370447782207,
+      "learning_rate": 1.9567622532009842e-06,
+      "loss": 0.7005,
+      "step": 4061
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.622025922313558,
+      "learning_rate": 1.9567341482879417e-06,
+      "loss": 0.7161,
+      "step": 4062
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3582454804070343,
+      "learning_rate": 1.9567060344456146e-06,
+      "loss": 0.6699,
+      "step": 4063
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.05576808461557,
+      "learning_rate": 1.9566779116742656e-06,
+      "loss": 0.6849,
+      "step": 4064
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.802588523011091,
+      "learning_rate": 1.9566497799741572e-06,
+      "loss": 0.6686,
+      "step": 4065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.044516525304923,
+      "learning_rate": 1.956621639345552e-06,
+      "loss": 0.6784,
+      "step": 4066
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2919354596622716,
+      "learning_rate": 1.956593489788712e-06,
+      "loss": 0.6732,
+      "step": 4067
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1860309055783156,
+      "learning_rate": 1.9565653313039008e-06,
+      "loss": 0.6686,
+      "step": 4068
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0705129014451185,
+      "learning_rate": 1.9565371638913803e-06,
+      "loss": 0.6797,
+      "step": 4069
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6001056723213165,
+      "learning_rate": 1.9565089875514137e-06,
+      "loss": 0.696,
+      "step": 4070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.232888759091858,
+      "learning_rate": 1.9564808022842645e-06,
+      "loss": 0.6862,
+      "step": 4071
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.53416863475318,
+      "learning_rate": 1.9564526080901957e-06,
+      "loss": 0.7057,
+      "step": 4072
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.407478958887545,
+      "learning_rate": 1.9564244049694695e-06,
+      "loss": 0.6758,
+      "step": 4073
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8395975075415735,
+      "learning_rate": 1.95639619292235e-06,
+      "loss": 0.6823,
+      "step": 4074
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9959129523698793,
+      "learning_rate": 1.9563679719491004e-06,
+      "loss": 0.6784,
+      "step": 4075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6186639873051004,
+      "learning_rate": 1.9563397420499836e-06,
+      "loss": 0.666,
+      "step": 4076
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.325254652146319,
+      "learning_rate": 1.9563115032252635e-06,
+      "loss": 0.6914,
+      "step": 4077
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.986802091238508,
+      "learning_rate": 1.956283255475204e-06,
+      "loss": 0.6764,
+      "step": 4078
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6346504051153264,
+      "learning_rate": 1.9562549988000678e-06,
+      "loss": 0.6784,
+      "step": 4079
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.714287781105607,
+      "learning_rate": 1.956226733200119e-06,
+      "loss": 0.6868,
+      "step": 4080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.320011142950522,
+      "learning_rate": 1.956198458675622e-06,
+      "loss": 0.696,
+      "step": 4081
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.85701226962054,
+      "learning_rate": 1.9561701752268404e-06,
+      "loss": 0.6654,
+      "step": 4082
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.371641274754898,
+      "learning_rate": 1.9561418828540374e-06,
+      "loss": 0.6999,
+      "step": 4083
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7646446542977416,
+      "learning_rate": 1.956113581557478e-06,
+      "loss": 0.6875,
+      "step": 4084
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6472112592791361,
+      "learning_rate": 1.956085271337426e-06,
+      "loss": 0.6823,
+      "step": 4085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.83600242593057,
+      "learning_rate": 1.9560569521941454e-06,
+      "loss": 0.6842,
+      "step": 4086
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.964624928329822,
+      "learning_rate": 1.9560286241279008e-06,
+      "loss": 0.7279,
+      "step": 4087
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.207567676620581,
+      "learning_rate": 1.9560002871389565e-06,
+      "loss": 0.6999,
+      "step": 4088
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.739990399510825,
+      "learning_rate": 1.955971941227577e-06,
+      "loss": 0.6764,
+      "step": 4089
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.021430939342034,
+      "learning_rate": 1.955943586394027e-06,
+      "loss": 0.6895,
+      "step": 4090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7897120484412785,
+      "learning_rate": 1.9559152226385707e-06,
+      "loss": 0.6777,
+      "step": 4091
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7126435579093503,
+      "learning_rate": 1.9558868499614733e-06,
+      "loss": 0.6934,
+      "step": 4092
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.207362716406132,
+      "learning_rate": 1.9558584683629992e-06,
+      "loss": 0.6628,
+      "step": 4093
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3618056747164842,
+      "learning_rate": 1.9558300778434137e-06,
+      "loss": 0.6823,
+      "step": 4094
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6256646483694887,
+      "learning_rate": 1.9558016784029814e-06,
+      "loss": 0.6706,
+      "step": 4095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.000378567130769,
+      "learning_rate": 1.9557732700419677e-06,
+      "loss": 0.6738,
+      "step": 4096
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.444108857460889,
+      "learning_rate": 1.955744852760637e-06,
+      "loss": 0.6855,
+      "step": 4097
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.497843608850112,
+      "learning_rate": 1.9557164265592555e-06,
+      "loss": 0.6751,
+      "step": 4098
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.1969008282766405,
+      "learning_rate": 1.9556879914380883e-06,
+      "loss": 0.6771,
+      "step": 4099
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7060918600121062,
+      "learning_rate": 1.9556595473974e-06,
+      "loss": 0.6712,
+      "step": 4100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.100710352774798,
+      "learning_rate": 1.9556310944374573e-06,
+      "loss": 0.6908,
+      "step": 4101
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8267797665845775,
+      "learning_rate": 1.9556026325585244e-06,
+      "loss": 0.724,
+      "step": 4102
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9916786860139205,
+      "learning_rate": 1.9555741617608682e-06,
+      "loss": 0.679,
+      "step": 4103
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6988267948305379,
+      "learning_rate": 1.9555456820447535e-06,
+      "loss": 0.6888,
+      "step": 4104
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8748325617345508,
+      "learning_rate": 1.9555171934104465e-06,
+      "loss": 0.7038,
+      "step": 4105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.273041168778825,
+      "learning_rate": 1.955488695858213e-06,
+      "loss": 0.6777,
+      "step": 4106
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1385330700828,
+      "learning_rate": 1.955460189388319e-06,
+      "loss": 0.6719,
+      "step": 4107
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.471802393442134,
+      "learning_rate": 1.9554316740010303e-06,
+      "loss": 0.6914,
+      "step": 4108
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.60225655354363,
+      "learning_rate": 1.9554031496966137e-06,
+      "loss": 0.6966,
+      "step": 4109
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.880698833904422,
+      "learning_rate": 1.955374616475335e-06,
+      "loss": 0.6621,
+      "step": 4110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1981205891515465,
+      "learning_rate": 1.9553460743374607e-06,
+      "loss": 0.6868,
+      "step": 4111
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4394497775676525,
+      "learning_rate": 1.9553175232832566e-06,
+      "loss": 0.7083,
+      "step": 4112
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4563103330701213,
+      "learning_rate": 1.9552889633129897e-06,
+      "loss": 0.6882,
+      "step": 4113
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8958873668022056,
+      "learning_rate": 1.955260394426926e-06,
+      "loss": 0.6836,
+      "step": 4114
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.288636772308869,
+      "learning_rate": 1.9552318166253332e-06,
+      "loss": 0.6842,
+      "step": 4115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7377629867514531,
+      "learning_rate": 1.9552032299084772e-06,
+      "loss": 0.6953,
+      "step": 4116
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9824247007454538,
+      "learning_rate": 1.955174634276625e-06,
+      "loss": 0.6868,
+      "step": 4117
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33270897654326337,
+      "learning_rate": 1.9551460297300434e-06,
+      "loss": 0.6842,
+      "step": 4118
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.206620515780319,
+      "learning_rate": 1.9551174162689992e-06,
+      "loss": 0.681,
+      "step": 4119
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3663717289747557,
+      "learning_rate": 1.95508879389376e-06,
+      "loss": 0.6934,
+      "step": 4120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.37055828989072204,
+      "learning_rate": 1.9550601626045925e-06,
+      "loss": 0.6855,
+      "step": 4121
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1188731659655793,
+      "learning_rate": 1.9550315224017644e-06,
+      "loss": 0.6888,
+      "step": 4122
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.965669399036667,
+      "learning_rate": 1.955002873285542e-06,
+      "loss": 0.6992,
+      "step": 4123
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.371014056858924,
+      "learning_rate": 1.9549742152561937e-06,
+      "loss": 0.709,
+      "step": 4124
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.39357063278099597,
+      "learning_rate": 1.954945548313987e-06,
+      "loss": 0.6829,
+      "step": 4125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.771296444108307,
+      "learning_rate": 1.9549168724591883e-06,
+      "loss": 0.6882,
+      "step": 4126
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7696142205659247,
+      "learning_rate": 1.9548881876920663e-06,
+      "loss": 0.6745,
+      "step": 4127
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8356061137736008,
+      "learning_rate": 1.9548594940128884e-06,
+      "loss": 0.6868,
+      "step": 4128
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1730499246704151,
+      "learning_rate": 1.9548307914219223e-06,
+      "loss": 0.6862,
+      "step": 4129
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5062654750385044,
+      "learning_rate": 1.9548020799194364e-06,
+      "loss": 0.6953,
+      "step": 4130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.2937594168301185,
+      "learning_rate": 1.9547733595056977e-06,
+      "loss": 0.7012,
+      "step": 4131
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.863578859767148,
+      "learning_rate": 1.954744630180975e-06,
+      "loss": 0.6947,
+      "step": 4132
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8225237834697268,
+      "learning_rate": 1.9547158919455364e-06,
+      "loss": 0.6758,
+      "step": 4133
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7607928872730487,
+      "learning_rate": 1.9546871447996496e-06,
+      "loss": 0.6895,
+      "step": 4134
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5869722276612857,
+      "learning_rate": 1.9546583887435835e-06,
+      "loss": 0.6947,
+      "step": 4135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3321112713247905,
+      "learning_rate": 1.954629623777606e-06,
+      "loss": 0.6947,
+      "step": 4136
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7952671108436595,
+      "learning_rate": 1.9546008499019862e-06,
+      "loss": 0.6751,
+      "step": 4137
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6342282794371352,
+      "learning_rate": 1.954572067116992e-06,
+      "loss": 0.679,
+      "step": 4138
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1927859685353313,
+      "learning_rate": 1.9545432754228923e-06,
+      "loss": 0.6992,
+      "step": 4139
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4323577460198612,
+      "learning_rate": 1.9545144748199557e-06,
+      "loss": 0.6836,
+      "step": 4140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7256016321269123,
+      "learning_rate": 1.9544856653084514e-06,
+      "loss": 0.6842,
+      "step": 4141
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4373553795964231,
+      "learning_rate": 1.9544568468886477e-06,
+      "loss": 0.6797,
+      "step": 4142
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6636642458744444,
+      "learning_rate": 1.9544280195608137e-06,
+      "loss": 0.6777,
+      "step": 4143
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.314647349541706,
+      "learning_rate": 1.954399183325219e-06,
+      "loss": 0.6686,
+      "step": 4144
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.958120659583271,
+      "learning_rate": 1.954370338182132e-06,
+      "loss": 0.6986,
+      "step": 4145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4028061784981056,
+      "learning_rate": 1.9543414841318223e-06,
+      "loss": 0.694,
+      "step": 4146
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0193035869633762,
+      "learning_rate": 1.954312621174559e-06,
+      "loss": 0.6751,
+      "step": 4147
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7613367163371814,
+      "learning_rate": 1.954283749310612e-06,
+      "loss": 0.6771,
+      "step": 4148
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.789599544089084,
+      "learning_rate": 1.95425486854025e-06,
+      "loss": 0.6868,
+      "step": 4149
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7952586511569316,
+      "learning_rate": 1.9542259788637435e-06,
+      "loss": 0.6615,
+      "step": 4150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0726554058488729,
+      "learning_rate": 1.954197080281361e-06,
+      "loss": 0.6699,
+      "step": 4151
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.421653641474186,
+      "learning_rate": 1.9541681727933726e-06,
+      "loss": 0.6764,
+      "step": 4152
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0985999135880866,
+      "learning_rate": 1.9541392564000487e-06,
+      "loss": 0.696,
+      "step": 4153
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.233185013152147,
+      "learning_rate": 1.9541103311016587e-06,
+      "loss": 0.696,
+      "step": 4154
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2588425071331848,
+      "learning_rate": 1.9540813968984727e-06,
+      "loss": 0.681,
+      "step": 4155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.076989831941326,
+      "learning_rate": 1.95405245379076e-06,
+      "loss": 0.6888,
+      "step": 4156
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1664806921232618,
+      "learning_rate": 1.954023501778792e-06,
+      "loss": 0.6803,
+      "step": 4157
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.097725117590515,
+      "learning_rate": 1.953994540862838e-06,
+      "loss": 0.6953,
+      "step": 4158
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.352728659664546,
+      "learning_rate": 1.953965571043169e-06,
+      "loss": 0.6719,
+      "step": 4159
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6753098493358649,
+      "learning_rate": 1.9539365923200546e-06,
+      "loss": 0.6927,
+      "step": 4160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2304742896238494,
+      "learning_rate": 1.953907604693766e-06,
+      "loss": 0.6882,
+      "step": 4161
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.204827068015192,
+      "learning_rate": 1.953878608164573e-06,
+      "loss": 0.6953,
+      "step": 4162
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.097934337331125,
+      "learning_rate": 1.953849602732746e-06,
+      "loss": 0.7077,
+      "step": 4163
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9781291020207437,
+      "learning_rate": 1.9538205883985573e-06,
+      "loss": 0.6719,
+      "step": 4164
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.182942485114819,
+      "learning_rate": 1.9537915651622763e-06,
+      "loss": 0.6647,
+      "step": 4165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3307320409059273,
+      "learning_rate": 1.953762533024174e-06,
+      "loss": 0.7005,
+      "step": 4166
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5592081163328801,
+      "learning_rate": 1.953733491984522e-06,
+      "loss": 0.6816,
+      "step": 4167
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.069126444951943,
+      "learning_rate": 1.953704442043591e-06,
+      "loss": 0.6979,
+      "step": 4168
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3536039385477654,
+      "learning_rate": 1.953675383201652e-06,
+      "loss": 0.6829,
+      "step": 4169
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6492139097338323,
+      "learning_rate": 1.9536463154589764e-06,
+      "loss": 0.6602,
+      "step": 4170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.501588515828579,
+      "learning_rate": 1.9536172388158358e-06,
+      "loss": 0.6999,
+      "step": 4171
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5745831126512009,
+      "learning_rate": 1.9535881532725006e-06,
+      "loss": 0.6719,
+      "step": 4172
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.2532796874893695,
+      "learning_rate": 1.953559058829243e-06,
+      "loss": 0.679,
+      "step": 4173
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.263019057028312,
+      "learning_rate": 1.953529955486334e-06,
+      "loss": 0.679,
+      "step": 4174
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2060340306904744,
+      "learning_rate": 1.953500843244046e-06,
+      "loss": 0.6803,
+      "step": 4175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.037520951983313,
+      "learning_rate": 1.9534717221026504e-06,
+      "loss": 0.6777,
+      "step": 4176
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5349442380853524,
+      "learning_rate": 1.953442592062419e-06,
+      "loss": 0.6888,
+      "step": 4177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.059904043121049,
+      "learning_rate": 1.953413453123623e-06,
+      "loss": 0.6888,
+      "step": 4178
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.9625035810334825,
+      "learning_rate": 1.9533843052865355e-06,
+      "loss": 0.6628,
+      "step": 4179
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1875511784796746,
+      "learning_rate": 1.953355148551427e-06,
+      "loss": 0.6803,
+      "step": 4180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7767569224966479,
+      "learning_rate": 1.9533259829185717e-06,
+      "loss": 0.6888,
+      "step": 4181
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6620502875382707,
+      "learning_rate": 1.9532968083882404e-06,
+      "loss": 0.6875,
+      "step": 4182
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.630677885057153,
+      "learning_rate": 1.9532676249607055e-06,
+      "loss": 0.6999,
+      "step": 4183
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3271571628425898,
+      "learning_rate": 1.9532384326362393e-06,
+      "loss": 0.6543,
+      "step": 4184
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5869078419518403,
+      "learning_rate": 1.953209231415115e-06,
+      "loss": 0.6953,
+      "step": 4185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4746859434770783,
+      "learning_rate": 1.9531800212976046e-06,
+      "loss": 0.6647,
+      "step": 4186
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6278224802940695,
+      "learning_rate": 1.9531508022839803e-06,
+      "loss": 0.6901,
+      "step": 4187
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.029294787302973,
+      "learning_rate": 1.9531215743745157e-06,
+      "loss": 0.6836,
+      "step": 4188
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2995445680001887,
+      "learning_rate": 1.953092337569483e-06,
+      "loss": 0.6868,
+      "step": 4189
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3348678564195486,
+      "learning_rate": 1.9530630918691554e-06,
+      "loss": 0.6569,
+      "step": 4190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6771494539552827,
+      "learning_rate": 1.953033837273806e-06,
+      "loss": 0.6738,
+      "step": 4191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.2115554499909456,
+      "learning_rate": 1.953004573783707e-06,
+      "loss": 0.6823,
+      "step": 4192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3577845221855702,
+      "learning_rate": 1.9529753013991315e-06,
+      "loss": 0.694,
+      "step": 4193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5538696066719634,
+      "learning_rate": 1.952946020120354e-06,
+      "loss": 0.6979,
+      "step": 4194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6036596213420913,
+      "learning_rate": 1.9529167299476473e-06,
+      "loss": 0.696,
+      "step": 4195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1044746158946455,
+      "learning_rate": 1.952887430881284e-06,
+      "loss": 0.6725,
+      "step": 4196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3262003781287484,
+      "learning_rate": 1.952858122921538e-06,
+      "loss": 0.696,
+      "step": 4197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1232938500682852,
+      "learning_rate": 1.9528288060686828e-06,
+      "loss": 0.6914,
+      "step": 4198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.367097872435004,
+      "learning_rate": 1.9527994803229923e-06,
+      "loss": 0.6868,
+      "step": 4199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.89099790012284,
+      "learning_rate": 1.95277014568474e-06,
+      "loss": 0.6829,
+      "step": 4200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.569110182526372,
+      "learning_rate": 1.9527408021541998e-06,
+      "loss": 0.6875,
+      "step": 4201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6137793974212262,
+      "learning_rate": 1.952711449731645e-06,
+      "loss": 0.679,
+      "step": 4202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.842952597803891,
+      "learning_rate": 1.9526820884173503e-06,
+      "loss": 0.6934,
+      "step": 4203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.651999818308638,
+      "learning_rate": 1.9526527182115894e-06,
+      "loss": 0.6973,
+      "step": 4204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.491120958266712,
+      "learning_rate": 1.9526233391146364e-06,
+      "loss": 0.6895,
+      "step": 4205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5344651143132962,
+      "learning_rate": 1.9525939511267656e-06,
+      "loss": 0.6908,
+      "step": 4206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.108809957656665,
+      "learning_rate": 1.952564554248251e-06,
+      "loss": 0.7031,
+      "step": 4207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.762282069038684,
+      "learning_rate": 1.9525351484793672e-06,
+      "loss": 0.6992,
+      "step": 4208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8149851562075997,
+      "learning_rate": 1.9525057338203886e-06,
+      "loss": 0.6803,
+      "step": 4209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.24451328240114,
+      "learning_rate": 1.95247631027159e-06,
+      "loss": 0.6758,
+      "step": 4210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7875648647998053,
+      "learning_rate": 1.9524468778332456e-06,
+      "loss": 0.7031,
+      "step": 4211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5473316313109122,
+      "learning_rate": 1.9524174365056303e-06,
+      "loss": 0.6803,
+      "step": 4212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.4017270566081255,
+      "learning_rate": 1.952387986289019e-06,
+      "loss": 0.6797,
+      "step": 4213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5022385189284118,
+      "learning_rate": 1.952358527183686e-06,
+      "loss": 0.694,
+      "step": 4214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.1407055450349715,
+      "learning_rate": 1.952329059189907e-06,
+      "loss": 0.6732,
+      "step": 4215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5096519130289492,
+      "learning_rate": 1.9522995823079565e-06,
+      "loss": 0.6888,
+      "step": 4216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.3971554366135015,
+      "learning_rate": 1.9522700965381097e-06,
+      "loss": 0.6999,
+      "step": 4217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2900704012049786,
+      "learning_rate": 1.9522406018806423e-06,
+      "loss": 0.707,
+      "step": 4218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7502533806694988,
+      "learning_rate": 1.952211098335829e-06,
+      "loss": 0.6849,
+      "step": 4219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8532153567730596,
+      "learning_rate": 1.952181585903945e-06,
+      "loss": 0.681,
+      "step": 4220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.797489761589331,
+      "learning_rate": 1.952152064585266e-06,
+      "loss": 0.6732,
+      "step": 4221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4033773467514252,
+      "learning_rate": 1.952122534380068e-06,
+      "loss": 0.6829,
+      "step": 4222
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6956200073348711,
+      "learning_rate": 1.952092995288626e-06,
+      "loss": 0.6868,
+      "step": 4223
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.056607873959254,
+      "learning_rate": 1.952063447311216e-06,
+      "loss": 0.6888,
+      "step": 4224
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1266547858163816,
+      "learning_rate": 1.9520338904481135e-06,
+      "loss": 0.6628,
+      "step": 4225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0757548653500533,
+      "learning_rate": 1.9520043246995943e-06,
+      "loss": 0.6875,
+      "step": 4226
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.493999202167379,
+      "learning_rate": 1.951974750065935e-06,
+      "loss": 0.6849,
+      "step": 4227
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.615832187022963,
+      "learning_rate": 1.9519451665474108e-06,
+      "loss": 0.6868,
+      "step": 4228
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6395909485293386,
+      "learning_rate": 1.9519155741442983e-06,
+      "loss": 0.668,
+      "step": 4229
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.878014997303444,
+      "learning_rate": 1.9518859728568736e-06,
+      "loss": 0.6953,
+      "step": 4230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8150771902779476,
+      "learning_rate": 1.951856362685413e-06,
+      "loss": 0.6953,
+      "step": 4231
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5658059223610259,
+      "learning_rate": 1.9518267436301923e-06,
+      "loss": 0.6862,
+      "step": 4232
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5131029539483583,
+      "learning_rate": 1.9517971156914892e-06,
+      "loss": 0.6719,
+      "step": 4233
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.29689791621872,
+      "learning_rate": 1.951767478869579e-06,
+      "loss": 0.6706,
+      "step": 4234
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.859550788658977,
+      "learning_rate": 1.951737833164739e-06,
+      "loss": 0.6914,
+      "step": 4235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9930812360316752,
+      "learning_rate": 1.9517081785772454e-06,
+      "loss": 0.6836,
+      "step": 4236
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6470354263144658,
+      "learning_rate": 1.951678515107375e-06,
+      "loss": 0.6797,
+      "step": 4237
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.496471279670719,
+      "learning_rate": 1.951648842755405e-06,
+      "loss": 0.6947,
+      "step": 4238
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3761182373226597,
+      "learning_rate": 1.9516191615216122e-06,
+      "loss": 0.6745,
+      "step": 4239
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.178042417691724,
+      "learning_rate": 1.951589471406274e-06,
+      "loss": 0.6738,
+      "step": 4240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.900531376684805,
+      "learning_rate": 1.951559772409667e-06,
+      "loss": 0.6901,
+      "step": 4241
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8350422135642965,
+      "learning_rate": 1.951530064532068e-06,
+      "loss": 0.6647,
+      "step": 4242
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.216153854702794,
+      "learning_rate": 1.9515003477737552e-06,
+      "loss": 0.6751,
+      "step": 4243
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3312152029019744,
+      "learning_rate": 1.9514706221350054e-06,
+      "loss": 0.6816,
+      "step": 4244
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.763801843815841,
+      "learning_rate": 1.951440887616096e-06,
+      "loss": 0.6758,
+      "step": 4245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1947618432423197,
+      "learning_rate": 1.951411144217305e-06,
+      "loss": 0.6953,
+      "step": 4246
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8276324411985485,
+      "learning_rate": 1.9513813919389094e-06,
+      "loss": 0.7044,
+      "step": 4247
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1362251051128947,
+      "learning_rate": 1.951351630781187e-06,
+      "loss": 0.7005,
+      "step": 4248
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6518930202249202,
+      "learning_rate": 1.951321860744416e-06,
+      "loss": 0.7018,
+      "step": 4249
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2967724380389973,
+      "learning_rate": 1.9512920818288736e-06,
+      "loss": 0.6758,
+      "step": 4250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4184246520790906,
+      "learning_rate": 1.9512622940348385e-06,
+      "loss": 0.6784,
+      "step": 4251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.1591135079325599,
+      "learning_rate": 1.951232497362588e-06,
+      "loss": 0.6953,
+      "step": 4252
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.22689536997153,
+      "learning_rate": 1.9512026918124005e-06,
+      "loss": 0.6953,
+      "step": 4253
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5943987232104725,
+      "learning_rate": 1.9511728773845545e-06,
+      "loss": 0.6842,
+      "step": 4254
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.9414894759686177,
+      "learning_rate": 1.951143054079327e-06,
+      "loss": 0.6764,
+      "step": 4255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.792655612533685,
+      "learning_rate": 1.9511132218969984e-06,
+      "loss": 0.696,
+      "step": 4256
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.988067031721738,
+      "learning_rate": 1.9510833808378453e-06,
+      "loss": 0.6882,
+      "step": 4257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.281325675666828,
+      "learning_rate": 1.951053530902147e-06,
+      "loss": 0.6823,
+      "step": 4258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.292294698704286,
+      "learning_rate": 1.9510236720901823e-06,
+      "loss": 0.6882,
+      "step": 4259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7791734572008657,
+      "learning_rate": 1.9509938044022293e-06,
+      "loss": 0.6816,
+      "step": 4260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.588255214465422,
+      "learning_rate": 1.950963927838567e-06,
+      "loss": 0.7018,
+      "step": 4261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8209263828135651,
+      "learning_rate": 1.950934042399474e-06,
+      "loss": 0.681,
+      "step": 4262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.430803174304989,
+      "learning_rate": 1.9509041480852298e-06,
+      "loss": 0.6712,
+      "step": 4263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7466610499611543,
+      "learning_rate": 1.9508742448961134e-06,
+      "loss": 0.6706,
+      "step": 4264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.753661906305285,
+      "learning_rate": 1.9508443328324033e-06,
+      "loss": 0.6628,
+      "step": 4265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.1517968950312667,
+      "learning_rate": 1.9508144118943786e-06,
+      "loss": 0.6901,
+      "step": 4266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.627687314713695,
+      "learning_rate": 1.9507844820823196e-06,
+      "loss": 0.6868,
+      "step": 4267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8751765480975096,
+      "learning_rate": 1.950754543396504e-06,
+      "loss": 0.6855,
+      "step": 4268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.241915554717159,
+      "learning_rate": 1.950724595837213e-06,
+      "loss": 0.6725,
+      "step": 4269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8847298841668723,
+      "learning_rate": 1.950694639404725e-06,
+      "loss": 0.6842,
+      "step": 4270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.242748178616972,
+      "learning_rate": 1.95066467409932e-06,
+      "loss": 0.6706,
+      "step": 4271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3804360737743244,
+      "learning_rate": 1.9506346999212773e-06,
+      "loss": 0.7005,
+      "step": 4272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3987439011334675,
+      "learning_rate": 1.950604716870877e-06,
+      "loss": 0.6699,
+      "step": 4273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.01586975444677,
+      "learning_rate": 1.9505747249483985e-06,
+      "loss": 0.6882,
+      "step": 4274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2110924533431375,
+      "learning_rate": 1.9505447241541225e-06,
+      "loss": 0.6816,
+      "step": 4275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6177476342523456,
+      "learning_rate": 1.950514714488328e-06,
+      "loss": 0.6764,
+      "step": 4276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4337799901849095,
+      "learning_rate": 1.950484695951296e-06,
+      "loss": 0.7005,
+      "step": 4277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.92464401649624,
+      "learning_rate": 1.9504546685433065e-06,
+      "loss": 0.6999,
+      "step": 4278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.713926938065606,
+      "learning_rate": 1.9504246322646393e-06,
+      "loss": 0.7077,
+      "step": 4279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.967359792617599,
+      "learning_rate": 1.950394587115575e-06,
+      "loss": 0.6914,
+      "step": 4280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4890057304549755,
+      "learning_rate": 1.9503645330963936e-06,
+      "loss": 0.6725,
+      "step": 4281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4012441099220403,
+      "learning_rate": 1.950334470207376e-06,
+      "loss": 0.6784,
+      "step": 4282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2047922653123764,
+      "learning_rate": 1.9503043984488032e-06,
+      "loss": 0.6953,
+      "step": 4283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.930934544421041,
+      "learning_rate": 1.950274317820955e-06,
+      "loss": 0.694,
+      "step": 4284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.878269287197783,
+      "learning_rate": 1.950244228324113e-06,
+      "loss": 0.6862,
+      "step": 4285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.031999151454824,
+      "learning_rate": 1.9502141299585574e-06,
+      "loss": 0.6934,
+      "step": 4286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.868490087082502,
+      "learning_rate": 1.9501840227245688e-06,
+      "loss": 0.6986,
+      "step": 4287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.536906763276398,
+      "learning_rate": 1.9501539066224293e-06,
+      "loss": 0.6888,
+      "step": 4288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5276224796173875,
+      "learning_rate": 1.950123781652419e-06,
+      "loss": 0.6895,
+      "step": 4289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.075286948627591,
+      "learning_rate": 1.95009364781482e-06,
+      "loss": 0.6686,
+      "step": 4290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.168320125447261,
+      "learning_rate": 1.9500635051099125e-06,
+      "loss": 0.6745,
+      "step": 4291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.574887066778595,
+      "learning_rate": 1.9500333535379783e-06,
+      "loss": 0.6693,
+      "step": 4292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9456125014092531,
+      "learning_rate": 1.9500031930992985e-06,
+      "loss": 0.6979,
+      "step": 4293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7558959573554094,
+      "learning_rate": 1.9499730237941553e-06,
+      "loss": 0.6992,
+      "step": 4294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7659452347343915,
+      "learning_rate": 1.94994284562283e-06,
+      "loss": 0.6927,
+      "step": 4295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.6122501946205325,
+      "learning_rate": 1.9499126585856036e-06,
+      "loss": 0.6992,
+      "step": 4296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.051910683049865,
+      "learning_rate": 1.949882462682759e-06,
+      "loss": 0.6914,
+      "step": 4297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.089321703464452,
+      "learning_rate": 1.9498522579145767e-06,
+      "loss": 0.6784,
+      "step": 4298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.050157378339197,
+      "learning_rate": 1.9498220442813397e-06,
+      "loss": 0.6862,
+      "step": 4299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.356213331073814,
+      "learning_rate": 1.9497918217833295e-06,
+      "loss": 0.6914,
+      "step": 4300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4529404746195979,
+      "learning_rate": 1.949761590420828e-06,
+      "loss": 0.6966,
+      "step": 4301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5083628881595825,
+      "learning_rate": 1.949731350194118e-06,
+      "loss": 0.6784,
+      "step": 4302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.877819051702648,
+      "learning_rate": 1.9497011011034806e-06,
+      "loss": 0.6849,
+      "step": 4303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.079601677664231,
+      "learning_rate": 1.9496708431491993e-06,
+      "loss": 0.6829,
+      "step": 4304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7089216811009107,
+      "learning_rate": 1.9496405763315564e-06,
+      "loss": 0.6953,
+      "step": 4305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2872231477241165,
+      "learning_rate": 1.9496103006508335e-06,
+      "loss": 0.681,
+      "step": 4306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4574389038013815,
+      "learning_rate": 1.949580016107314e-06,
+      "loss": 0.681,
+      "step": 4307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.793967237736386,
+      "learning_rate": 1.94954972270128e-06,
+      "loss": 0.6686,
+      "step": 4308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0809151738695877,
+      "learning_rate": 1.9495194204330145e-06,
+      "loss": 0.6647,
+      "step": 4309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.309023910451725,
+      "learning_rate": 1.9494891093028002e-06,
+      "loss": 0.6738,
+      "step": 4310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6984581433107327,
+      "learning_rate": 1.9494587893109204e-06,
+      "loss": 0.6823,
+      "step": 4311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.19293156032089,
+      "learning_rate": 1.9494284604576572e-06,
+      "loss": 0.7109,
+      "step": 4312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0359375095252503,
+      "learning_rate": 1.949398122743295e-06,
+      "loss": 0.6849,
+      "step": 4313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6782089182816349,
+      "learning_rate": 1.9493677761681156e-06,
+      "loss": 0.6849,
+      "step": 4314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.079978014081709,
+      "learning_rate": 1.949337420732403e-06,
+      "loss": 0.6973,
+      "step": 4315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.870849983594082,
+      "learning_rate": 1.9493070564364404e-06,
+      "loss": 0.679,
+      "step": 4316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6735773472121735,
+      "learning_rate": 1.9492766832805106e-06,
+      "loss": 0.6836,
+      "step": 4317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.938652164543074,
+      "learning_rate": 1.9492463012648977e-06,
+      "loss": 0.6992,
+      "step": 4318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7865246755938715,
+      "learning_rate": 1.9492159103898857e-06,
+      "loss": 0.6602,
+      "step": 4319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5105029439294169,
+      "learning_rate": 1.949185510655757e-06,
+      "loss": 0.6686,
+      "step": 4320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7904261433842974,
+      "learning_rate": 1.9491551020627964e-06,
+      "loss": 0.6921,
+      "step": 4321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.2514237794542575,
+      "learning_rate": 1.949124684611287e-06,
+      "loss": 0.6842,
+      "step": 4322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.325815613504528,
+      "learning_rate": 1.949094258301513e-06,
+      "loss": 0.6582,
+      "step": 4323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.267815698957933,
+      "learning_rate": 1.9490638231337586e-06,
+      "loss": 0.7122,
+      "step": 4324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.234217114717781,
+      "learning_rate": 1.9490333791083075e-06,
+      "loss": 0.6927,
+      "step": 4325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0771540441453915,
+      "learning_rate": 1.949002926225444e-06,
+      "loss": 0.6953,
+      "step": 4326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1471262746690676,
+      "learning_rate": 1.9489724644854523e-06,
+      "loss": 0.6914,
+      "step": 4327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0249691638698506,
+      "learning_rate": 1.9489419938886166e-06,
+      "loss": 0.679,
+      "step": 4328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4648978351805138,
+      "learning_rate": 1.9489115144352213e-06,
+      "loss": 0.6966,
+      "step": 4329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9727681920127664,
+      "learning_rate": 1.9488810261255507e-06,
+      "loss": 0.6777,
+      "step": 4330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1087194130833946,
+      "learning_rate": 1.9488505289598897e-06,
+      "loss": 0.6641,
+      "step": 4331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9033568688307647,
+      "learning_rate": 1.948820022938523e-06,
+      "loss": 0.6849,
+      "step": 4332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3286449572478056,
+      "learning_rate": 1.948789508061735e-06,
+      "loss": 0.6693,
+      "step": 4333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7580054100779687,
+      "learning_rate": 1.9487589843298107e-06,
+      "loss": 0.6849,
+      "step": 4334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.900332060570873,
+      "learning_rate": 1.9487284517430347e-06,
+      "loss": 0.7057,
+      "step": 4335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6716299347900863,
+      "learning_rate": 1.9486979103016923e-06,
+      "loss": 0.7057,
+      "step": 4336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4048284874156887,
+      "learning_rate": 1.9486673600060684e-06,
+      "loss": 0.6758,
+      "step": 4337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.6307656768430645,
+      "learning_rate": 1.948636800856448e-06,
+      "loss": 0.6868,
+      "step": 4338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.587420095198285,
+      "learning_rate": 1.9486062328531168e-06,
+      "loss": 0.7005,
+      "step": 4339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.777053382143625,
+      "learning_rate": 1.9485756559963595e-06,
+      "loss": 0.6764,
+      "step": 4340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3982073176858623,
+      "learning_rate": 1.9485450702864617e-06,
+      "loss": 0.6842,
+      "step": 4341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7308977851650629,
+      "learning_rate": 1.948514475723709e-06,
+      "loss": 0.6712,
+      "step": 4342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7333472808400587,
+      "learning_rate": 1.948483872308387e-06,
+      "loss": 0.679,
+      "step": 4343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2486683087005188,
+      "learning_rate": 1.9484532600407807e-06,
+      "loss": 0.6895,
+      "step": 4344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5921630853592034,
+      "learning_rate": 1.9484226389211763e-06,
+      "loss": 0.6777,
+      "step": 4345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.326334511880046,
+      "learning_rate": 1.9483920089498594e-06,
+      "loss": 0.6836,
+      "step": 4346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4165419184286665,
+      "learning_rate": 1.948361370127116e-06,
+      "loss": 0.6803,
+      "step": 4347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3751323085217537,
+      "learning_rate": 1.9483307224532324e-06,
+      "loss": 0.6849,
+      "step": 4348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3783681780979369,
+      "learning_rate": 1.9483000659284944e-06,
+      "loss": 0.6908,
+      "step": 4349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2133982767950084,
+      "learning_rate": 1.948269400553188e-06,
+      "loss": 0.694,
+      "step": 4350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8294361372790812,
+      "learning_rate": 1.9482387263275992e-06,
+      "loss": 0.6966,
+      "step": 4351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.329670723616537,
+      "learning_rate": 1.9482080432520146e-06,
+      "loss": 0.6823,
+      "step": 4352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.765836533871267,
+      "learning_rate": 1.9481773513267205e-06,
+      "loss": 0.6966,
+      "step": 4353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8607326979339438,
+      "learning_rate": 1.9481466505520034e-06,
+      "loss": 0.6797,
+      "step": 4354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.351875379493428,
+      "learning_rate": 1.9481159409281493e-06,
+      "loss": 0.6621,
+      "step": 4355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.182812969928094,
+      "learning_rate": 1.948085222455446e-06,
+      "loss": 0.6953,
+      "step": 4356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.363700813012888,
+      "learning_rate": 1.9480544951341787e-06,
+      "loss": 0.6803,
+      "step": 4357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.405297432278351,
+      "learning_rate": 1.9480237589646352e-06,
+      "loss": 0.6654,
+      "step": 4358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.600485065031592,
+      "learning_rate": 1.9479930139471025e-06,
+      "loss": 0.6751,
+      "step": 4359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8677072320390962,
+      "learning_rate": 1.9479622600818667e-06,
+      "loss": 0.6699,
+      "step": 4360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.421250626387373,
+      "learning_rate": 1.9479314973692156e-06,
+      "loss": 0.6888,
+      "step": 4361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.7297049919060665,
+      "learning_rate": 1.9479007258094357e-06,
+      "loss": 0.6719,
+      "step": 4362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.316573485676514,
+      "learning_rate": 1.9478699454028147e-06,
+      "loss": 0.6842,
+      "step": 4363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2139762191696604,
+      "learning_rate": 1.9478391561496398e-06,
+      "loss": 0.6855,
+      "step": 4364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0189600125491063,
+      "learning_rate": 1.947808358050198e-06,
+      "loss": 0.6914,
+      "step": 4365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8671354683724904,
+      "learning_rate": 1.9477775511047773e-06,
+      "loss": 0.681,
+      "step": 4366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2902563280247406,
+      "learning_rate": 1.9477467353136646e-06,
+      "loss": 0.6823,
+      "step": 4367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.073163610009197,
+      "learning_rate": 1.947715910677148e-06,
+      "loss": 0.6673,
+      "step": 4368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3491883786747287,
+      "learning_rate": 1.947685077195515e-06,
+      "loss": 0.6751,
+      "step": 4369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2824940628825348,
+      "learning_rate": 1.947654234869053e-06,
+      "loss": 0.694,
+      "step": 4370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49976697195277625,
+      "learning_rate": 1.947623383698051e-06,
+      "loss": 0.6849,
+      "step": 4371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9529297216522556,
+      "learning_rate": 1.9475925236827952e-06,
+      "loss": 0.6914,
+      "step": 4372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.931378011967905,
+      "learning_rate": 1.947561654823575e-06,
+      "loss": 0.6901,
+      "step": 4373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.942184431558608,
+      "learning_rate": 1.947530777120678e-06,
+      "loss": 0.7064,
+      "step": 4374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4684169878002087,
+      "learning_rate": 1.9474998905743927e-06,
+      "loss": 0.6875,
+      "step": 4375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7668009966715554,
+      "learning_rate": 1.947468995185007e-06,
+      "loss": 0.6979,
+      "step": 4376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8272532038400406,
+      "learning_rate": 1.9474380909528092e-06,
+      "loss": 0.6992,
+      "step": 4377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.416958448320274,
+      "learning_rate": 1.9474071778780884e-06,
+      "loss": 0.6836,
+      "step": 4378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3525924142179386,
+      "learning_rate": 1.947376255961132e-06,
+      "loss": 0.6693,
+      "step": 4379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3702277315489835,
+      "learning_rate": 1.9473453252022295e-06,
+      "loss": 0.6829,
+      "step": 4380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5217406226207262,
+      "learning_rate": 1.9473143856016695e-06,
+      "loss": 0.6777,
+      "step": 4381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0683768552183257,
+      "learning_rate": 1.94728343715974e-06,
+      "loss": 0.6953,
+      "step": 4382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0310525170935267,
+      "learning_rate": 1.947252479876731e-06,
+      "loss": 0.6647,
+      "step": 4383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.220494888601468,
+      "learning_rate": 1.9472215137529307e-06,
+      "loss": 0.6888,
+      "step": 4384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.879314633750355,
+      "learning_rate": 1.947190538788628e-06,
+      "loss": 0.6855,
+      "step": 4385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.822143404832425,
+      "learning_rate": 1.9471595549841123e-06,
+      "loss": 0.666,
+      "step": 4386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6464637669115394,
+      "learning_rate": 1.947128562339673e-06,
+      "loss": 0.6921,
+      "step": 4387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7614317701416926,
+      "learning_rate": 1.947097560855599e-06,
+      "loss": 0.696,
+      "step": 4388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2037413421216303,
+      "learning_rate": 1.9470665505321793e-06,
+      "loss": 0.6901,
+      "step": 4389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7983608265969113,
+      "learning_rate": 1.947035531369704e-06,
+      "loss": 0.6966,
+      "step": 4390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6130462140384142,
+      "learning_rate": 1.9470045033684624e-06,
+      "loss": 0.6777,
+      "step": 4391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7517741302292564,
+      "learning_rate": 1.946973466528744e-06,
+      "loss": 0.6836,
+      "step": 4392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5946600441459164,
+      "learning_rate": 1.9469424208508383e-06,
+      "loss": 0.694,
+      "step": 4393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.191788412910382,
+      "learning_rate": 1.9469113663350357e-06,
+      "loss": 0.6921,
+      "step": 4394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2469101954723882,
+      "learning_rate": 1.946880302981625e-06,
+      "loss": 0.6673,
+      "step": 4395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8406551817464591,
+      "learning_rate": 1.9468492307908973e-06,
+      "loss": 0.6882,
+      "step": 4396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8798229236463184,
+      "learning_rate": 1.9468181497631413e-06,
+      "loss": 0.6875,
+      "step": 4397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9231330201302144,
+      "learning_rate": 1.946787059898648e-06,
+      "loss": 0.6829,
+      "step": 4398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.585985003214602,
+      "learning_rate": 1.9467559611977076e-06,
+      "loss": 0.6914,
+      "step": 4399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.7923941192961665,
+      "learning_rate": 1.94672485366061e-06,
+      "loss": 0.6855,
+      "step": 4400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6291324706549148,
+      "learning_rate": 1.9466937372876456e-06,
+      "loss": 0.6732,
+      "step": 4401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.608010312911152,
+      "learning_rate": 1.946662612079105e-06,
+      "loss": 0.6706,
+      "step": 4402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7329307818655871,
+      "learning_rate": 1.9466314780352783e-06,
+      "loss": 0.7057,
+      "step": 4403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.881452448581508,
+      "learning_rate": 1.9466003351564565e-06,
+      "loss": 0.6849,
+      "step": 4404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3781140623353432,
+      "learning_rate": 1.94656918344293e-06,
+      "loss": 0.6745,
+      "step": 4405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.691734821699818,
+      "learning_rate": 1.9465380228949894e-06,
+      "loss": 0.679,
+      "step": 4406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.59487716052632,
+      "learning_rate": 1.9465068535129256e-06,
+      "loss": 0.6986,
+      "step": 4407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7356463189215126,
+      "learning_rate": 1.9464756752970298e-06,
+      "loss": 0.6901,
+      "step": 4408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0857762887894151,
+      "learning_rate": 1.946444488247593e-06,
+      "loss": 0.6908,
+      "step": 4409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.988311728841555,
+      "learning_rate": 1.946413292364906e-06,
+      "loss": 0.7129,
+      "step": 4410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.39642472018077,
+      "learning_rate": 1.94638208764926e-06,
+      "loss": 0.6901,
+      "step": 4411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5048648408853417,
+      "learning_rate": 1.9463508741009464e-06,
+      "loss": 0.6706,
+      "step": 4412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4917016133146708,
+      "learning_rate": 1.9463196517202564e-06,
+      "loss": 0.6914,
+      "step": 4413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5057303307208255,
+      "learning_rate": 1.9462884205074814e-06,
+      "loss": 0.6738,
+      "step": 4414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.375863162184344,
+      "learning_rate": 1.9462571804629124e-06,
+      "loss": 0.6771,
+      "step": 4415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8565615525484866,
+      "learning_rate": 1.946225931586842e-06,
+      "loss": 0.6589,
+      "step": 4416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3419218738313277,
+      "learning_rate": 1.9461946738795613e-06,
+      "loss": 0.6771,
+      "step": 4417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.480835783019396,
+      "learning_rate": 1.946163407341362e-06,
+      "loss": 0.6895,
+      "step": 4418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7547548958047632,
+      "learning_rate": 1.9461321319725354e-06,
+      "loss": 0.6751,
+      "step": 4419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.82808719962887,
+      "learning_rate": 1.9461008477733746e-06,
+      "loss": 0.6901,
+      "step": 4420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.434418992631334,
+      "learning_rate": 1.9460695547441707e-06,
+      "loss": 0.6732,
+      "step": 4421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8153891654331717,
+      "learning_rate": 1.946038252885216e-06,
+      "loss": 0.6751,
+      "step": 4422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.813716032375922,
+      "learning_rate": 1.9460069421968022e-06,
+      "loss": 0.6979,
+      "step": 4423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8368987299761954,
+      "learning_rate": 1.9459756226792226e-06,
+      "loss": 0.6465,
+      "step": 4424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6988408326246804,
+      "learning_rate": 1.9459442943327686e-06,
+      "loss": 0.6901,
+      "step": 4425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.946238401390829,
+      "learning_rate": 1.9459129571577325e-06,
+      "loss": 0.6921,
+      "step": 4426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1602362708425598,
+      "learning_rate": 1.9458816111544075e-06,
+      "loss": 0.653,
+      "step": 4427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.369064955539727,
+      "learning_rate": 1.9458502563230857e-06,
+      "loss": 0.7246,
+      "step": 4428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.090456062617542,
+      "learning_rate": 1.94581889266406e-06,
+      "loss": 0.696,
+      "step": 4429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.276506730823712,
+      "learning_rate": 1.9457875201776225e-06,
+      "loss": 0.6908,
+      "step": 4430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.758927564452291,
+      "learning_rate": 1.9457561388640668e-06,
+      "loss": 0.6562,
+      "step": 4431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0098241805145065,
+      "learning_rate": 1.9457247487236853e-06,
+      "loss": 0.6745,
+      "step": 4432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0239623739047623,
+      "learning_rate": 1.945693349756771e-06,
+      "loss": 0.7018,
+      "step": 4433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.446224221835294,
+      "learning_rate": 1.945661941963617e-06,
+      "loss": 0.6797,
+      "step": 4434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5452059324312172,
+      "learning_rate": 1.945630525344517e-06,
+      "loss": 0.6914,
+      "step": 4435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8526536442244393,
+      "learning_rate": 1.945599099899763e-06,
+      "loss": 0.6875,
+      "step": 4436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4464613511929025,
+      "learning_rate": 1.9455676656296495e-06,
+      "loss": 0.696,
+      "step": 4437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.021951193286683,
+      "learning_rate": 1.945536222534469e-06,
+      "loss": 0.6803,
+      "step": 4438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1825320190189594,
+      "learning_rate": 1.945504770614515e-06,
+      "loss": 0.7077,
+      "step": 4439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.859129745292048,
+      "learning_rate": 1.9454733098700823e-06,
+      "loss": 0.6901,
+      "step": 4440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1618030872798353,
+      "learning_rate": 1.945441840301463e-06,
+      "loss": 0.6732,
+      "step": 4441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.044289831434746,
+      "learning_rate": 1.9454103619089513e-06,
+      "loss": 0.6673,
+      "step": 4442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.374127785904934,
+      "learning_rate": 1.9453788746928415e-06,
+      "loss": 0.6901,
+      "step": 4443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3716015120982927,
+      "learning_rate": 1.9453473786534266e-06,
+      "loss": 0.6842,
+      "step": 4444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5287641785493735,
+      "learning_rate": 1.945315873791001e-06,
+      "loss": 0.6758,
+      "step": 4445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9341198326521367,
+      "learning_rate": 1.9452843601058595e-06,
+      "loss": 0.6836,
+      "step": 4446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3926565205622605,
+      "learning_rate": 1.9452528375982947e-06,
+      "loss": 0.6836,
+      "step": 4447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.991209034245841,
+      "learning_rate": 1.945221306268602e-06,
+      "loss": 0.7044,
+      "step": 4448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.597124017174842,
+      "learning_rate": 1.9451897661170747e-06,
+      "loss": 0.6771,
+      "step": 4449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.632686887299034,
+      "learning_rate": 1.9451582171440083e-06,
+      "loss": 0.6901,
+      "step": 4450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.271062306661701,
+      "learning_rate": 1.9451266593496963e-06,
+      "loss": 0.6797,
+      "step": 4451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5793495587635134,
+      "learning_rate": 1.9450950927344337e-06,
+      "loss": 0.6836,
+      "step": 4452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.596703318096197,
+      "learning_rate": 1.945063517298515e-06,
+      "loss": 0.6855,
+      "step": 4453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6333842612003067,
+      "learning_rate": 1.9450319330422353e-06,
+      "loss": 0.6953,
+      "step": 4454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7585047593706293,
+      "learning_rate": 1.9450003399658887e-06,
+      "loss": 0.6908,
+      "step": 4455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9282247555783196,
+      "learning_rate": 1.9449687380697704e-06,
+      "loss": 0.6816,
+      "step": 4456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4376482218422666,
+      "learning_rate": 1.944937127354175e-06,
+      "loss": 0.6862,
+      "step": 4457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4913275593076998,
+      "learning_rate": 1.9449055078193985e-06,
+      "loss": 0.6934,
+      "step": 4458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.37468055125571,
+      "learning_rate": 1.9448738794657347e-06,
+      "loss": 0.6842,
+      "step": 4459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.273009961139964,
+      "learning_rate": 1.9448422422934797e-06,
+      "loss": 0.6934,
+      "step": 4460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6272721626614826,
+      "learning_rate": 1.944810596302928e-06,
+      "loss": 0.7044,
+      "step": 4461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.666803547850005,
+      "learning_rate": 1.944778941494376e-06,
+      "loss": 0.6855,
+      "step": 4462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3494164470287937,
+      "learning_rate": 1.9447472778681186e-06,
+      "loss": 0.6738,
+      "step": 4463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.484934935798241,
+      "learning_rate": 1.9447156054244515e-06,
+      "loss": 0.6934,
+      "step": 4464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.730993656123881,
+      "learning_rate": 1.9446839241636695e-06,
+      "loss": 0.6882,
+      "step": 4465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.0753801581306104,
+      "learning_rate": 1.944652234086069e-06,
+      "loss": 0.694,
+      "step": 4466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5826377735389067,
+      "learning_rate": 1.944620535191946e-06,
+      "loss": 0.7025,
+      "step": 4467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.167371823007878,
+      "learning_rate": 1.9445888274815956e-06,
+      "loss": 0.6771,
+      "step": 4468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6326354687504567,
+      "learning_rate": 1.9445571109553147e-06,
+      "loss": 0.6979,
+      "step": 4469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.297225190105134,
+      "learning_rate": 1.9445253856133982e-06,
+      "loss": 0.6862,
+      "step": 4470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7096168341963641,
+      "learning_rate": 1.944493651456143e-06,
+      "loss": 0.6758,
+      "step": 4471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5238738142498804,
+      "learning_rate": 1.9444619084838448e-06,
+      "loss": 0.7005,
+      "step": 4472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.949112231654392,
+      "learning_rate": 1.9444301566968003e-06,
+      "loss": 0.6882,
+      "step": 4473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6456974172194725,
+      "learning_rate": 1.9443983960953057e-06,
+      "loss": 0.6823,
+      "step": 4474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.25248477437888,
+      "learning_rate": 1.9443666266796574e-06,
+      "loss": 0.6836,
+      "step": 4475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.558523480775492,
+      "learning_rate": 1.944334848450152e-06,
+      "loss": 0.6895,
+      "step": 4476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7330063602752914,
+      "learning_rate": 1.9443030614070855e-06,
+      "loss": 0.6725,
+      "step": 4477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.259327728322829,
+      "learning_rate": 1.9442712655507552e-06,
+      "loss": 0.6771,
+      "step": 4478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.079146534839978,
+      "learning_rate": 1.9442394608814576e-06,
+      "loss": 0.6979,
+      "step": 4479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3279138812555993,
+      "learning_rate": 1.9442076473994894e-06,
+      "loss": 0.6875,
+      "step": 4480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1297716223602507,
+      "learning_rate": 1.944175825105148e-06,
+      "loss": 0.6953,
+      "step": 4481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8673988063173215,
+      "learning_rate": 1.94414399399873e-06,
+      "loss": 0.6797,
+      "step": 4482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5055607322771046,
+      "learning_rate": 1.944112154080533e-06,
+      "loss": 0.6986,
+      "step": 4483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8463054372337657,
+      "learning_rate": 1.9440803053508535e-06,
+      "loss": 0.6777,
+      "step": 4484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5637274754653178,
+      "learning_rate": 1.944048447809989e-06,
+      "loss": 0.6849,
+      "step": 4485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4530317955261198,
+      "learning_rate": 1.944016581458237e-06,
+      "loss": 0.6875,
+      "step": 4486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2373233305755265,
+      "learning_rate": 1.9439847062958946e-06,
+      "loss": 0.6855,
+      "step": 4487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3906604546366252,
+      "learning_rate": 1.9439528223232596e-06,
+      "loss": 0.6777,
+      "step": 4488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.474020146038208,
+      "learning_rate": 1.943920929540629e-06,
+      "loss": 0.668,
+      "step": 4489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.029656803372926,
+      "learning_rate": 1.9438890279483015e-06,
+      "loss": 0.6862,
+      "step": 4490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.433063723603159,
+      "learning_rate": 1.9438571175465738e-06,
+      "loss": 0.6875,
+      "step": 4491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6069535475237148,
+      "learning_rate": 1.943825198335744e-06,
+      "loss": 0.6875,
+      "step": 4492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.959295535895532,
+      "learning_rate": 1.9437932703161104e-06,
+      "loss": 0.6699,
+      "step": 4493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6779930709621932,
+      "learning_rate": 1.943761333487971e-06,
+      "loss": 0.694,
+      "step": 4494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4652695697542516,
+      "learning_rate": 1.9437293878516235e-06,
+      "loss": 0.6777,
+      "step": 4495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0402796631238846,
+      "learning_rate": 1.943697433407366e-06,
+      "loss": 0.679,
+      "step": 4496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7788726205529323,
+      "learning_rate": 1.943665470155497e-06,
+      "loss": 0.6562,
+      "step": 4497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5988299652088636,
+      "learning_rate": 1.9436334980963145e-06,
+      "loss": 0.6914,
+      "step": 4498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3274627202503626,
+      "learning_rate": 1.9436015172301175e-06,
+      "loss": 0.6699,
+      "step": 4499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2771882119094313,
+      "learning_rate": 1.9435695275572037e-06,
+      "loss": 0.6895,
+      "step": 4500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.602329109873144,
+      "learning_rate": 1.943537529077872e-06,
+      "loss": 0.6725,
+      "step": 4501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.442106365750087,
+      "learning_rate": 1.9435055217924215e-06,
+      "loss": 0.7122,
+      "step": 4502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2121675723382177,
+      "learning_rate": 1.9434735057011506e-06,
+      "loss": 0.6901,
+      "step": 4503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.261770171577307,
+      "learning_rate": 1.9434414808043577e-06,
+      "loss": 0.6667,
+      "step": 4504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.854971591545948,
+      "learning_rate": 1.9434094471023423e-06,
+      "loss": 0.6999,
+      "step": 4505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1737911843219793,
+      "learning_rate": 1.9433774045954024e-06,
+      "loss": 0.6673,
+      "step": 4506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0513650821418796,
+      "learning_rate": 1.9433453532838384e-06,
+      "loss": 0.6771,
+      "step": 4507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.424208376937677,
+      "learning_rate": 1.9433132931679488e-06,
+      "loss": 0.6647,
+      "step": 4508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.227153442774269,
+      "learning_rate": 1.9432812242480326e-06,
+      "loss": 0.681,
+      "step": 4509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.491386646865371,
+      "learning_rate": 1.9432491465243893e-06,
+      "loss": 0.668,
+      "step": 4510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.709356891211863,
+      "learning_rate": 1.9432170599973182e-06,
+      "loss": 0.6745,
+      "step": 4511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.873794444505156,
+      "learning_rate": 1.9431849646671193e-06,
+      "loss": 0.6647,
+      "step": 4512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.072238036828796,
+      "learning_rate": 1.9431528605340915e-06,
+      "loss": 0.6803,
+      "step": 4513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.9409862249645755,
+      "learning_rate": 1.9431207475985343e-06,
+      "loss": 0.7135,
+      "step": 4514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.52709849628212,
+      "learning_rate": 1.9430886258607483e-06,
+      "loss": 0.6706,
+      "step": 4515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.107384897927765,
+      "learning_rate": 1.9430564953210326e-06,
+      "loss": 0.7096,
+      "step": 4516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5726964407584751,
+      "learning_rate": 1.943024355979687e-06,
+      "loss": 0.6771,
+      "step": 4517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1633452070132904,
+      "learning_rate": 1.9429922078370118e-06,
+      "loss": 0.6641,
+      "step": 4518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.081431071863616,
+      "learning_rate": 1.942960050893307e-06,
+      "loss": 0.6706,
+      "step": 4519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4174695441415859,
+      "learning_rate": 1.9429278851488727e-06,
+      "loss": 0.6673,
+      "step": 4520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7693233782674973,
+      "learning_rate": 1.942895710604009e-06,
+      "loss": 0.6882,
+      "step": 4521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1217481984540107,
+      "learning_rate": 1.9428635272590163e-06,
+      "loss": 0.6719,
+      "step": 4522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6043258634952771,
+      "learning_rate": 1.9428313351141946e-06,
+      "loss": 0.6686,
+      "step": 4523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7409610363667816,
+      "learning_rate": 1.942799134169845e-06,
+      "loss": 0.6758,
+      "step": 4524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9092279158515453,
+      "learning_rate": 1.9427669244262678e-06,
+      "loss": 0.6725,
+      "step": 4525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3611230185687204,
+      "learning_rate": 1.942734705883763e-06,
+      "loss": 0.6706,
+      "step": 4526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.768420774942328,
+      "learning_rate": 1.942702478542632e-06,
+      "loss": 0.6947,
+      "step": 4527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3556815073004074,
+      "learning_rate": 1.9426702424031757e-06,
+      "loss": 0.6908,
+      "step": 4528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6824785269343197,
+      "learning_rate": 1.942637997465695e-06,
+      "loss": 0.6673,
+      "step": 4529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.491803567979242,
+      "learning_rate": 1.9426057437304897e-06,
+      "loss": 0.6927,
+      "step": 4530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.397241884377238,
+      "learning_rate": 1.942573481197862e-06,
+      "loss": 0.6706,
+      "step": 4531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9564569572119014,
+      "learning_rate": 1.9425412098681125e-06,
+      "loss": 0.6745,
+      "step": 4532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0311001211924835,
+      "learning_rate": 1.9425089297415428e-06,
+      "loss": 0.6901,
+      "step": 4533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8319664594694054,
+      "learning_rate": 1.9424766408184536e-06,
+      "loss": 0.6647,
+      "step": 4534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9738818949631047,
+      "learning_rate": 1.9424443430991467e-06,
+      "loss": 0.666,
+      "step": 4535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8381553877449273,
+      "learning_rate": 1.942412036583923e-06,
+      "loss": 0.6855,
+      "step": 4536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8460506378320364,
+      "learning_rate": 1.9423797212730854e-06,
+      "loss": 0.6771,
+      "step": 4537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1448586300393755,
+      "learning_rate": 1.9423473971669336e-06,
+      "loss": 0.6895,
+      "step": 4538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.639377282857431,
+      "learning_rate": 1.9423150642657706e-06,
+      "loss": 0.7038,
+      "step": 4539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.77735006872778,
+      "learning_rate": 1.9422827225698976e-06,
+      "loss": 0.6628,
+      "step": 4540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.75121689969217,
+      "learning_rate": 1.942250372079617e-06,
+      "loss": 0.6966,
+      "step": 4541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2054110573387624,
+      "learning_rate": 1.94221801279523e-06,
+      "loss": 0.668,
+      "step": 4542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.550101206531818,
+      "learning_rate": 1.942185644717039e-06,
+      "loss": 0.7227,
+      "step": 4543
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.080166159260949,
+      "learning_rate": 1.942153267845346e-06,
+      "loss": 0.6901,
+      "step": 4544
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.13194905544647,
+      "learning_rate": 1.9421208821804538e-06,
+      "loss": 0.6654,
+      "step": 4545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5310239771397824,
+      "learning_rate": 1.9420884877226634e-06,
+      "loss": 0.6849,
+      "step": 4546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5775369755171336,
+      "learning_rate": 1.942056084472278e-06,
+      "loss": 0.696,
+      "step": 4547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5186996941781012,
+      "learning_rate": 1.9420236724296e-06,
+      "loss": 0.696,
+      "step": 4548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8463125515033169,
+      "learning_rate": 1.9419912515949317e-06,
+      "loss": 0.6803,
+      "step": 4549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1959849505269413,
+      "learning_rate": 1.9419588219685756e-06,
+      "loss": 0.6829,
+      "step": 4550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9348287254665508,
+      "learning_rate": 1.9419263835508345e-06,
+      "loss": 0.6868,
+      "step": 4551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.133311905997823,
+      "learning_rate": 1.941893936342011e-06,
+      "loss": 0.6608,
+      "step": 4552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.6933538181010155,
+      "learning_rate": 1.941861480342409e-06,
+      "loss": 0.6758,
+      "step": 4553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.604582313028975,
+      "learning_rate": 1.94182901555233e-06,
+      "loss": 0.6966,
+      "step": 4554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4421770381464052,
+      "learning_rate": 1.941796541972077e-06,
+      "loss": 0.6712,
+      "step": 4555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5976224543042137,
+      "learning_rate": 1.941764059601954e-06,
+      "loss": 0.6777,
+      "step": 4556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.91872124383277,
+      "learning_rate": 1.9417315684422636e-06,
+      "loss": 0.6875,
+      "step": 4557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7695945220833449,
+      "learning_rate": 1.9416990684933094e-06,
+      "loss": 0.7161,
+      "step": 4558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6708754921987589,
+      "learning_rate": 1.9416665597553945e-06,
+      "loss": 0.6953,
+      "step": 4559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9548697104759052,
+      "learning_rate": 1.9416340422288224e-06,
+      "loss": 0.6725,
+      "step": 4560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.4910015454085555,
+      "learning_rate": 1.9416015159138963e-06,
+      "loss": 0.696,
+      "step": 4561
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6832168845830883,
+      "learning_rate": 1.94156898081092e-06,
+      "loss": 0.6868,
+      "step": 4562
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6023346518104261,
+      "learning_rate": 1.941536436920197e-06,
+      "loss": 0.666,
+      "step": 4563
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.801018197804071,
+      "learning_rate": 1.9415038842420317e-06,
+      "loss": 0.6901,
+      "step": 4564
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.193880325749101,
+      "learning_rate": 1.941471322776727e-06,
+      "loss": 0.7051,
+      "step": 4565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.826071329496873,
+      "learning_rate": 1.941438752524587e-06,
+      "loss": 0.7201,
+      "step": 4566
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6260820803764442,
+      "learning_rate": 1.941406173485916e-06,
+      "loss": 0.7129,
+      "step": 4567
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3227032298345793,
+      "learning_rate": 1.941373585661018e-06,
+      "loss": 0.6888,
+      "step": 4568
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9832203388888803,
+      "learning_rate": 1.941340989050197e-06,
+      "loss": 0.6947,
+      "step": 4569
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7948478883573509,
+      "learning_rate": 1.9413083836537578e-06,
+      "loss": 0.6842,
+      "step": 4570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.135369273172944,
+      "learning_rate": 1.9412757694720036e-06,
+      "loss": 0.6908,
+      "step": 4571
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.781969236539982,
+      "learning_rate": 1.9412431465052397e-06,
+      "loss": 0.6784,
+      "step": 4572
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7793513455982067,
+      "learning_rate": 1.94121051475377e-06,
+      "loss": 0.6999,
+      "step": 4573
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4821255388712866,
+      "learning_rate": 1.9411778742179e-06,
+      "loss": 0.7005,
+      "step": 4574
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.090045253576871,
+      "learning_rate": 1.941145224897933e-06,
+      "loss": 0.6816,
+      "step": 4575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4860060996668661,
+      "learning_rate": 1.9411125667941743e-06,
+      "loss": 0.6914,
+      "step": 4576
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4607806522813893,
+      "learning_rate": 1.941079899906929e-06,
+      "loss": 0.6816,
+      "step": 4577
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.152073746174679,
+      "learning_rate": 1.941047224236502e-06,
+      "loss": 0.668,
+      "step": 4578
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.117409988734107,
+      "learning_rate": 1.9410145397831976e-06,
+      "loss": 0.6875,
+      "step": 4579
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.662626142526115,
+      "learning_rate": 1.940981846547322e-06,
+      "loss": 0.6895,
+      "step": 4580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9039331611767536,
+      "learning_rate": 1.9409491445291786e-06,
+      "loss": 0.6829,
+      "step": 4581
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4406917652842848,
+      "learning_rate": 1.9409164337290744e-06,
+      "loss": 0.6842,
+      "step": 4582
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4245709936119697,
+      "learning_rate": 1.9408837141473133e-06,
+      "loss": 0.6751,
+      "step": 4583
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.559608506518679,
+      "learning_rate": 1.940850985784202e-06,
+      "loss": 0.6908,
+      "step": 4584
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7958033917158036,
+      "learning_rate": 1.940818248640045e-06,
+      "loss": 0.7064,
+      "step": 4585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.274157686631596,
+      "learning_rate": 1.940785502715148e-06,
+      "loss": 0.694,
+      "step": 4586
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.541694874234467,
+      "learning_rate": 1.9407527480098165e-06,
+      "loss": 0.6888,
+      "step": 4587
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9770902666905674,
+      "learning_rate": 1.9407199845243566e-06,
+      "loss": 0.6654,
+      "step": 4588
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5045303272787047,
+      "learning_rate": 1.940687212259074e-06,
+      "loss": 0.6908,
+      "step": 4589
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.158744163828834,
+      "learning_rate": 1.9406544312142742e-06,
+      "loss": 0.6868,
+      "step": 4590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.0421864790926838,
+      "learning_rate": 1.940621641390264e-06,
+      "loss": 0.6699,
+      "step": 4591
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.212733672204899,
+      "learning_rate": 1.940588842787348e-06,
+      "loss": 0.6654,
+      "step": 4592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1965852630733687,
+      "learning_rate": 1.9405560354058338e-06,
+      "loss": 0.6868,
+      "step": 4593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.191612091870629,
+      "learning_rate": 1.9405232192460266e-06,
+      "loss": 0.6647,
+      "step": 4594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1328304435626406,
+      "learning_rate": 1.940490394308233e-06,
+      "loss": 0.6738,
+      "step": 4595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.105980342280992,
+      "learning_rate": 1.9404575605927597e-06,
+      "loss": 0.6855,
+      "step": 4596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.29571355236885,
+      "learning_rate": 1.9404247180999124e-06,
+      "loss": 0.6888,
+      "step": 4597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1031783298403148,
+      "learning_rate": 1.9403918668299983e-06,
+      "loss": 0.6914,
+      "step": 4598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.902911569538608,
+      "learning_rate": 1.9403590067833236e-06,
+      "loss": 0.6562,
+      "step": 4599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6732791008764911,
+      "learning_rate": 1.9403261379601954e-06,
+      "loss": 0.6849,
+      "step": 4600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1867861133550601,
+      "learning_rate": 1.94029326036092e-06,
+      "loss": 0.6654,
+      "step": 4601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3704623743513067,
+      "learning_rate": 1.9402603739858045e-06,
+      "loss": 0.7038,
+      "step": 4602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.631443757590841,
+      "learning_rate": 1.9402274788351554e-06,
+      "loss": 0.6908,
+      "step": 4603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1038144127974814,
+      "learning_rate": 1.9401945749092806e-06,
+      "loss": 0.6966,
+      "step": 4604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.716205210903247,
+      "learning_rate": 1.9401616622084863e-06,
+      "loss": 0.6947,
+      "step": 4605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3581841159752353,
+      "learning_rate": 1.9401287407330803e-06,
+      "loss": 0.7064,
+      "step": 4606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9112253305738953,
+      "learning_rate": 1.9400958104833696e-06,
+      "loss": 0.6842,
+      "step": 4607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8081084849806484,
+      "learning_rate": 1.9400628714596614e-06,
+      "loss": 0.6888,
+      "step": 4608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.071337354338524,
+      "learning_rate": 1.9400299236622636e-06,
+      "loss": 0.6738,
+      "step": 4609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.761038167713839,
+      "learning_rate": 1.939996967091483e-06,
+      "loss": 0.7057,
+      "step": 4610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6922450186893402,
+      "learning_rate": 1.939964001747628e-06,
+      "loss": 0.7051,
+      "step": 4611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.902314129637913,
+      "learning_rate": 1.939931027631006e-06,
+      "loss": 0.6986,
+      "step": 4612
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5284157303821155,
+      "learning_rate": 1.939898044741924e-06,
+      "loss": 0.694,
+      "step": 4613
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.43321576273256107,
+      "learning_rate": 1.9398650530806905e-06,
+      "loss": 0.6908,
+      "step": 4614
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0995080259872902,
+      "learning_rate": 1.9398320526476137e-06,
+      "loss": 0.6784,
+      "step": 4615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.951948478299124,
+      "learning_rate": 1.939799043443001e-06,
+      "loss": 0.6803,
+      "step": 4616
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48029235469037884,
+      "learning_rate": 1.9397660254671606e-06,
+      "loss": 0.6914,
+      "step": 4617
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7072911522755735,
+      "learning_rate": 1.939732998720401e-06,
+      "loss": 0.6882,
+      "step": 4618
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4989518687888888,
+      "learning_rate": 1.93969996320303e-06,
+      "loss": 0.694,
+      "step": 4619
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3906962090221362,
+      "learning_rate": 1.9396669189153566e-06,
+      "loss": 0.6738,
+      "step": 4620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5765388704891201,
+      "learning_rate": 1.9396338658576886e-06,
+      "loss": 0.6901,
+      "step": 4621
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5349766711656379,
+      "learning_rate": 1.9396008040303343e-06,
+      "loss": 0.6927,
+      "step": 4622
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.366399486455268,
+      "learning_rate": 1.9395677334336027e-06,
+      "loss": 0.6803,
+      "step": 4623
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.152531925475314,
+      "learning_rate": 1.9395346540678024e-06,
+      "loss": 0.6842,
+      "step": 4624
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.489004114544809,
+      "learning_rate": 1.939501565933242e-06,
+      "loss": 0.6777,
+      "step": 4625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.276888325049014,
+      "learning_rate": 1.9394684690302307e-06,
+      "loss": 0.6908,
+      "step": 4626
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2579125823539852,
+      "learning_rate": 1.939435363359077e-06,
+      "loss": 0.6797,
+      "step": 4627
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8014572297154965,
+      "learning_rate": 1.93940224892009e-06,
+      "loss": 0.6862,
+      "step": 4628
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.044169160887862,
+      "learning_rate": 1.939369125713579e-06,
+      "loss": 0.6797,
+      "step": 4629
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.030245279370902,
+      "learning_rate": 1.939335993739852e-06,
+      "loss": 0.6934,
+      "step": 4630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6422805630580934,
+      "learning_rate": 1.9393028529992198e-06,
+      "loss": 0.6836,
+      "step": 4631
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7123181271403418,
+      "learning_rate": 1.9392697034919907e-06,
+      "loss": 0.6855,
+      "step": 4632
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.812792267988028,
+      "learning_rate": 1.9392365452184743e-06,
+      "loss": 0.6816,
+      "step": 4633
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.388922974100465,
+      "learning_rate": 1.9392033781789804e-06,
+      "loss": 0.6875,
+      "step": 4634
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.309722557100435,
+      "learning_rate": 1.939170202373818e-06,
+      "loss": 0.679,
+      "step": 4635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.264028712008374,
+      "learning_rate": 1.939137017803298e-06,
+      "loss": 0.6934,
+      "step": 4636
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8334782513440696,
+      "learning_rate": 1.9391038244677283e-06,
+      "loss": 0.6901,
+      "step": 4637
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.962769984861091,
+      "learning_rate": 1.9390706223674197e-06,
+      "loss": 0.6784,
+      "step": 4638
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.576555473821989,
+      "learning_rate": 1.939037411502682e-06,
+      "loss": 0.6855,
+      "step": 4639
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9320754513093252,
+      "learning_rate": 1.939004191873825e-06,
+      "loss": 0.679,
+      "step": 4640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7055094360350899,
+      "learning_rate": 1.938970963481159e-06,
+      "loss": 0.6882,
+      "step": 4641
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9942293914644408,
+      "learning_rate": 1.938937726324994e-06,
+      "loss": 0.6888,
+      "step": 4642
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.201398924085725,
+      "learning_rate": 1.9389044804056398e-06,
+      "loss": 0.6895,
+      "step": 4643
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.162824661151475,
+      "learning_rate": 1.938871225723407e-06,
+      "loss": 0.707,
+      "step": 4644
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1187850897245024,
+      "learning_rate": 1.9388379622786067e-06,
+      "loss": 0.696,
+      "step": 4645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7693301994142272,
+      "learning_rate": 1.9388046900715483e-06,
+      "loss": 0.6771,
+      "step": 4646
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.0500943973166057,
+      "learning_rate": 1.9387714091025426e-06,
+      "loss": 0.6706,
+      "step": 4647
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7340264217449948,
+      "learning_rate": 1.9387381193719005e-06,
+      "loss": 0.6797,
+      "step": 4648
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.431567362695743,
+      "learning_rate": 1.9387048208799324e-06,
+      "loss": 0.7064,
+      "step": 4649
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5451260324124858,
+      "learning_rate": 1.9386715136269493e-06,
+      "loss": 0.696,
+      "step": 4650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3724832428135445,
+      "learning_rate": 1.938638197613262e-06,
+      "loss": 0.6816,
+      "step": 4651
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6049657404921454,
+      "learning_rate": 1.938604872839181e-06,
+      "loss": 0.6693,
+      "step": 4652
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.829689903957837,
+      "learning_rate": 1.938571539305018e-06,
+      "loss": 0.6849,
+      "step": 4653
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9040573660471478,
+      "learning_rate": 1.9385381970110835e-06,
+      "loss": 0.6797,
+      "step": 4654
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6407854203235669,
+      "learning_rate": 1.938504845957689e-06,
+      "loss": 0.6712,
+      "step": 4655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9332937127317347,
+      "learning_rate": 1.9384714861451466e-06,
+      "loss": 0.6758,
+      "step": 4656
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.805330744579195,
+      "learning_rate": 1.938438117573766e-06,
+      "loss": 0.6758,
+      "step": 4657
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7622381374425076,
+      "learning_rate": 1.9384047402438597e-06,
+      "loss": 0.7044,
+      "step": 4658
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.90286324806706,
+      "learning_rate": 1.9383713541557393e-06,
+      "loss": 0.6901,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.3355062697444104,
+      "learning_rate": 1.9383379593097158e-06,
+      "loss": 0.6849,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4162581772560805,
+      "learning_rate": 1.938304555706101e-06,
+      "loss": 0.6615,
+      "step": 4661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.072647290292624,
+      "learning_rate": 1.938271143345207e-06,
+      "loss": 0.679,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9667173621117834,
+      "learning_rate": 1.9382377222273455e-06,
+      "loss": 0.6842,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.5347195404446765,
+      "learning_rate": 1.938204292352828e-06,
+      "loss": 0.6667,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5122963144739505,
+      "learning_rate": 1.9381708537219672e-06,
+      "loss": 0.6764,
+      "step": 4665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3574941183975835,
+      "learning_rate": 1.9381374063350745e-06,
+      "loss": 0.6836,
+      "step": 4666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9042500665137655,
+      "learning_rate": 1.9381039501924627e-06,
+      "loss": 0.6816,
+      "step": 4667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.180596290318754,
+      "learning_rate": 1.938070485294444e-06,
+      "loss": 0.6875,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.996751309845459,
+      "learning_rate": 1.93803701164133e-06,
+      "loss": 0.6992,
+      "step": 4669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7256676786052247,
+      "learning_rate": 1.938003529233434e-06,
+      "loss": 0.679,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.990306077184205,
+      "learning_rate": 1.937970038071068e-06,
+      "loss": 0.6973,
+      "step": 4671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.18639492310863,
+      "learning_rate": 1.937936538154545e-06,
+      "loss": 0.6862,
+      "step": 4672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.060703978102196,
+      "learning_rate": 1.9379030294841767e-06,
+      "loss": 0.6536,
+      "step": 4673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.559172301312083,
+      "learning_rate": 1.9378695120602766e-06,
+      "loss": 0.707,
+      "step": 4674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0076470548913052,
+      "learning_rate": 1.937835985883158e-06,
+      "loss": 0.6751,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.318673220223257,
+      "learning_rate": 1.9378024509531324e-06,
+      "loss": 0.6927,
+      "step": 4676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5907134846551685,
+      "learning_rate": 1.9377689072705142e-06,
+      "loss": 0.6784,
+      "step": 4677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7645098208841323,
+      "learning_rate": 1.9377353548356156e-06,
+      "loss": 0.7038,
+      "step": 4678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.141752957352954,
+      "learning_rate": 1.93770179364875e-06,
+      "loss": 0.6784,
+      "step": 4679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6716262009492748,
+      "learning_rate": 1.9376682237102305e-06,
+      "loss": 0.6966,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.404699284317886,
+      "learning_rate": 1.937634645020371e-06,
+      "loss": 0.6979,
+      "step": 4681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5102397014080715,
+      "learning_rate": 1.9376010575794843e-06,
+      "loss": 0.6901,
+      "step": 4682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2771539211791687,
+      "learning_rate": 1.937567461387884e-06,
+      "loss": 0.6745,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.427584872747752,
+      "learning_rate": 1.9375338564458833e-06,
+      "loss": 0.6895,
+      "step": 4684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5400507097097709,
+      "learning_rate": 1.9375002427537965e-06,
+      "loss": 0.696,
+      "step": 4685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.46290758991668,
+      "learning_rate": 1.937466620311937e-06,
+      "loss": 0.6829,
+      "step": 4686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.264838348116425,
+      "learning_rate": 1.9374329891206183e-06,
+      "loss": 0.6829,
+      "step": 4687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.40562465563685296,
+      "learning_rate": 1.937399349180155e-06,
+      "loss": 0.6823,
+      "step": 4688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.229247823206577,
+      "learning_rate": 1.9373657004908606e-06,
+      "loss": 0.6927,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.809314744133374,
+      "learning_rate": 1.937332043053049e-06,
+      "loss": 0.6914,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.134862857159744,
+      "learning_rate": 1.937298376867035e-06,
+      "loss": 0.6842,
+      "step": 4691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.30343672667496,
+      "learning_rate": 1.9372647019331324e-06,
+      "loss": 0.6849,
+      "step": 4692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8673964334679187,
+      "learning_rate": 1.937231018251655e-06,
+      "loss": 0.668,
+      "step": 4693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.48503154231043155,
+      "learning_rate": 1.9371973258229177e-06,
+      "loss": 0.7057,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0028639269518895,
+      "learning_rate": 1.9371636246472353e-06,
+      "loss": 0.6888,
+      "step": 4695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9851796343864776,
+      "learning_rate": 1.9371299147249212e-06,
+      "loss": 0.679,
+      "step": 4696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0748006714747054,
+      "learning_rate": 1.937096196056291e-06,
+      "loss": 0.6712,
+      "step": 4697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.188295816898359,
+      "learning_rate": 1.9370624686416595e-06,
+      "loss": 0.6882,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.36381881542718,
+      "learning_rate": 1.9370287324813407e-06,
+      "loss": 0.6803,
+      "step": 4699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.536210286141105,
+      "learning_rate": 1.9369949875756503e-06,
+      "loss": 0.6738,
+      "step": 4700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3298107119892935,
+      "learning_rate": 1.9369612339249025e-06,
+      "loss": 0.6764,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9167886841043678,
+      "learning_rate": 1.936927471529413e-06,
+      "loss": 0.6973,
+      "step": 4702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.80969929752904,
+      "learning_rate": 1.936893700389496e-06,
+      "loss": 0.6764,
+      "step": 4703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9844260137917855,
+      "learning_rate": 1.9368599205054673e-06,
+      "loss": 0.6803,
+      "step": 4704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4044342580330442,
+      "learning_rate": 1.9368261318776424e-06,
+      "loss": 0.6719,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.233125246594602,
+      "learning_rate": 1.936792334506336e-06,
+      "loss": 0.6908,
+      "step": 4706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2507680459068187,
+      "learning_rate": 1.9367585283918643e-06,
+      "loss": 0.6699,
+      "step": 4707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.429741707267216,
+      "learning_rate": 1.9367247135345424e-06,
+      "loss": 0.7038,
+      "step": 4708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.038027453097706,
+      "learning_rate": 1.9366908899346855e-06,
+      "loss": 0.6849,
+      "step": 4709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.336757968952408,
+      "learning_rate": 1.93665705759261e-06,
+      "loss": 0.6621,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6943542037059798,
+      "learning_rate": 1.936623216508631e-06,
+      "loss": 0.6986,
+      "step": 4711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5011315046830234,
+      "learning_rate": 1.936589366683065e-06,
+      "loss": 0.6608,
+      "step": 4712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5426806018302771,
+      "learning_rate": 1.936555508116228e-06,
+      "loss": 0.6745,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9216644750365588,
+      "learning_rate": 1.936521640808435e-06,
+      "loss": 0.6803,
+      "step": 4714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.164147654887322,
+      "learning_rate": 1.936487764760003e-06,
+      "loss": 0.7103,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1065701646908095,
+      "learning_rate": 1.936453879971248e-06,
+      "loss": 0.6504,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7134259955839363,
+      "learning_rate": 1.936419986442486e-06,
+      "loss": 0.6927,
+      "step": 4717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.273895737880805,
+      "learning_rate": 1.9363860841740332e-06,
+      "loss": 0.679,
+      "step": 4718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3338901915702712,
+      "learning_rate": 1.9363521731662063e-06,
+      "loss": 0.696,
+      "step": 4719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.04318545323967,
+      "learning_rate": 1.936318253419322e-06,
+      "loss": 0.6973,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8390104691785973,
+      "learning_rate": 1.9362843249336966e-06,
+      "loss": 0.6914,
+      "step": 4721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7596618804808181,
+      "learning_rate": 1.936250387709647e-06,
+      "loss": 0.6582,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.524548645169548,
+      "learning_rate": 1.9362164417474892e-06,
+      "loss": 0.6803,
+      "step": 4723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.44283140021041,
+      "learning_rate": 1.936182487047541e-06,
+      "loss": 0.6868,
+      "step": 4724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0635924257935725,
+      "learning_rate": 1.9361485236101186e-06,
+      "loss": 0.6764,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.503073328880959,
+      "learning_rate": 1.936114551435539e-06,
+      "loss": 0.6738,
+      "step": 4726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.431551312123146,
+      "learning_rate": 1.93608057052412e-06,
+      "loss": 0.6836,
+      "step": 4727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.5782745051659335,
+      "learning_rate": 1.9360465808761782e-06,
+      "loss": 0.666,
+      "step": 4728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6159602708004788,
+      "learning_rate": 1.9360125824920307e-06,
+      "loss": 0.6908,
+      "step": 4729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.324174106467415,
+      "learning_rate": 1.935978575371995e-06,
+      "loss": 0.6966,
+      "step": 4730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.936741860816285,
+      "learning_rate": 1.9359445595163885e-06,
+      "loss": 0.6719,
+      "step": 4731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.339589666343436,
+      "learning_rate": 1.9359105349255286e-06,
+      "loss": 0.696,
+      "step": 4732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.618661932007486,
+      "learning_rate": 1.935876501599733e-06,
+      "loss": 0.681,
+      "step": 4733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.817252262091888,
+      "learning_rate": 1.9358424595393196e-06,
+      "loss": 0.6615,
+      "step": 4734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8203009766566176,
+      "learning_rate": 1.9358084087446052e-06,
+      "loss": 0.6966,
+      "step": 4735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.226761122452084,
+      "learning_rate": 1.9357743492159085e-06,
+      "loss": 0.679,
+      "step": 4736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.260295417811491,
+      "learning_rate": 1.935740280953547e-06,
+      "loss": 0.6829,
+      "step": 4737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9819618960716285,
+      "learning_rate": 1.935706203957838e-06,
+      "loss": 0.6862,
+      "step": 4738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7901860537090722,
+      "learning_rate": 1.9356721182291015e-06,
+      "loss": 0.6725,
+      "step": 4739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.190699392434953,
+      "learning_rate": 1.9356380237676534e-06,
+      "loss": 0.6927,
+      "step": 4740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.2533694581646095,
+      "learning_rate": 1.935603920573813e-06,
+      "loss": 0.6673,
+      "step": 4741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.328039485646577,
+      "learning_rate": 1.935569808647899e-06,
+      "loss": 0.6921,
+      "step": 4742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.501475157301153,
+      "learning_rate": 1.935535687990229e-06,
+      "loss": 0.6654,
+      "step": 4743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5967532397356112,
+      "learning_rate": 1.9355015586011218e-06,
+      "loss": 0.6803,
+      "step": 4744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.760411030872428,
+      "learning_rate": 1.9354674204808952e-06,
+      "loss": 0.6836,
+      "step": 4745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.500827878200978,
+      "learning_rate": 1.935433273629869e-06,
+      "loss": 0.6816,
+      "step": 4746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7157121858565523,
+      "learning_rate": 1.9353991180483615e-06,
+      "loss": 0.6895,
+      "step": 4747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.911587810757561,
+      "learning_rate": 1.9353649537366908e-06,
+      "loss": 0.7057,
+      "step": 4748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9902374880181716,
+      "learning_rate": 1.9353307806951762e-06,
+      "loss": 0.6927,
+      "step": 4749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8512459901324817,
+      "learning_rate": 1.935296598924137e-06,
+      "loss": 0.6758,
+      "step": 4750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4569899543879241,
+      "learning_rate": 1.935262408423892e-06,
+      "loss": 0.6921,
+      "step": 4751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.698130247834169,
+      "learning_rate": 1.9352282091947607e-06,
+      "loss": 0.6855,
+      "step": 4752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.271337173325692,
+      "learning_rate": 1.935194001237061e-06,
+      "loss": 0.6803,
+      "step": 4753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9274806634794124,
+      "learning_rate": 1.935159784551113e-06,
+      "loss": 0.6758,
+      "step": 4754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6282085576176926,
+      "learning_rate": 1.9351255591372363e-06,
+      "loss": 0.7005,
+      "step": 4755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8926071223787373,
+      "learning_rate": 1.93509132499575e-06,
+      "loss": 0.6914,
+      "step": 4756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.988473113161297,
+      "learning_rate": 1.935057082126974e-06,
+      "loss": 0.7012,
+      "step": 4757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2764787731415606,
+      "learning_rate": 1.9350228305312274e-06,
+      "loss": 0.6862,
+      "step": 4758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.44006405003422755,
+      "learning_rate": 1.9349885702088296e-06,
+      "loss": 0.6803,
+      "step": 4759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.47164011349076307,
+      "learning_rate": 1.934954301160101e-06,
+      "loss": 0.6816,
+      "step": 4760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.5166321132277245,
+      "learning_rate": 1.9349200233853617e-06,
+      "loss": 0.6784,
+      "step": 4761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8731434034551975,
+      "learning_rate": 1.9348857368849307e-06,
+      "loss": 0.6777,
+      "step": 4762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6945732966507476,
+      "learning_rate": 1.9348514416591282e-06,
+      "loss": 0.6947,
+      "step": 4763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3380508464297933,
+      "learning_rate": 1.934817137708275e-06,
+      "loss": 0.6764,
+      "step": 4764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5748463426479633,
+      "learning_rate": 1.9347828250326903e-06,
+      "loss": 0.6699,
+      "step": 4765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8175450690724022,
+      "learning_rate": 1.9347485036326952e-06,
+      "loss": 0.6875,
+      "step": 4766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.733459672202913,
+      "learning_rate": 1.9347141735086098e-06,
+      "loss": 0.6673,
+      "step": 4767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1405375170040082,
+      "learning_rate": 1.934679834660754e-06,
+      "loss": 0.6829,
+      "step": 4768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.8076122851726115,
+      "learning_rate": 1.934645487089449e-06,
+      "loss": 0.6999,
+      "step": 4769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.149052122327441,
+      "learning_rate": 1.9346111307950145e-06,
+      "loss": 0.6771,
+      "step": 4770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4210328857023806,
+      "learning_rate": 1.934576765777772e-06,
+      "loss": 0.6875,
+      "step": 4771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6824997095602688,
+      "learning_rate": 1.934542392038042e-06,
+      "loss": 0.6706,
+      "step": 4772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3553910188688991,
+      "learning_rate": 1.9345080095761446e-06,
+      "loss": 0.6836,
+      "step": 4773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.036396851781547,
+      "learning_rate": 1.934473618392402e-06,
+      "loss": 0.6751,
+      "step": 4774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.60146947142649,
+      "learning_rate": 1.9344392184871344e-06,
+      "loss": 0.6868,
+      "step": 4775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7828410158209875,
+      "learning_rate": 1.9344048098606626e-06,
+      "loss": 0.6667,
+      "step": 4776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6122713179931263,
+      "learning_rate": 1.9343703925133086e-06,
+      "loss": 0.6836,
+      "step": 4777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.1803457236847965,
+      "learning_rate": 1.9343359664453924e-06,
+      "loss": 0.6908,
+      "step": 4778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3284736049581793,
+      "learning_rate": 1.9343015316572366e-06,
+      "loss": 0.7044,
+      "step": 4779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.507894860967619,
+      "learning_rate": 1.9342670881491615e-06,
+      "loss": 0.679,
+      "step": 4780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.777319125047473,
+      "learning_rate": 1.9342326359214892e-06,
+      "loss": 0.6647,
+      "step": 4781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6333920618551887,
+      "learning_rate": 1.9341981749745415e-06,
+      "loss": 0.7025,
+      "step": 4782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.659029691386153,
+      "learning_rate": 1.934163705308639e-06,
+      "loss": 0.6855,
+      "step": 4783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1478235497831506,
+      "learning_rate": 1.9341292269241042e-06,
+      "loss": 0.6699,
+      "step": 4784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.34493244920275,
+      "learning_rate": 1.934094739821259e-06,
+      "loss": 0.6452,
+      "step": 4785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.08735303422741,
+      "learning_rate": 1.934060244000425e-06,
+      "loss": 0.6432,
+      "step": 4786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.940494644845843,
+      "learning_rate": 1.934025739461924e-06,
+      "loss": 0.6523,
+      "step": 4787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.504691662797854,
+      "learning_rate": 1.9339912262060782e-06,
+      "loss": 0.6621,
+      "step": 4788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.478116278441096,
+      "learning_rate": 1.933956704233209e-06,
+      "loss": 0.6901,
+      "step": 4789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.203140323880074,
+      "learning_rate": 1.9339221735436402e-06,
+      "loss": 0.6712,
+      "step": 4790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0228433958510115,
+      "learning_rate": 1.933887634137693e-06,
+      "loss": 0.6712,
+      "step": 4791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9935846318528497,
+      "learning_rate": 1.9338530860156897e-06,
+      "loss": 0.6719,
+      "step": 4792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.842351267554437,
+      "learning_rate": 1.933818529177953e-06,
+      "loss": 0.6699,
+      "step": 4793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.318055206513973,
+      "learning_rate": 1.9337839636248054e-06,
+      "loss": 0.7129,
+      "step": 4794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.44835555780951,
+      "learning_rate": 1.9337493893565692e-06,
+      "loss": 0.7168,
+      "step": 4795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.118500488542466,
+      "learning_rate": 1.933714806373568e-06,
+      "loss": 0.6771,
+      "step": 4796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.011353402884152,
+      "learning_rate": 1.933680214676123e-06,
+      "loss": 0.679,
+      "step": 4797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0582177162501545,
+      "learning_rate": 1.933645614264559e-06,
+      "loss": 0.6849,
+      "step": 4798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0213217637991154,
+      "learning_rate": 1.9336110051391976e-06,
+      "loss": 0.679,
+      "step": 4799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.683027723790547,
+      "learning_rate": 1.9335763873003617e-06,
+      "loss": 0.6686,
+      "step": 4800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5804456009816411,
+      "learning_rate": 1.9335417607483757e-06,
+      "loss": 0.6777,
+      "step": 4801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.018551991431303,
+      "learning_rate": 1.933507125483561e-06,
+      "loss": 0.6927,
+      "step": 4802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.67658878308912,
+      "learning_rate": 1.933472481506242e-06,
+      "loss": 0.6654,
+      "step": 4803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.605523441799136,
+      "learning_rate": 1.933437828816742e-06,
+      "loss": 0.6797,
+      "step": 4804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.879618411392443,
+      "learning_rate": 1.9334031674153843e-06,
+      "loss": 0.6823,
+      "step": 4805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0593024664615671,
+      "learning_rate": 1.933368497302492e-06,
+      "loss": 0.6979,
+      "step": 4806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.09109944768797,
+      "learning_rate": 1.933333818478389e-06,
+      "loss": 0.7116,
+      "step": 4807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7299671790988922,
+      "learning_rate": 1.933299130943399e-06,
+      "loss": 0.6777,
+      "step": 4808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8288869813274156,
+      "learning_rate": 1.9332644346978456e-06,
+      "loss": 0.7103,
+      "step": 4809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9641520504813703,
+      "learning_rate": 1.9332297297420527e-06,
+      "loss": 0.7077,
+      "step": 4810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2114051706880922,
+      "learning_rate": 1.933195016076344e-06,
+      "loss": 0.6855,
+      "step": 4811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.778711965223833,
+      "learning_rate": 1.933160293701044e-06,
+      "loss": 0.6686,
+      "step": 4812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.526955233166704,
+      "learning_rate": 1.9331255626164764e-06,
+      "loss": 0.6986,
+      "step": 4813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9041673651011326,
+      "learning_rate": 1.9330908228229655e-06,
+      "loss": 0.7012,
+      "step": 4814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2597785956342897,
+      "learning_rate": 1.933056074320835e-06,
+      "loss": 0.6719,
+      "step": 4815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1500662853532466,
+      "learning_rate": 1.93302131711041e-06,
+      "loss": 0.679,
+      "step": 4816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.302249606942449,
+      "learning_rate": 1.9329865511920147e-06,
+      "loss": 0.6706,
+      "step": 4817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.364260913368708,
+      "learning_rate": 1.932951776565973e-06,
+      "loss": 0.6803,
+      "step": 4818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.703383425462866,
+      "learning_rate": 1.9329169932326104e-06,
+      "loss": 0.6888,
+      "step": 4819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.661113455527881,
+      "learning_rate": 1.9328822011922505e-06,
+      "loss": 0.7012,
+      "step": 4820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9871203670481905,
+      "learning_rate": 1.9328474004452182e-06,
+      "loss": 0.6882,
+      "step": 4821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1914845884041976,
+      "learning_rate": 1.9328125909918397e-06,
+      "loss": 0.6751,
+      "step": 4822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4761895412445973,
+      "learning_rate": 1.932777772832438e-06,
+      "loss": 0.681,
+      "step": 4823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9710654787820716,
+      "learning_rate": 1.932742945967339e-06,
+      "loss": 0.6862,
+      "step": 4824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.097237621990895,
+      "learning_rate": 1.9327081103968677e-06,
+      "loss": 0.6816,
+      "step": 4825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.97096905301523,
+      "learning_rate": 1.932673266121349e-06,
+      "loss": 0.6745,
+      "step": 4826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.282163856823253,
+      "learning_rate": 1.9326384131411083e-06,
+      "loss": 0.6947,
+      "step": 4827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8740956642757367,
+      "learning_rate": 1.932603551456471e-06,
+      "loss": 0.6999,
+      "step": 4828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3149310785724895,
+      "learning_rate": 1.932568681067762e-06,
+      "loss": 0.6868,
+      "step": 4829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5379296576069397,
+      "learning_rate": 1.9325338019753066e-06,
+      "loss": 0.6849,
+      "step": 4830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4095332363379685,
+      "learning_rate": 1.9324989141794314e-06,
+      "loss": 0.6712,
+      "step": 4831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6025840244580812,
+      "learning_rate": 1.9324640176804613e-06,
+      "loss": 0.6908,
+      "step": 4832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0574032954461843,
+      "learning_rate": 1.9324291124787223e-06,
+      "loss": 0.6823,
+      "step": 4833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8287160340356876,
+      "learning_rate": 1.9323941985745393e-06,
+      "loss": 0.6999,
+      "step": 4834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9647063406388364,
+      "learning_rate": 1.9323592759682392e-06,
+      "loss": 0.6816,
+      "step": 4835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.3024598600067145,
+      "learning_rate": 1.9323243446601473e-06,
+      "loss": 0.6999,
+      "step": 4836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.683719107656228,
+      "learning_rate": 1.93228940465059e-06,
+      "loss": 0.7012,
+      "step": 4837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4945575073555126,
+      "learning_rate": 1.9322544559398933e-06,
+      "loss": 0.7083,
+      "step": 4838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.071414703426855,
+      "learning_rate": 1.932219498528383e-06,
+      "loss": 0.696,
+      "step": 4839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5837215785761619,
+      "learning_rate": 1.9321845324163863e-06,
+      "loss": 0.6829,
+      "step": 4840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.492044713642034,
+      "learning_rate": 1.932149557604228e-06,
+      "loss": 0.6927,
+      "step": 4841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.226776791423348,
+      "learning_rate": 1.932114574092236e-06,
+      "loss": 0.6745,
+      "step": 4842
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3535013555897124,
+      "learning_rate": 1.932079581880737e-06,
+      "loss": 0.6914,
+      "step": 4843
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7618557491594706,
+      "learning_rate": 1.932044580970056e-06,
+      "loss": 0.6823,
+      "step": 4844
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.768398068886179,
+      "learning_rate": 1.93200957136052e-06,
+      "loss": 0.6849,
+      "step": 4845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.900467271572317,
+      "learning_rate": 1.931974553052457e-06,
+      "loss": 0.6862,
+      "step": 4846
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.178028141489832,
+      "learning_rate": 1.9319395260461926e-06,
+      "loss": 0.6947,
+      "step": 4847
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1506840716215048,
+      "learning_rate": 1.931904490342055e-06,
+      "loss": 0.6934,
+      "step": 4848
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.593564323434981,
+      "learning_rate": 1.9318694459403697e-06,
+      "loss": 0.7038,
+      "step": 4849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5813161853309251,
+      "learning_rate": 1.9318343928414642e-06,
+      "loss": 0.6569,
+      "step": 4850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4045850487523526,
+      "learning_rate": 1.9317993310456666e-06,
+      "loss": 0.6764,
+      "step": 4851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.111368666135262,
+      "learning_rate": 1.931764260553303e-06,
+      "loss": 0.6725,
+      "step": 4852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44293673332907635,
+      "learning_rate": 1.9317291813647014e-06,
+      "loss": 0.6823,
+      "step": 4853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.444094241288651,
+      "learning_rate": 1.931694093480189e-06,
+      "loss": 0.696,
+      "step": 4854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3410750405202139,
+      "learning_rate": 1.9316589969000928e-06,
+      "loss": 0.6973,
+      "step": 4855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.514422924009261,
+      "learning_rate": 1.9316238916247412e-06,
+      "loss": 0.666,
+      "step": 4856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.711712311216047,
+      "learning_rate": 1.9315887776544614e-06,
+      "loss": 0.681,
+      "step": 4857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.192810095349653,
+      "learning_rate": 1.9315536549895806e-06,
+      "loss": 0.6849,
+      "step": 4858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.094871938715038,
+      "learning_rate": 1.931518523630428e-06,
+      "loss": 0.6764,
+      "step": 4859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9763434301952265,
+      "learning_rate": 1.9314833835773305e-06,
+      "loss": 0.6797,
+      "step": 4860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5823219461355453,
+      "learning_rate": 1.9314482348306154e-06,
+      "loss": 0.6921,
+      "step": 4861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.667478039476154,
+      "learning_rate": 1.9314130773906124e-06,
+      "loss": 0.6693,
+      "step": 4862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.192927064703819,
+      "learning_rate": 1.9313779112576482e-06,
+      "loss": 0.6784,
+      "step": 4863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7843012355411408,
+      "learning_rate": 1.931342736432052e-06,
+      "loss": 0.681,
+      "step": 4864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3602339497066105,
+      "learning_rate": 1.9313075529141517e-06,
+      "loss": 0.6725,
+      "step": 4865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3795427291566726,
+      "learning_rate": 1.9312723607042754e-06,
+      "loss": 0.6647,
+      "step": 4866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8602730480372975,
+      "learning_rate": 1.931237159802752e-06,
+      "loss": 0.7044,
+      "step": 4867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.458576514069478,
+      "learning_rate": 1.9312019502099095e-06,
+      "loss": 0.681,
+      "step": 4868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.020187800305548,
+      "learning_rate": 1.931166731926077e-06,
+      "loss": 0.6842,
+      "step": 4869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.509785113024289,
+      "learning_rate": 1.9311315049515835e-06,
+      "loss": 0.7057,
+      "step": 4870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.466447460738513,
+      "learning_rate": 1.9310962692867566e-06,
+      "loss": 0.679,
+      "step": 4871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.889253449165543,
+      "learning_rate": 1.9310610249319265e-06,
+      "loss": 0.7057,
+      "step": 4872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0701796699400525,
+      "learning_rate": 1.931025771887421e-06,
+      "loss": 0.6842,
+      "step": 4873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5272820838147309,
+      "learning_rate": 1.93099051015357e-06,
+      "loss": 0.6732,
+      "step": 4874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.803979473935386,
+      "learning_rate": 1.930955239730702e-06,
+      "loss": 0.6992,
+      "step": 4875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7469564985082762,
+      "learning_rate": 1.9309199606191468e-06,
+      "loss": 0.6849,
+      "step": 4876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3342868974996094,
+      "learning_rate": 1.930884672819233e-06,
+      "loss": 0.681,
+      "step": 4877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.552625904880203,
+      "learning_rate": 1.93084937633129e-06,
+      "loss": 0.7188,
+      "step": 4878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7833128443550443,
+      "learning_rate": 1.930814071155648e-06,
+      "loss": 0.7044,
+      "step": 4879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6472803861643478,
+      "learning_rate": 1.930778757292635e-06,
+      "loss": 0.7038,
+      "step": 4880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.20181805763803,
+      "learning_rate": 1.9307434347425826e-06,
+      "loss": 0.6745,
+      "step": 4881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.181166007401924,
+      "learning_rate": 1.9307081035058187e-06,
+      "loss": 0.6947,
+      "step": 4882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8674981130133568,
+      "learning_rate": 1.9306727635826744e-06,
+      "loss": 0.6921,
+      "step": 4883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6744968433119748,
+      "learning_rate": 1.930637414973478e-06,
+      "loss": 0.6732,
+      "step": 4884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.908725200891529,
+      "learning_rate": 1.9306020576785614e-06,
+      "loss": 0.6771,
+      "step": 4885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9640514356983547,
+      "learning_rate": 1.9305666916982526e-06,
+      "loss": 0.6699,
+      "step": 4886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0227575086083238,
+      "learning_rate": 1.930531317032883e-06,
+      "loss": 0.6764,
+      "step": 4887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.502954997791241,
+      "learning_rate": 1.930495933682782e-06,
+      "loss": 0.694,
+      "step": 4888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4755718944077802,
+      "learning_rate": 1.930460541648281e-06,
+      "loss": 0.6797,
+      "step": 4889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8979654587446695,
+      "learning_rate": 1.9304251409297083e-06,
+      "loss": 0.6745,
+      "step": 4890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.409270328869719,
+      "learning_rate": 1.930389731527396e-06,
+      "loss": 0.6849,
+      "step": 4891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5710711223811027,
+      "learning_rate": 1.9303543134416743e-06,
+      "loss": 0.668,
+      "step": 4892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.41211547365088724,
+      "learning_rate": 1.9303188866728733e-06,
+      "loss": 0.6875,
+      "step": 4893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.1757427441753383,
+      "learning_rate": 1.930283451221324e-06,
+      "loss": 0.6751,
+      "step": 4894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.988298893754278,
+      "learning_rate": 1.930248007087357e-06,
+      "loss": 0.6823,
+      "step": 4895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.151908656105882,
+      "learning_rate": 1.930212554271303e-06,
+      "loss": 0.6751,
+      "step": 4896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.727377806565963,
+      "learning_rate": 1.9301770927734928e-06,
+      "loss": 0.6777,
+      "step": 4897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0191023382990247,
+      "learning_rate": 1.9301416225942575e-06,
+      "loss": 0.6882,
+      "step": 4898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.896541971508066,
+      "learning_rate": 1.9301061437339285e-06,
+      "loss": 0.6745,
+      "step": 4899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.046957729426741,
+      "learning_rate": 1.9300706561928366e-06,
+      "loss": 0.6895,
+      "step": 4900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.295547510333853,
+      "learning_rate": 1.930035159971313e-06,
+      "loss": 0.6615,
+      "step": 4901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6533316818745891,
+      "learning_rate": 1.9299996550696886e-06,
+      "loss": 0.6751,
+      "step": 4902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5599859899418057,
+      "learning_rate": 1.9299641414882954e-06,
+      "loss": 0.6764,
+      "step": 4903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.313150212360428,
+      "learning_rate": 1.9299286192274644e-06,
+      "loss": 0.679,
+      "step": 4904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7313283601224674,
+      "learning_rate": 1.9298930882875277e-06,
+      "loss": 0.653,
+      "step": 4905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.955871965122361,
+      "learning_rate": 1.9298575486688165e-06,
+      "loss": 0.6758,
+      "step": 4906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.529941413050044,
+      "learning_rate": 1.9298220003716627e-06,
+      "loss": 0.6751,
+      "step": 4907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5498294659664578,
+      "learning_rate": 1.9297864433963976e-06,
+      "loss": 0.668,
+      "step": 4908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0081631105768638,
+      "learning_rate": 1.9297508777433537e-06,
+      "loss": 0.6634,
+      "step": 4909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.858559082031405,
+      "learning_rate": 1.9297153034128626e-06,
+      "loss": 0.6914,
+      "step": 4910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5780912093885147,
+      "learning_rate": 1.929679720405256e-06,
+      "loss": 0.6927,
+      "step": 4911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3093655424778645,
+      "learning_rate": 1.929644128720867e-06,
+      "loss": 0.6836,
+      "step": 4912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0937248099823926,
+      "learning_rate": 1.929608528360027e-06,
+      "loss": 0.6764,
+      "step": 4913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.714855652708421,
+      "learning_rate": 1.929572919323068e-06,
+      "loss": 0.7116,
+      "step": 4914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.922132561602789,
+      "learning_rate": 1.929537301610323e-06,
+      "loss": 0.6803,
+      "step": 4915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7503868014149488,
+      "learning_rate": 1.929501675222124e-06,
+      "loss": 0.6602,
+      "step": 4916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9095548547436259,
+      "learning_rate": 1.9294660401588034e-06,
+      "loss": 0.6823,
+      "step": 4917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.728892838187432,
+      "learning_rate": 1.9294303964206944e-06,
+      "loss": 0.6947,
+      "step": 4918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.60910156058529,
+      "learning_rate": 1.9293947440081295e-06,
+      "loss": 0.6888,
+      "step": 4919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.188943512551134,
+      "learning_rate": 1.929359082921441e-06,
+      "loss": 0.6523,
+      "step": 4920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7560486432045961,
+      "learning_rate": 1.929323413160962e-06,
+      "loss": 0.6602,
+      "step": 4921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7771123668653983,
+      "learning_rate": 1.9292877347270254e-06,
+      "loss": 0.6777,
+      "step": 4922
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7825665282424902,
+      "learning_rate": 1.9292520476199643e-06,
+      "loss": 0.7038,
+      "step": 4923
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.371672901896353,
+      "learning_rate": 1.9292163518401116e-06,
+      "loss": 0.6738,
+      "step": 4924
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3571344162193255,
+      "learning_rate": 1.9291806473878004e-06,
+      "loss": 0.6745,
+      "step": 4925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.691561697207337,
+      "learning_rate": 1.929144934263364e-06,
+      "loss": 0.6543,
+      "step": 4926
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.346419882685644,
+      "learning_rate": 1.9291092124671364e-06,
+      "loss": 0.6777,
+      "step": 4927
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3210312706564717,
+      "learning_rate": 1.9290734819994495e-06,
+      "loss": 0.6693,
+      "step": 4928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3039413845740808,
+      "learning_rate": 1.929037742860638e-06,
+      "loss": 0.6719,
+      "step": 4929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1881783982580865,
+      "learning_rate": 1.9290019950510352e-06,
+      "loss": 0.6986,
+      "step": 4930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.23541688811473,
+      "learning_rate": 1.9289662385709745e-06,
+      "loss": 0.6823,
+      "step": 4931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7114828035702128,
+      "learning_rate": 1.9289304734207896e-06,
+      "loss": 0.6706,
+      "step": 4932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5801684420637143,
+      "learning_rate": 1.9288946996008147e-06,
+      "loss": 0.6908,
+      "step": 4933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.190670487336772,
+      "learning_rate": 1.928858917111383e-06,
+      "loss": 0.6693,
+      "step": 4934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.77416365660095,
+      "learning_rate": 1.9288231259528295e-06,
+      "loss": 0.6667,
+      "step": 4935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.07134536937542,
+      "learning_rate": 1.928787326125487e-06,
+      "loss": 0.6654,
+      "step": 4936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3966395819139361,
+      "learning_rate": 1.9287515176296905e-06,
+      "loss": 0.6758,
+      "step": 4937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7090847938867517,
+      "learning_rate": 1.928715700465774e-06,
+      "loss": 0.6842,
+      "step": 4938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1289329657048472,
+      "learning_rate": 1.928679874634072e-06,
+      "loss": 0.6719,
+      "step": 4939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.837610139356017,
+      "learning_rate": 1.9286440401349185e-06,
+      "loss": 0.6849,
+      "step": 4940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1912708142563047,
+      "learning_rate": 1.928608196968648e-06,
+      "loss": 0.7044,
+      "step": 4941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.817560046085256,
+      "learning_rate": 1.9285723451355947e-06,
+      "loss": 0.7044,
+      "step": 4942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.225185842434117,
+      "learning_rate": 1.9285364846360943e-06,
+      "loss": 0.6823,
+      "step": 4943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.116987180930424,
+      "learning_rate": 1.92850061547048e-06,
+      "loss": 0.6934,
+      "step": 4944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.964040430071127,
+      "learning_rate": 1.9284647376390878e-06,
+      "loss": 0.7135,
+      "step": 4945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.685634485379869,
+      "learning_rate": 1.9284288511422518e-06,
+      "loss": 0.7174,
+      "step": 4946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.087043876424659,
+      "learning_rate": 1.9283929559803072e-06,
+      "loss": 0.6882,
+      "step": 4947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.573703830216191,
+      "learning_rate": 1.9283570521535896e-06,
+      "loss": 0.6999,
+      "step": 4948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.363711225073967,
+      "learning_rate": 1.928321139662433e-06,
+      "loss": 0.6751,
+      "step": 4949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.082334204755485,
+      "learning_rate": 1.9282852185071735e-06,
+      "loss": 0.6849,
+      "step": 4950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.671989980258979,
+      "learning_rate": 1.9282492886881453e-06,
+      "loss": 0.6693,
+      "step": 4951
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.965944427552028,
+      "learning_rate": 1.928213350205685e-06,
+      "loss": 0.6953,
+      "step": 4952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.389377937417062,
+      "learning_rate": 1.9281774030601274e-06,
+      "loss": 0.6966,
+      "step": 4953
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.20933766820728,
+      "learning_rate": 1.9281414472518076e-06,
+      "loss": 0.6921,
+      "step": 4954
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.795159031252245,
+      "learning_rate": 1.928105482781062e-06,
+      "loss": 0.7201,
+      "step": 4955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5232960396566528,
+      "learning_rate": 1.9280695096482255e-06,
+      "loss": 0.6849,
+      "step": 4956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8629527247831907,
+      "learning_rate": 1.9280335278536345e-06,
+      "loss": 0.6797,
+      "step": 4957
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.427192450883182,
+      "learning_rate": 1.9279975373976247e-06,
+      "loss": 0.666,
+      "step": 4958
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.804231897328308,
+      "learning_rate": 1.9279615382805314e-06,
+      "loss": 0.6868,
+      "step": 4959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.701918667055409,
+      "learning_rate": 1.927925530502691e-06,
+      "loss": 0.6758,
+      "step": 4960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.248470440495258,
+      "learning_rate": 1.92788951406444e-06,
+      "loss": 0.6777,
+      "step": 4961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3519338271586586,
+      "learning_rate": 1.9278534889661135e-06,
+      "loss": 0.6888,
+      "step": 4962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4767752717640106,
+      "learning_rate": 1.9278174552080485e-06,
+      "loss": 0.6797,
+      "step": 4963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9453993855064584,
+      "learning_rate": 1.9277814127905815e-06,
+      "loss": 0.6712,
+      "step": 4964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3102769367643474,
+      "learning_rate": 1.9277453617140482e-06,
+      "loss": 0.6777,
+      "step": 4965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8627439582804972,
+      "learning_rate": 1.9277093019787855e-06,
+      "loss": 0.6875,
+      "step": 4966
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.350142317384139,
+      "learning_rate": 1.92767323358513e-06,
+      "loss": 0.6908,
+      "step": 4967
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.898591886834909,
+      "learning_rate": 1.927637156533418e-06,
+      "loss": 0.6934,
+      "step": 4968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.4887686936169855,
+      "learning_rate": 1.9276010708239863e-06,
+      "loss": 0.6914,
+      "step": 4969
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.897003739104938,
+      "learning_rate": 1.9275649764571716e-06,
+      "loss": 0.6621,
+      "step": 4970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5992636446537338,
+      "learning_rate": 1.9275288734333113e-06,
+      "loss": 0.6745,
+      "step": 4971
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.113937454338086,
+      "learning_rate": 1.927492761752742e-06,
+      "loss": 0.7031,
+      "step": 4972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6428630163077593,
+      "learning_rate": 1.9274566414158008e-06,
+      "loss": 0.6908,
+      "step": 4973
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1770472780381784,
+      "learning_rate": 1.9274205124228243e-06,
+      "loss": 0.6829,
+      "step": 4974
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.612131812646875,
+      "learning_rate": 1.9273843747741507e-06,
+      "loss": 0.679,
+      "step": 4975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8995111366233651,
+      "learning_rate": 1.927348228470116e-06,
+      "loss": 0.6855,
+      "step": 4976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.300950206224386,
+      "learning_rate": 1.9273120735110593e-06,
+      "loss": 0.6771,
+      "step": 4977
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6477723227721746,
+      "learning_rate": 1.927275909897316e-06,
+      "loss": 0.6882,
+      "step": 4978
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8413344042524926,
+      "learning_rate": 1.927239737629225e-06,
+      "loss": 0.6745,
+      "step": 4979
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7923351059468877,
+      "learning_rate": 1.9272035567071236e-06,
+      "loss": 0.6751,
+      "step": 4980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.090852599227851,
+      "learning_rate": 1.9271673671313497e-06,
+      "loss": 0.6842,
+      "step": 4981
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.3562233634599545,
+      "learning_rate": 1.92713116890224e-06,
+      "loss": 0.6745,
+      "step": 4982
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3368601672200886,
+      "learning_rate": 1.927094962020134e-06,
+      "loss": 0.6862,
+      "step": 4983
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2958641298507503,
+      "learning_rate": 1.927058746485368e-06,
+      "loss": 0.666,
+      "step": 4984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6941027017982444,
+      "learning_rate": 1.927022522298281e-06,
+      "loss": 0.6797,
+      "step": 4985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8138429971941274,
+      "learning_rate": 1.926986289459211e-06,
+      "loss": 0.6862,
+      "step": 4986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.344991240428279,
+      "learning_rate": 1.926950047968496e-06,
+      "loss": 0.6999,
+      "step": 4987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4068046796540017,
+      "learning_rate": 1.926913797826474e-06,
+      "loss": 0.6966,
+      "step": 4988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8760816243735755,
+      "learning_rate": 1.9268775390334833e-06,
+      "loss": 0.6699,
+      "step": 4989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8227795777591125,
+      "learning_rate": 1.926841271589863e-06,
+      "loss": 0.7018,
+      "step": 4990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.234012610643305,
+      "learning_rate": 1.9268049954959513e-06,
+      "loss": 0.6719,
+      "step": 4991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.475439457551577,
+      "learning_rate": 1.926768710752086e-06,
+      "loss": 0.6758,
+      "step": 4992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.497804155510151,
+      "learning_rate": 1.9267324173586065e-06,
+      "loss": 0.6882,
+      "step": 4993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7563075304587514,
+      "learning_rate": 1.9266961153158515e-06,
+      "loss": 0.6842,
+      "step": 4994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6852941566759778,
+      "learning_rate": 1.92665980462416e-06,
+      "loss": 0.6517,
+      "step": 4995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.556780286815626,
+      "learning_rate": 1.9266234852838707e-06,
+      "loss": 0.7031,
+      "step": 4996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.30182299073212,
+      "learning_rate": 1.9265871572953216e-06,
+      "loss": 0.709,
+      "step": 4997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9225669512296288,
+      "learning_rate": 1.9265508206588534e-06,
+      "loss": 0.6641,
+      "step": 4998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.665375040077518,
+      "learning_rate": 1.9265144753748042e-06,
+      "loss": 0.6764,
+      "step": 4999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4677024852500926,
+      "learning_rate": 1.9264781214435135e-06,
+      "loss": 0.6784,
+      "step": 5000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.235362007897166,
+      "learning_rate": 1.9264417588653208e-06,
+      "loss": 0.696,
+      "step": 5001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.040671178791531,
+      "learning_rate": 1.926405387640565e-06,
+      "loss": 0.6803,
+      "step": 5002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.680628162912633,
+      "learning_rate": 1.9263690077695856e-06,
+      "loss": 0.6634,
+      "step": 5003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.980207976591543,
+      "learning_rate": 1.9263326192527224e-06,
+      "loss": 0.7005,
+      "step": 5004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.178139501553892,
+      "learning_rate": 1.926296222090315e-06,
+      "loss": 0.6849,
+      "step": 5005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4756439748336987,
+      "learning_rate": 1.926259816282703e-06,
+      "loss": 0.6693,
+      "step": 5006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.251353429792244,
+      "learning_rate": 1.9262234018302265e-06,
+      "loss": 0.6888,
+      "step": 5007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.27773521472061,
+      "learning_rate": 1.9261869787332246e-06,
+      "loss": 0.6855,
+      "step": 5008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4195316758980603,
+      "learning_rate": 1.926150546992038e-06,
+      "loss": 0.6751,
+      "step": 5009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0931471661542207,
+      "learning_rate": 1.9261141066070065e-06,
+      "loss": 0.7012,
+      "step": 5010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8792416090201662,
+      "learning_rate": 1.92607765757847e-06,
+      "loss": 0.681,
+      "step": 5011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.885849906607889,
+      "learning_rate": 1.926041199906769e-06,
+      "loss": 0.6732,
+      "step": 5012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.997705431581159,
+      "learning_rate": 1.9260047335922437e-06,
+      "loss": 0.6895,
+      "step": 5013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9516609621684893,
+      "learning_rate": 1.9259682586352337e-06,
+      "loss": 0.679,
+      "step": 5014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.276436004388301,
+      "learning_rate": 1.925931775036081e-06,
+      "loss": 0.679,
+      "step": 5015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6713646635021844,
+      "learning_rate": 1.9258952827951243e-06,
+      "loss": 0.6842,
+      "step": 5016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0854940687513635,
+      "learning_rate": 1.9258587819127055e-06,
+      "loss": 0.6706,
+      "step": 5017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.475552975476553,
+      "learning_rate": 1.925822272389165e-06,
+      "loss": 0.6901,
+      "step": 5018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1594535894303395,
+      "learning_rate": 1.9257857542248424e-06,
+      "loss": 0.7018,
+      "step": 5019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7205634832668975,
+      "learning_rate": 1.92574922742008e-06,
+      "loss": 0.6751,
+      "step": 5020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9518311929724317,
+      "learning_rate": 1.9257126919752187e-06,
+      "loss": 0.679,
+      "step": 5021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9182452715593398,
+      "learning_rate": 1.9256761478905984e-06,
+      "loss": 0.6855,
+      "step": 5022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0697037667747646,
+      "learning_rate": 1.925639595166561e-06,
+      "loss": 0.679,
+      "step": 5023
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0920663331525802,
+      "learning_rate": 1.9256030338034473e-06,
+      "loss": 0.6712,
+      "step": 5024
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.379844244906793,
+      "learning_rate": 1.9255664638015988e-06,
+      "loss": 0.707,
+      "step": 5025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4610782970706937,
+      "learning_rate": 1.9255298851613567e-06,
+      "loss": 0.6803,
+      "step": 5026
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7365221308191854,
+      "learning_rate": 1.925493297883062e-06,
+      "loss": 0.6738,
+      "step": 5027
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9129392784287338,
+      "learning_rate": 1.9254567019670567e-06,
+      "loss": 0.6992,
+      "step": 5028
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3312896297587973,
+      "learning_rate": 1.9254200974136818e-06,
+      "loss": 0.6979,
+      "step": 5029
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.159286672087882,
+      "learning_rate": 1.9253834842232796e-06,
+      "loss": 0.6764,
+      "step": 5030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.078987409089902,
+      "learning_rate": 1.9253468623961916e-06,
+      "loss": 0.6784,
+      "step": 5031
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.556176306713169,
+      "learning_rate": 1.9253102319327594e-06,
+      "loss": 0.6784,
+      "step": 5032
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.468821417577688,
+      "learning_rate": 1.925273592833325e-06,
+      "loss": 0.6901,
+      "step": 5033
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4087137983407362,
+      "learning_rate": 1.9252369450982303e-06,
+      "loss": 0.6842,
+      "step": 5034
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2078590940117218,
+      "learning_rate": 1.9252002887278173e-06,
+      "loss": 0.6875,
+      "step": 5035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.180172708686467,
+      "learning_rate": 1.925163623722428e-06,
+      "loss": 0.679,
+      "step": 5036
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.888708470428105,
+      "learning_rate": 1.925126950082405e-06,
+      "loss": 0.6875,
+      "step": 5037
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9487924503393504,
+      "learning_rate": 1.9250902678080906e-06,
+      "loss": 0.6686,
+      "step": 5038
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8835830313133762,
+      "learning_rate": 1.925053576899827e-06,
+      "loss": 0.6758,
+      "step": 5039
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.078781603888221,
+      "learning_rate": 1.9250168773579565e-06,
+      "loss": 0.6751,
+      "step": 5040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.390257063566098,
+      "learning_rate": 1.924980169182821e-06,
+      "loss": 0.6901,
+      "step": 5041
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.891990122582198,
+      "learning_rate": 1.9249434523747647e-06,
+      "loss": 0.6895,
+      "step": 5042
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.979403857475553,
+      "learning_rate": 1.9249067269341287e-06,
+      "loss": 0.6797,
+      "step": 5043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.316251558693802,
+      "learning_rate": 1.9248699928612563e-06,
+      "loss": 0.7057,
+      "step": 5044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2879678296380832,
+      "learning_rate": 1.924833250156491e-06,
+      "loss": 0.6686,
+      "step": 5045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3526386780402346,
+      "learning_rate": 1.924796498820175e-06,
+      "loss": 0.6803,
+      "step": 5046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1373946731790063,
+      "learning_rate": 1.9247597388526514e-06,
+      "loss": 0.7044,
+      "step": 5047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.358717341205867,
+      "learning_rate": 1.9247229702542637e-06,
+      "loss": 0.6771,
+      "step": 5048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.60547527930621,
+      "learning_rate": 1.924686193025354e-06,
+      "loss": 0.6784,
+      "step": 5049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.522107764561483,
+      "learning_rate": 1.924649407166267e-06,
+      "loss": 0.6686,
+      "step": 5050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.0058930531065515,
+      "learning_rate": 1.924612612677345e-06,
+      "loss": 0.6992,
+      "step": 5051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.017835360158522,
+      "learning_rate": 1.9245758095589316e-06,
+      "loss": 0.6888,
+      "step": 5052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.268360649033404,
+      "learning_rate": 1.924538997811371e-06,
+      "loss": 0.7168,
+      "step": 5053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5914305988879397,
+      "learning_rate": 1.9245021774350058e-06,
+      "loss": 0.6966,
+      "step": 5054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7640576706324542,
+      "learning_rate": 1.9244653484301797e-06,
+      "loss": 0.6758,
+      "step": 5055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9374689532352106,
+      "learning_rate": 1.924428510797237e-06,
+      "loss": 0.6738,
+      "step": 5056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5484279591033343,
+      "learning_rate": 1.9243916645365217e-06,
+      "loss": 0.6901,
+      "step": 5057
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4922954185377094,
+      "learning_rate": 1.9243548096483766e-06,
+      "loss": 0.6888,
+      "step": 5058
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9842899301560162,
+      "learning_rate": 1.924317946133147e-06,
+      "loss": 0.6842,
+      "step": 5059
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.132164335608764,
+      "learning_rate": 1.9242810739911753e-06,
+      "loss": 0.6875,
+      "step": 5060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2676773428143349,
+      "learning_rate": 1.924244193222807e-06,
+      "loss": 0.6875,
+      "step": 5061
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4363286709270197,
+      "learning_rate": 1.9242073038283864e-06,
+      "loss": 0.6855,
+      "step": 5062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.66501636383612,
+      "learning_rate": 1.924170405808257e-06,
+      "loss": 0.709,
+      "step": 5063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9759044912108429,
+      "learning_rate": 1.9241334991627633e-06,
+      "loss": 0.6882,
+      "step": 5064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.927849892276194,
+      "learning_rate": 1.92409658389225e-06,
+      "loss": 0.6836,
+      "step": 5065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9557717330740916,
+      "learning_rate": 1.9240596599970613e-06,
+      "loss": 0.6875,
+      "step": 5066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.367905457830349,
+      "learning_rate": 1.9240227274775424e-06,
+      "loss": 0.679,
+      "step": 5067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7195188507687895,
+      "learning_rate": 1.923985786334038e-06,
+      "loss": 0.6849,
+      "step": 5068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5439244493506443,
+      "learning_rate": 1.923948836566892e-06,
+      "loss": 0.6712,
+      "step": 5069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.84645698442471,
+      "learning_rate": 1.92391187817645e-06,
+      "loss": 0.6934,
+      "step": 5070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9897220314723627,
+      "learning_rate": 1.9238749111630564e-06,
+      "loss": 0.6855,
+      "step": 5071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.899576100271968,
+      "learning_rate": 1.923837935527057e-06,
+      "loss": 0.6914,
+      "step": 5072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.35769802992667216,
+      "learning_rate": 1.9238009512687964e-06,
+      "loss": 0.6934,
+      "step": 5073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5641032556240462,
+      "learning_rate": 1.9237639583886196e-06,
+      "loss": 0.6829,
+      "step": 5074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9801106135547746,
+      "learning_rate": 1.9237269568868726e-06,
+      "loss": 0.6836,
+      "step": 5075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8544003578366586,
+      "learning_rate": 1.9236899467638997e-06,
+      "loss": 0.6875,
+      "step": 5076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.516142267571946,
+      "learning_rate": 1.923652928020047e-06,
+      "loss": 0.6908,
+      "step": 5077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1846956517658245,
+      "learning_rate": 1.9236159006556596e-06,
+      "loss": 0.6868,
+      "step": 5078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0924800446198513,
+      "learning_rate": 1.9235788646710835e-06,
+      "loss": 0.6927,
+      "step": 5079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5082021854981877,
+      "learning_rate": 1.9235418200666644e-06,
+      "loss": 0.6829,
+      "step": 5080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.445286597021625,
+      "learning_rate": 1.9235047668427474e-06,
+      "loss": 0.651,
+      "step": 5081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.096615740263812,
+      "learning_rate": 1.923467704999679e-06,
+      "loss": 0.6849,
+      "step": 5082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.150325518482485,
+      "learning_rate": 1.923430634537805e-06,
+      "loss": 0.6784,
+      "step": 5083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.155652584983833,
+      "learning_rate": 1.9233935554574707e-06,
+      "loss": 0.6875,
+      "step": 5084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.513793350033659,
+      "learning_rate": 1.923356467759023e-06,
+      "loss": 0.6882,
+      "step": 5085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0487666299749516,
+      "learning_rate": 1.923319371442808e-06,
+      "loss": 0.651,
+      "step": 5086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.481452187750333,
+      "learning_rate": 1.9232822665091715e-06,
+      "loss": 0.6738,
+      "step": 5087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8317638351774193,
+      "learning_rate": 1.9232451529584597e-06,
+      "loss": 0.6823,
+      "step": 5088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6147671711419296,
+      "learning_rate": 1.9232080307910193e-06,
+      "loss": 0.6921,
+      "step": 5089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.703316874442366,
+      "learning_rate": 1.9231709000071968e-06,
+      "loss": 0.6816,
+      "step": 5090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6638682068353373,
+      "learning_rate": 1.9231337606073386e-06,
+      "loss": 0.6992,
+      "step": 5091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0648726866890255,
+      "learning_rate": 1.9230966125917917e-06,
+      "loss": 0.6751,
+      "step": 5092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.0664979226280602,
+      "learning_rate": 1.923059455960902e-06,
+      "loss": 0.6888,
+      "step": 5093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8614393887535705,
+      "learning_rate": 1.923022290715017e-06,
+      "loss": 0.6634,
+      "step": 5094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1719576239263763,
+      "learning_rate": 1.9229851168544833e-06,
+      "loss": 0.6823,
+      "step": 5095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.370267530443113,
+      "learning_rate": 1.922947934379648e-06,
+      "loss": 0.7044,
+      "step": 5096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.177519122869517,
+      "learning_rate": 1.922910743290858e-06,
+      "loss": 0.6803,
+      "step": 5097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9738941242934729,
+      "learning_rate": 1.9228735435884606e-06,
+      "loss": 0.6868,
+      "step": 5098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.599335711753268,
+      "learning_rate": 1.922836335272802e-06,
+      "loss": 0.7083,
+      "step": 5099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0898182652425534,
+      "learning_rate": 1.9227991183442312e-06,
+      "loss": 0.7018,
+      "step": 5100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9550285927866295,
+      "learning_rate": 1.922761892803094e-06,
+      "loss": 0.679,
+      "step": 5101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5512685273157563,
+      "learning_rate": 1.9227246586497388e-06,
+      "loss": 0.679,
+      "step": 5102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.176006293319146,
+      "learning_rate": 1.9226874158845125e-06,
+      "loss": 0.6895,
+      "step": 5103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.464016722144726,
+      "learning_rate": 1.922650164507763e-06,
+      "loss": 0.6927,
+      "step": 5104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.765636745534902,
+      "learning_rate": 1.922612904519838e-06,
+      "loss": 0.6738,
+      "step": 5105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.158399617529925,
+      "learning_rate": 1.922575635921085e-06,
+      "loss": 0.6621,
+      "step": 5106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9302735844059835,
+      "learning_rate": 1.9225383587118522e-06,
+      "loss": 0.6855,
+      "step": 5107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.076948147620345,
+      "learning_rate": 1.9225010728924876e-06,
+      "loss": 0.6699,
+      "step": 5108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.964419031790629,
+      "learning_rate": 1.9224637784633384e-06,
+      "loss": 0.7064,
+      "step": 5109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1877509814991303,
+      "learning_rate": 1.9224264754247536e-06,
+      "loss": 0.722,
+      "step": 5110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8432882536882317,
+      "learning_rate": 1.9223891637770805e-06,
+      "loss": 0.666,
+      "step": 5111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.8028406646836785,
+      "learning_rate": 1.9223518435206677e-06,
+      "loss": 0.6914,
+      "step": 5112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.754136568555937,
+      "learning_rate": 1.922314514655864e-06,
+      "loss": 0.6953,
+      "step": 5113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.129251857541044,
+      "learning_rate": 1.922277177183017e-06,
+      "loss": 0.7018,
+      "step": 5114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.698065941617655,
+      "learning_rate": 1.922239831102476e-06,
+      "loss": 0.6712,
+      "step": 5115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4267135470626373,
+      "learning_rate": 1.9222024764145884e-06,
+      "loss": 0.6921,
+      "step": 5116
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1564375085516394,
+      "learning_rate": 1.9221651131197044e-06,
+      "loss": 0.6966,
+      "step": 5117
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.195961346495355,
+      "learning_rate": 1.9221277412181715e-06,
+      "loss": 0.6868,
+      "step": 5118
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.251588066294339,
+      "learning_rate": 1.9220903607103383e-06,
+      "loss": 0.7051,
+      "step": 5119
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.400041616204578,
+      "learning_rate": 1.922052971596555e-06,
+      "loss": 0.6836,
+      "step": 5120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7798029255793986,
+      "learning_rate": 1.922015573877169e-06,
+      "loss": 0.6986,
+      "step": 5121
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7199763748484393,
+      "learning_rate": 1.9219781675525307e-06,
+      "loss": 0.6797,
+      "step": 5122
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.415319209793105,
+      "learning_rate": 1.9219407526229887e-06,
+      "loss": 0.6777,
+      "step": 5123
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.450410273717887,
+      "learning_rate": 1.921903329088892e-06,
+      "loss": 0.681,
+      "step": 5124
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.42904463676618404,
+      "learning_rate": 1.9218658969505896e-06,
+      "loss": 0.6823,
+      "step": 5125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2359442415492023,
+      "learning_rate": 1.9218284562084318e-06,
+      "loss": 0.6797,
+      "step": 5126
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3816113996183385,
+      "learning_rate": 1.921791006862767e-06,
+      "loss": 0.6836,
+      "step": 5127
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3974052019865893,
+      "learning_rate": 1.9217535489139456e-06,
+      "loss": 0.6784,
+      "step": 5128
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.347835601366578,
+      "learning_rate": 1.9217160823623165e-06,
+      "loss": 0.6836,
+      "step": 5129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.438141495820985,
+      "learning_rate": 1.92167860720823e-06,
+      "loss": 0.6745,
+      "step": 5130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5378566037000513,
+      "learning_rate": 1.9216411234520354e-06,
+      "loss": 0.6764,
+      "step": 5131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8068297991961422,
+      "learning_rate": 1.921603631094083e-06,
+      "loss": 0.6686,
+      "step": 5132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8653719826809416,
+      "learning_rate": 1.921566130134722e-06,
+      "loss": 0.6908,
+      "step": 5133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5350584471433684,
+      "learning_rate": 1.921528620574303e-06,
+      "loss": 0.6947,
+      "step": 5134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.652431647328719,
+      "learning_rate": 1.9214911024131757e-06,
+      "loss": 0.6901,
+      "step": 5135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0279637099977132,
+      "learning_rate": 1.9214535756516906e-06,
+      "loss": 0.6927,
+      "step": 5136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.879161441808333,
+      "learning_rate": 1.9214160402901974e-06,
+      "loss": 0.6842,
+      "step": 5137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6630115003699333,
+      "learning_rate": 1.921378496329047e-06,
+      "loss": 0.6908,
+      "step": 5138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.577469028790561,
+      "learning_rate": 1.92134094376859e-06,
+      "loss": 0.6914,
+      "step": 5139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7176898854911028,
+      "learning_rate": 1.9213033826091763e-06,
+      "loss": 0.6855,
+      "step": 5140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.069033149795668,
+      "learning_rate": 1.921265812851157e-06,
+      "loss": 0.6595,
+      "step": 5141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6987482834839938,
+      "learning_rate": 1.9212282344948816e-06,
+      "loss": 0.6849,
+      "step": 5142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0777941495240506,
+      "learning_rate": 1.921190647540702e-06,
+      "loss": 0.6654,
+      "step": 5143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9767904417191738,
+      "learning_rate": 1.921153051988969e-06,
+      "loss": 0.668,
+      "step": 5144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.6471950457733975,
+      "learning_rate": 1.9211154478400326e-06,
+      "loss": 0.696,
+      "step": 5145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.368620248420106,
+      "learning_rate": 1.9210778350942446e-06,
+      "loss": 0.6868,
+      "step": 5146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.406773833897521,
+      "learning_rate": 1.9210402137519556e-06,
+      "loss": 0.6797,
+      "step": 5147
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7296456720431217,
+      "learning_rate": 1.9210025838135168e-06,
+      "loss": 0.6647,
+      "step": 5148
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2679879012481376,
+      "learning_rate": 1.920964945279279e-06,
+      "loss": 0.6569,
+      "step": 5149
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.580213805994913,
+      "learning_rate": 1.9209272981495947e-06,
+      "loss": 0.6784,
+      "step": 5150
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6671725614384765,
+      "learning_rate": 1.920889642424814e-06,
+      "loss": 0.6934,
+      "step": 5151
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3314603902642217,
+      "learning_rate": 1.920851978105289e-06,
+      "loss": 0.668,
+      "step": 5152
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.492567497229608,
+      "learning_rate": 1.9208143051913706e-06,
+      "loss": 0.7051,
+      "step": 5153
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0690737422091536,
+      "learning_rate": 1.9207766236834115e-06,
+      "loss": 0.6719,
+      "step": 5154
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.206787292206513,
+      "learning_rate": 1.9207389335817627e-06,
+      "loss": 0.6699,
+      "step": 5155
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.349409370228818,
+      "learning_rate": 1.9207012348867754e-06,
+      "loss": 0.7083,
+      "step": 5156
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.087562509256677,
+      "learning_rate": 1.9206635275988028e-06,
+      "loss": 0.6849,
+      "step": 5157
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6183786906549782,
+      "learning_rate": 1.9206258117181954e-06,
+      "loss": 0.6914,
+      "step": 5158
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2032132540721903,
+      "learning_rate": 1.920588087245306e-06,
+      "loss": 0.6842,
+      "step": 5159
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7948104076811605,
+      "learning_rate": 1.920550354180487e-06,
+      "loss": 0.6621,
+      "step": 5160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.153245675561223,
+      "learning_rate": 1.92051261252409e-06,
+      "loss": 0.7292,
+      "step": 5161
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.169394859917476,
+      "learning_rate": 1.9204748622764675e-06,
+      "loss": 0.6784,
+      "step": 5162
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.407339503085876,
+      "learning_rate": 1.9204371034379714e-06,
+      "loss": 0.6966,
+      "step": 5163
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7385966937479674,
+      "learning_rate": 1.9203993360089544e-06,
+      "loss": 0.6966,
+      "step": 5164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5201668655439455,
+      "learning_rate": 1.9203615599897693e-06,
+      "loss": 0.6829,
+      "step": 5165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.89701701905455,
+      "learning_rate": 1.920323775380768e-06,
+      "loss": 0.6836,
+      "step": 5166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.552017736013891,
+      "learning_rate": 1.9202859821823038e-06,
+      "loss": 0.6615,
+      "step": 5167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0013523298104055,
+      "learning_rate": 1.920248180394729e-06,
+      "loss": 0.6784,
+      "step": 5168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7027260826675716,
+      "learning_rate": 1.9202103700183967e-06,
+      "loss": 0.6875,
+      "step": 5169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5259123602246278,
+      "learning_rate": 1.9201725510536594e-06,
+      "loss": 0.6589,
+      "step": 5170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.827988211831475,
+      "learning_rate": 1.9201347235008704e-06,
+      "loss": 0.6842,
+      "step": 5171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.71231156332935,
+      "learning_rate": 1.9200968873603827e-06,
+      "loss": 0.6842,
+      "step": 5172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8925409277378984,
+      "learning_rate": 1.9200590426325493e-06,
+      "loss": 0.6777,
+      "step": 5173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5074998039780627,
+      "learning_rate": 1.9200211893177237e-06,
+      "loss": 0.6569,
+      "step": 5174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4832061483873415,
+      "learning_rate": 1.9199833274162587e-06,
+      "loss": 0.6706,
+      "step": 5175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9575151831079998,
+      "learning_rate": 1.919945456928508e-06,
+      "loss": 0.6934,
+      "step": 5176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4786507388511414,
+      "learning_rate": 1.919907577854825e-06,
+      "loss": 0.6758,
+      "step": 5177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5983247167125327,
+      "learning_rate": 1.9198696901955634e-06,
+      "loss": 0.6725,
+      "step": 5178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0549328915758487,
+      "learning_rate": 1.919831793951077e-06,
+      "loss": 0.6745,
+      "step": 5179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3648938759223146,
+      "learning_rate": 1.919793889121718e-06,
+      "loss": 0.681,
+      "step": 5180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.403790200164839,
+      "learning_rate": 1.9197559757078424e-06,
+      "loss": 0.6934,
+      "step": 5181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.11836677863783,
+      "learning_rate": 1.9197180537098024e-06,
+      "loss": 0.7311,
+      "step": 5182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.414646113241014,
+      "learning_rate": 1.919680123127953e-06,
+      "loss": 0.7064,
+      "step": 5183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.204501634827672,
+      "learning_rate": 1.919642183962647e-06,
+      "loss": 0.6784,
+      "step": 5184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0108367609058428,
+      "learning_rate": 1.919604236214239e-06,
+      "loss": 0.6875,
+      "step": 5185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.415862161554144,
+      "learning_rate": 1.919566279883084e-06,
+      "loss": 0.6908,
+      "step": 5186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6080045967786997,
+      "learning_rate": 1.9195283149695355e-06,
+      "loss": 0.6842,
+      "step": 5187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.230580366397807,
+      "learning_rate": 1.9194903414739476e-06,
+      "loss": 0.6999,
+      "step": 5188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.01421943827295,
+      "learning_rate": 1.9194523593966754e-06,
+      "loss": 0.7135,
+      "step": 5189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.725345443931961,
+      "learning_rate": 1.9194143687380726e-06,
+      "loss": 0.709,
+      "step": 5190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.24374121028374,
+      "learning_rate": 1.919376369498494e-06,
+      "loss": 0.6973,
+      "step": 5191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4779383777087123,
+      "learning_rate": 1.919338361678295e-06,
+      "loss": 0.6882,
+      "step": 5192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.2791807040524015,
+      "learning_rate": 1.9193003452778296e-06,
+      "loss": 0.6908,
+      "step": 5193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.786264890643834,
+      "learning_rate": 1.9192623202974527e-06,
+      "loss": 0.6953,
+      "step": 5194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.790419040334855,
+      "learning_rate": 1.919224286737519e-06,
+      "loss": 0.7064,
+      "step": 5195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2549362854425765,
+      "learning_rate": 1.9191862445983846e-06,
+      "loss": 0.6777,
+      "step": 5196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.1796302284878175,
+      "learning_rate": 1.919148193880403e-06,
+      "loss": 0.6602,
+      "step": 5197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9709865988004287,
+      "learning_rate": 1.9191101345839302e-06,
+      "loss": 0.6777,
+      "step": 5198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.337124593507324,
+      "learning_rate": 1.919072066709321e-06,
+      "loss": 0.6816,
+      "step": 5199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2278142648941435,
+      "learning_rate": 1.9190339902569314e-06,
+      "loss": 0.681,
+      "step": 5200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3178904478544053,
+      "learning_rate": 1.9189959052271163e-06,
+      "loss": 0.6842,
+      "step": 5201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9592556234508612,
+      "learning_rate": 1.9189578116202307e-06,
+      "loss": 0.6849,
+      "step": 5202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1412437432476477,
+      "learning_rate": 1.9189197094366304e-06,
+      "loss": 0.6732,
+      "step": 5203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.984087261848875,
+      "learning_rate": 1.9188815986766717e-06,
+      "loss": 0.7096,
+      "step": 5204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.251028001144143,
+      "learning_rate": 1.9188434793407097e-06,
+      "loss": 0.6908,
+      "step": 5205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.233816076135131,
+      "learning_rate": 1.9188053514291e-06,
+      "loss": 0.6875,
+      "step": 5206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4467441939200607,
+      "learning_rate": 1.918767214942199e-06,
+      "loss": 0.679,
+      "step": 5207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.426166902273389,
+      "learning_rate": 1.9187290698803623e-06,
+      "loss": 0.6836,
+      "step": 5208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0403718120229577,
+      "learning_rate": 1.918690916243946e-06,
+      "loss": 0.6849,
+      "step": 5209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.656421133283909,
+      "learning_rate": 1.9186527540333056e-06,
+      "loss": 0.6803,
+      "step": 5210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.775241068350747,
+      "learning_rate": 1.9186145832487984e-06,
+      "loss": 0.6784,
+      "step": 5211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.593782398176366,
+      "learning_rate": 1.91857640389078e-06,
+      "loss": 0.6992,
+      "step": 5212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.0268466256823021,
+      "learning_rate": 1.9185382159596063e-06,
+      "loss": 0.6816,
+      "step": 5213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.836752197483894,
+      "learning_rate": 1.918500019455635e-06,
+      "loss": 0.6934,
+      "step": 5214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5497102406794387,
+      "learning_rate": 1.918461814379221e-06,
+      "loss": 0.7025,
+      "step": 5215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4426955628474918,
+      "learning_rate": 1.9184236007307215e-06,
+      "loss": 0.6615,
+      "step": 5216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5875161068361667,
+      "learning_rate": 1.9183853785104943e-06,
+      "loss": 0.696,
+      "step": 5217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.526577055827364,
+      "learning_rate": 1.9183471477188946e-06,
+      "loss": 0.6908,
+      "step": 5218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7313920883726799,
+      "learning_rate": 1.9183089083562795e-06,
+      "loss": 0.6836,
+      "step": 5219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0074573706843046,
+      "learning_rate": 1.9182706604230068e-06,
+      "loss": 0.6641,
+      "step": 5220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9659421020134467,
+      "learning_rate": 1.9182324039194323e-06,
+      "loss": 0.6745,
+      "step": 5221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1290928762331447,
+      "learning_rate": 1.9181941388459134e-06,
+      "loss": 0.6829,
+      "step": 5222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.580297989375622,
+      "learning_rate": 1.918155865202808e-06,
+      "loss": 0.7051,
+      "step": 5223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9765386971620598,
+      "learning_rate": 1.9181175829904724e-06,
+      "loss": 0.6823,
+      "step": 5224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5185311776456006,
+      "learning_rate": 1.9180792922092643e-06,
+      "loss": 0.6751,
+      "step": 5225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.022780066980127,
+      "learning_rate": 1.918040992859541e-06,
+      "loss": 0.696,
+      "step": 5226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.130163845614596,
+      "learning_rate": 1.9180026849416598e-06,
+      "loss": 0.6582,
+      "step": 5227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.995861165441475,
+      "learning_rate": 1.917964368455978e-06,
+      "loss": 0.6803,
+      "step": 5228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3214532064959865,
+      "learning_rate": 1.917926043402854e-06,
+      "loss": 0.6758,
+      "step": 5229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6320866197685475,
+      "learning_rate": 1.9178877097826454e-06,
+      "loss": 0.6908,
+      "step": 5230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.033214131841345,
+      "learning_rate": 1.917849367595709e-06,
+      "loss": 0.6686,
+      "step": 5231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2068698611908288,
+      "learning_rate": 1.9178110168424037e-06,
+      "loss": 0.6908,
+      "step": 5232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.677288833735646,
+      "learning_rate": 1.9177726575230868e-06,
+      "loss": 0.7018,
+      "step": 5233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7747602144250911,
+      "learning_rate": 1.9177342896381166e-06,
+      "loss": 0.6803,
+      "step": 5234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5969152241165303,
+      "learning_rate": 1.9176959131878513e-06,
+      "loss": 0.6849,
+      "step": 5235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.545123520734429,
+      "learning_rate": 1.9176575281726486e-06,
+      "loss": 0.6738,
+      "step": 5236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9536482287193089,
+      "learning_rate": 1.917619134592867e-06,
+      "loss": 0.6667,
+      "step": 5237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9253076182618392,
+      "learning_rate": 1.9175807324488653e-06,
+      "loss": 0.6895,
+      "step": 5238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.914853437550412,
+      "learning_rate": 1.917542321741001e-06,
+      "loss": 0.6829,
+      "step": 5239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3136230302829746,
+      "learning_rate": 1.917503902469633e-06,
+      "loss": 0.6934,
+      "step": 5240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7862442091318633,
+      "learning_rate": 1.91746547463512e-06,
+      "loss": 0.6699,
+      "step": 5241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.8683164058673105,
+      "learning_rate": 1.9174270382378207e-06,
+      "loss": 0.6725,
+      "step": 5242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6709064552454,
+      "learning_rate": 1.917388593278094e-06,
+      "loss": 0.6784,
+      "step": 5243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9736205452264726,
+      "learning_rate": 1.917350139756298e-06,
+      "loss": 0.6784,
+      "step": 5244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2535789510524122,
+      "learning_rate": 1.917311677672792e-06,
+      "loss": 0.6608,
+      "step": 5245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.141781675782395,
+      "learning_rate": 1.9172732070279357e-06,
+      "loss": 0.6803,
+      "step": 5246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.701634714227708,
+      "learning_rate": 1.917234727822087e-06,
+      "loss": 0.7155,
+      "step": 5247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.889265067961097,
+      "learning_rate": 1.917196240055605e-06,
+      "loss": 0.6947,
+      "step": 5248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.70915310801982,
+      "learning_rate": 1.91715774372885e-06,
+      "loss": 0.7168,
+      "step": 5249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.3154800181256245,
+      "learning_rate": 1.9171192388421807e-06,
+      "loss": 0.6706,
+      "step": 5250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6917591974403074,
+      "learning_rate": 1.917080725395956e-06,
+      "loss": 0.6803,
+      "step": 5251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.908787650576514,
+      "learning_rate": 1.917042203390536e-06,
+      "loss": 0.707,
+      "step": 5252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.040648821685389,
+      "learning_rate": 1.91700367282628e-06,
+      "loss": 0.6973,
+      "step": 5253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.010462416914631,
+      "learning_rate": 1.9169651337035473e-06,
+      "loss": 0.6927,
+      "step": 5254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2680119221735622,
+      "learning_rate": 1.9169265860226987e-06,
+      "loss": 0.6647,
+      "step": 5255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.285494867645112,
+      "learning_rate": 1.916888029784093e-06,
+      "loss": 0.6803,
+      "step": 5256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.918505295409657,
+      "learning_rate": 1.9168494649880896e-06,
+      "loss": 0.7025,
+      "step": 5257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.658532867539108,
+      "learning_rate": 1.9168108916350492e-06,
+      "loss": 0.6921,
+      "step": 5258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.272965671237271,
+      "learning_rate": 1.916772309725332e-06,
+      "loss": 0.7044,
+      "step": 5259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.493318732001443,
+      "learning_rate": 1.9167337192592978e-06,
+      "loss": 0.6842,
+      "step": 5260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8061856568224763,
+      "learning_rate": 1.9166951202373063e-06,
+      "loss": 0.6849,
+      "step": 5261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0656746380419118,
+      "learning_rate": 1.916656512659719e-06,
+      "loss": 0.6882,
+      "step": 5262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.182325528756099,
+      "learning_rate": 1.9166178965268944e-06,
+      "loss": 0.6803,
+      "step": 5263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.970524274152441,
+      "learning_rate": 1.9165792718391944e-06,
+      "loss": 0.6862,
+      "step": 5264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1149962621253136,
+      "learning_rate": 1.916540638596979e-06,
+      "loss": 0.681,
+      "step": 5265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.415960623323468,
+      "learning_rate": 1.916501996800609e-06,
+      "loss": 0.6908,
+      "step": 5266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3872293135724258,
+      "learning_rate": 1.916463346450444e-06,
+      "loss": 0.6816,
+      "step": 5267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0939619116808057,
+      "learning_rate": 1.9164246875468464e-06,
+      "loss": 0.6771,
+      "step": 5268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5416882648567247,
+      "learning_rate": 1.9163860200901756e-06,
+      "loss": 0.6725,
+      "step": 5269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.666106239728497,
+      "learning_rate": 1.9163473440807934e-06,
+      "loss": 0.6751,
+      "step": 5270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.412183348302087,
+      "learning_rate": 1.9163086595190605e-06,
+      "loss": 0.6738,
+      "step": 5271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.880662307553055,
+      "learning_rate": 1.9162699664053373e-06,
+      "loss": 0.6764,
+      "step": 5272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4360233820959825,
+      "learning_rate": 1.916231264739986e-06,
+      "loss": 0.6855,
+      "step": 5273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7942903640245291,
+      "learning_rate": 1.916192554523367e-06,
+      "loss": 0.6758,
+      "step": 5274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5170136060518417,
+      "learning_rate": 1.9161538357558417e-06,
+      "loss": 0.6751,
+      "step": 5275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.316361163671118,
+      "learning_rate": 1.916115108437772e-06,
+      "loss": 0.6628,
+      "step": 5276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.719055550256077,
+      "learning_rate": 1.916076372569519e-06,
+      "loss": 0.6647,
+      "step": 5277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5550965442359631,
+      "learning_rate": 1.916037628151444e-06,
+      "loss": 0.6901,
+      "step": 5278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.631050830378552,
+      "learning_rate": 1.915998875183909e-06,
+      "loss": 0.6888,
+      "step": 5279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5204219917111207,
+      "learning_rate": 1.915960113667275e-06,
+      "loss": 0.6875,
+      "step": 5280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.197663041055757,
+      "learning_rate": 1.9159213436019048e-06,
+      "loss": 0.6608,
+      "step": 5281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4816972804690998,
+      "learning_rate": 1.9158825649881592e-06,
+      "loss": 0.6725,
+      "step": 5282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.859748873202748,
+      "learning_rate": 1.915843777826401e-06,
+      "loss": 0.6908,
+      "step": 5283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3738369645438135,
+      "learning_rate": 1.9158049821169918e-06,
+      "loss": 0.696,
+      "step": 5284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3946684401357903,
+      "learning_rate": 1.9157661778602936e-06,
+      "loss": 0.6621,
+      "step": 5285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.658156802998355,
+      "learning_rate": 1.915727365056669e-06,
+      "loss": 0.6654,
+      "step": 5286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.273495561876913,
+      "learning_rate": 1.9156885437064795e-06,
+      "loss": 0.6882,
+      "step": 5287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8730964249208273,
+      "learning_rate": 1.915649713810088e-06,
+      "loss": 0.6764,
+      "step": 5288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6957977434864091,
+      "learning_rate": 1.9156108753678568e-06,
+      "loss": 0.6686,
+      "step": 5289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2821212744970933,
+      "learning_rate": 1.915572028380148e-06,
+      "loss": 0.6901,
+      "step": 5290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.0016874962768085,
+      "learning_rate": 1.915533172847325e-06,
+      "loss": 0.6855,
+      "step": 5291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4925523149771247,
+      "learning_rate": 1.9154943087697495e-06,
+      "loss": 0.6986,
+      "step": 5292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.565772404423167,
+      "learning_rate": 1.915455436147785e-06,
+      "loss": 0.6875,
+      "step": 5293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.479693330475718,
+      "learning_rate": 1.915416554981794e-06,
+      "loss": 0.666,
+      "step": 5294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.432220701239354,
+      "learning_rate": 1.915377665272139e-06,
+      "loss": 0.7227,
+      "step": 5295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4716210598302022,
+      "learning_rate": 1.9153387670191834e-06,
+      "loss": 0.6934,
+      "step": 5296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7209885257498145,
+      "learning_rate": 1.9152998602232898e-06,
+      "loss": 0.6927,
+      "step": 5297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2204283054769767,
+      "learning_rate": 1.9152609448848224e-06,
+      "loss": 0.6908,
+      "step": 5298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4519340467516675,
+      "learning_rate": 1.9152220210041433e-06,
+      "loss": 0.653,
+      "step": 5299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.270508461687908,
+      "learning_rate": 1.915183088581616e-06,
+      "loss": 0.679,
+      "step": 5300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8152637561545615,
+      "learning_rate": 1.915144147617604e-06,
+      "loss": 0.6536,
+      "step": 5301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5082242778265713,
+      "learning_rate": 1.915105198112471e-06,
+      "loss": 0.6784,
+      "step": 5302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5566355306947642,
+      "learning_rate": 1.91506624006658e-06,
+      "loss": 0.6836,
+      "step": 5303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.665023721051778,
+      "learning_rate": 1.9150272734802952e-06,
+      "loss": 0.6543,
+      "step": 5304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.781433302139859,
+      "learning_rate": 1.9149882983539797e-06,
+      "loss": 0.6921,
+      "step": 5305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.102891479409787,
+      "learning_rate": 1.9149493146879975e-06,
+      "loss": 0.6921,
+      "step": 5306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.8749213259659145,
+      "learning_rate": 1.9149103224827123e-06,
+      "loss": 0.6667,
+      "step": 5307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.077293123415245,
+      "learning_rate": 1.9148713217384883e-06,
+      "loss": 0.6673,
+      "step": 5308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.169989496138715,
+      "learning_rate": 1.9148323124556895e-06,
+      "loss": 0.6947,
+      "step": 5309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.46093231984859,
+      "learning_rate": 1.9147932946346796e-06,
+      "loss": 0.6875,
+      "step": 5310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.638656427932575,
+      "learning_rate": 1.9147542682758227e-06,
+      "loss": 0.6484,
+      "step": 5311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2558826446806655,
+      "learning_rate": 1.9147152333794836e-06,
+      "loss": 0.6888,
+      "step": 5312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7210099547041231,
+      "learning_rate": 1.9146761899460265e-06,
+      "loss": 0.679,
+      "step": 5313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7445369107447437,
+      "learning_rate": 1.9146371379758156e-06,
+      "loss": 0.6569,
+      "step": 5314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.283186825976683,
+      "learning_rate": 1.9145980774692156e-06,
+      "loss": 0.6758,
+      "step": 5315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8484434690737648,
+      "learning_rate": 1.9145590084265906e-06,
+      "loss": 0.6868,
+      "step": 5316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.127849239617561,
+      "learning_rate": 1.9145199308483053e-06,
+      "loss": 0.6816,
+      "step": 5317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.252005523332189,
+      "learning_rate": 1.914480844734725e-06,
+      "loss": 0.6973,
+      "step": 5318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.741612831472131,
+      "learning_rate": 1.9144417500862134e-06,
+      "loss": 0.707,
+      "step": 5319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.801829654910554,
+      "learning_rate": 1.914402646903137e-06,
+      "loss": 0.6712,
+      "step": 5320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5623531298403651,
+      "learning_rate": 1.914363535185859e-06,
+      "loss": 0.6895,
+      "step": 5321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7383243392260396,
+      "learning_rate": 1.914324414934746e-06,
+      "loss": 0.7018,
+      "step": 5322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3067305133821983,
+      "learning_rate": 1.9142852861501622e-06,
+      "loss": 0.6862,
+      "step": 5323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6765967510045325,
+      "learning_rate": 1.914246148832473e-06,
+      "loss": 0.6621,
+      "step": 5324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.106651762843581,
+      "learning_rate": 1.9142070029820432e-06,
+      "loss": 0.6979,
+      "step": 5325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4214601018212831,
+      "learning_rate": 1.9141678485992386e-06,
+      "loss": 0.6803,
+      "step": 5326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2960984626937355,
+      "learning_rate": 1.914128685684425e-06,
+      "loss": 0.6855,
+      "step": 5327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.153183350090814,
+      "learning_rate": 1.9140895142379674e-06,
+      "loss": 0.6823,
+      "step": 5328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1486123137013178,
+      "learning_rate": 1.9140503342602315e-06,
+      "loss": 0.6816,
+      "step": 5329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7171794153103246,
+      "learning_rate": 1.9140111457515832e-06,
+      "loss": 0.6895,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.079561887053426,
+      "learning_rate": 1.913971948712388e-06,
+      "loss": 0.6803,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.0161156232739295,
+      "learning_rate": 1.9139327431430117e-06,
+      "loss": 0.6784,
+      "step": 5332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.587731313638269,
+      "learning_rate": 1.91389352904382e-06,
+      "loss": 0.6784,
+      "step": 5333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5342202865902017,
+      "learning_rate": 1.913854306415179e-06,
+      "loss": 0.6895,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.797199066881254,
+      "learning_rate": 1.9138150752574556e-06,
+      "loss": 0.7051,
+      "step": 5335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.773504582919025,
+      "learning_rate": 1.9137758355710153e-06,
+      "loss": 0.6836,
+      "step": 5336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.224965559867011,
+      "learning_rate": 1.9137365873562238e-06,
+      "loss": 0.6803,
+      "step": 5337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6001877528190334,
+      "learning_rate": 1.913697330613448e-06,
+      "loss": 0.6797,
+      "step": 5338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6132890414950464,
+      "learning_rate": 1.913658065343055e-06,
+      "loss": 0.6777,
+      "step": 5339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7805824530101828,
+      "learning_rate": 1.9136187915454095e-06,
+      "loss": 0.6901,
+      "step": 5340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9730190710510055,
+      "learning_rate": 1.9135795092208794e-06,
+      "loss": 0.6816,
+      "step": 5341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6385307753946438,
+      "learning_rate": 1.9135402183698306e-06,
+      "loss": 0.6947,
+      "step": 5342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.318265106387141,
+      "learning_rate": 1.913500918992631e-06,
+      "loss": 0.7012,
+      "step": 5343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5337561410901617,
+      "learning_rate": 1.913461611089646e-06,
+      "loss": 0.6842,
+      "step": 5344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8099436391747247,
+      "learning_rate": 1.913422294661243e-06,
+      "loss": 0.6745,
+      "step": 5345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5390630386747435,
+      "learning_rate": 1.913382969707789e-06,
+      "loss": 0.6842,
+      "step": 5346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.8294675761157695,
+      "learning_rate": 1.913343636229651e-06,
+      "loss": 0.6999,
+      "step": 5347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.247949305267284,
+      "learning_rate": 1.913304294227196e-06,
+      "loss": 0.6836,
+      "step": 5348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.699146846387004,
+      "learning_rate": 1.913264943700791e-06,
+      "loss": 0.6966,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6646424836688758,
+      "learning_rate": 1.913225584650804e-06,
+      "loss": 0.694,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.51395552537227,
+      "learning_rate": 1.9131862170776014e-06,
+      "loss": 0.6732,
+      "step": 5351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2922720442069453,
+      "learning_rate": 1.913146840981551e-06,
+      "loss": 0.6914,
+      "step": 5352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.34595253883062527,
+      "learning_rate": 1.9131074563630213e-06,
+      "loss": 0.6797,
+      "step": 5353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.200146886002653,
+      "learning_rate": 1.913068063222378e-06,
+      "loss": 0.6855,
+      "step": 5354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5326391770291503,
+      "learning_rate": 1.91302866155999e-06,
+      "loss": 0.6589,
+      "step": 5355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6851408250296119,
+      "learning_rate": 1.912989251376225e-06,
+      "loss": 0.681,
+      "step": 5356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5101138886230872,
+      "learning_rate": 1.9129498326714505e-06,
+      "loss": 0.694,
+      "step": 5357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.736553762294813,
+      "learning_rate": 1.9129104054460343e-06,
+      "loss": 0.6797,
+      "step": 5358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5324244521895196,
+      "learning_rate": 1.9128709697003446e-06,
+      "loss": 0.6908,
+      "step": 5359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3702247524028457,
+      "learning_rate": 1.9128315254347494e-06,
+      "loss": 0.6771,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.484332641869686,
+      "learning_rate": 1.912792072649617e-06,
+      "loss": 0.6921,
+      "step": 5361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.473137049966319,
+      "learning_rate": 1.9127526113453153e-06,
+      "loss": 0.6836,
+      "step": 5362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.92384623372227,
+      "learning_rate": 1.912713141522213e-06,
+      "loss": 0.6686,
+      "step": 5363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4304416872233325,
+      "learning_rate": 1.912673663180678e-06,
+      "loss": 0.6992,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9237269121174538,
+      "learning_rate": 1.912634176321079e-06,
+      "loss": 0.6895,
+      "step": 5365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4156732686293485,
+      "learning_rate": 1.9125946809437843e-06,
+      "loss": 0.6914,
+      "step": 5366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.53281265184044,
+      "learning_rate": 1.912555177049163e-06,
+      "loss": 0.681,
+      "step": 5367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3254405582154636,
+      "learning_rate": 1.9125156646375837e-06,
+      "loss": 0.6784,
+      "step": 5368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4487463843697181,
+      "learning_rate": 1.912476143709415e-06,
+      "loss": 0.6771,
+      "step": 5369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.54781793833766,
+      "learning_rate": 1.9124366142650253e-06,
+      "loss": 0.6667,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7970463352233956,
+      "learning_rate": 1.912397076304784e-06,
+      "loss": 0.6803,
+      "step": 5371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.725442708099446,
+      "learning_rate": 1.9123575298290607e-06,
+      "loss": 0.694,
+      "step": 5372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.194644993104653,
+      "learning_rate": 1.9123179748382235e-06,
+      "loss": 0.7155,
+      "step": 5373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.194943323083387,
+      "learning_rate": 1.912278411332642e-06,
+      "loss": 0.681,
+      "step": 5374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5296387165706179,
+      "learning_rate": 1.9122388393126852e-06,
+      "loss": 0.668,
+      "step": 5375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.07366605031229,
+      "learning_rate": 1.912199258778723e-06,
+      "loss": 0.6947,
+      "step": 5376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.14571884465069,
+      "learning_rate": 1.9121596697311243e-06,
+      "loss": 0.7005,
+      "step": 5377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.073495102478522,
+      "learning_rate": 1.912120072170259e-06,
+      "loss": 0.7005,
+      "step": 5378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45309732440097916,
+      "learning_rate": 1.912080466096496e-06,
+      "loss": 0.6777,
+      "step": 5379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.687754273636608,
+      "learning_rate": 1.9120408515102053e-06,
+      "loss": 0.6751,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.259371757352644,
+      "learning_rate": 1.912001228411757e-06,
+      "loss": 0.6842,
+      "step": 5381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45632645494601815,
+      "learning_rate": 1.9119615968015207e-06,
+      "loss": 0.6771,
+      "step": 5382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.102308754263552,
+      "learning_rate": 1.911921956679866e-06,
+      "loss": 0.6914,
+      "step": 5383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7484208674912138,
+      "learning_rate": 1.911882308047163e-06,
+      "loss": 0.6953,
+      "step": 5384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.25626350528375,
+      "learning_rate": 1.9118426509037816e-06,
+      "loss": 0.6829,
+      "step": 5385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9833362424703695,
+      "learning_rate": 1.911802985250092e-06,
+      "loss": 0.6706,
+      "step": 5386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.604006485342938,
+      "learning_rate": 1.9117633110864648e-06,
+      "loss": 0.6901,
+      "step": 5387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7123367964736932,
+      "learning_rate": 1.91172362841327e-06,
+      "loss": 0.6849,
+      "step": 5388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4356696554466584,
+      "learning_rate": 1.911683937230878e-06,
+      "loss": 0.6738,
+      "step": 5389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.493886691473135,
+      "learning_rate": 1.911644237539659e-06,
+      "loss": 0.6719,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5655741208924117,
+      "learning_rate": 1.9116045293399836e-06,
+      "loss": 0.6979,
+      "step": 5391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.96454369782311,
+      "learning_rate": 1.9115648126322227e-06,
+      "loss": 0.6771,
+      "step": 5392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.658096152052235,
+      "learning_rate": 1.911525087416747e-06,
+      "loss": 0.6699,
+      "step": 5393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.584702528784411,
+      "learning_rate": 1.911485353693926e-06,
+      "loss": 0.6888,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3950599651343487,
+      "learning_rate": 1.911445611464133e-06,
+      "loss": 0.7038,
+      "step": 5395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.603041488486046,
+      "learning_rate": 1.911405860727737e-06,
+      "loss": 0.6497,
+      "step": 5396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.469833042841585,
+      "learning_rate": 1.911366101485109e-06,
+      "loss": 0.7064,
+      "step": 5397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.04644743112094,
+      "learning_rate": 1.9113263337366213e-06,
+      "loss": 0.6777,
+      "step": 5398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.845410376886332,
+      "learning_rate": 1.911286557482644e-06,
+      "loss": 0.6934,
+      "step": 5399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.204116562993643,
+      "learning_rate": 1.9112467727235485e-06,
+      "loss": 0.6908,
+      "step": 5400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5993468315424737,
+      "learning_rate": 1.9112069794597064e-06,
+      "loss": 0.7174,
+      "step": 5401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2862311561163153,
+      "learning_rate": 1.911167177691489e-06,
+      "loss": 0.679,
+      "step": 5402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.804012966127382,
+      "learning_rate": 1.9111273674192683e-06,
+      "loss": 0.6797,
+      "step": 5403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.978158372640676,
+      "learning_rate": 1.9110875486434142e-06,
+      "loss": 0.6908,
+      "step": 5404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7761711891364356,
+      "learning_rate": 1.9110477213643e-06,
+      "loss": 0.6628,
+      "step": 5405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.798688805756057,
+      "learning_rate": 1.911007885582297e-06,
+      "loss": 0.6921,
+      "step": 5406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.4303536648137625,
+      "learning_rate": 1.910968041297777e-06,
+      "loss": 0.6914,
+      "step": 5407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.575375114305759,
+      "learning_rate": 1.910928188511111e-06,
+      "loss": 0.6986,
+      "step": 5408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7154241064974287,
+      "learning_rate": 1.910888327222672e-06,
+      "loss": 0.6816,
+      "step": 5409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5471353877724394,
+      "learning_rate": 1.9108484574328325e-06,
+      "loss": 0.6797,
+      "step": 5410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.1427653372782505,
+      "learning_rate": 1.910808579141963e-06,
+      "loss": 0.6725,
+      "step": 5411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.587760062097268,
+      "learning_rate": 1.9107686923504365e-06,
+      "loss": 0.6693,
+      "step": 5412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.115576815671119,
+      "learning_rate": 1.910728797058625e-06,
+      "loss": 0.6732,
+      "step": 5413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3650511644413115,
+      "learning_rate": 1.910688893266902e-06,
+      "loss": 0.7018,
+      "step": 5414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.229592613513115,
+      "learning_rate": 1.910648980975638e-06,
+      "loss": 0.6855,
+      "step": 5415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4685420857483593,
+      "learning_rate": 1.910609060185207e-06,
+      "loss": 0.6803,
+      "step": 5416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.36210639265082,
+      "learning_rate": 1.910569130895981e-06,
+      "loss": 0.6868,
+      "step": 5417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0175201314164835,
+      "learning_rate": 1.9105291931083326e-06,
+      "loss": 0.6771,
+      "step": 5418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.806481632579675,
+      "learning_rate": 1.9104892468226353e-06,
+      "loss": 0.6849,
+      "step": 5419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.270693115808768,
+      "learning_rate": 1.910449292039261e-06,
+      "loss": 0.6914,
+      "step": 5420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.264300145001483,
+      "learning_rate": 1.910409328758583e-06,
+      "loss": 0.6823,
+      "step": 5421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3159868320991486,
+      "learning_rate": 1.910369356980974e-06,
+      "loss": 0.6797,
+      "step": 5422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.106489621120441,
+      "learning_rate": 1.9103293767068075e-06,
+      "loss": 0.6934,
+      "step": 5423
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.486634394969297,
+      "learning_rate": 1.9102893879364562e-06,
+      "loss": 0.6634,
+      "step": 5424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9443697107489045,
+      "learning_rate": 1.9102493906702938e-06,
+      "loss": 0.694,
+      "step": 5425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.728547745336364,
+      "learning_rate": 1.9102093849086932e-06,
+      "loss": 0.6517,
+      "step": 5426
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.505511492014726,
+      "learning_rate": 1.9101693706520278e-06,
+      "loss": 0.6758,
+      "step": 5427
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.697622274783304,
+      "learning_rate": 1.9101293479006713e-06,
+      "loss": 0.6699,
+      "step": 5428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.828783352672083,
+      "learning_rate": 1.9100893166549974e-06,
+      "loss": 0.6816,
+      "step": 5429
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3223891530755743,
+      "learning_rate": 1.910049276915379e-06,
+      "loss": 0.6647,
+      "step": 5430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6683240222460095,
+      "learning_rate": 1.9100092286821905e-06,
+      "loss": 0.6654,
+      "step": 5431
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.022447940787219,
+      "learning_rate": 1.909969171955805e-06,
+      "loss": 0.6855,
+      "step": 5432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.885855760577506,
+      "learning_rate": 1.9099291067365973e-06,
+      "loss": 0.6895,
+      "step": 5433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.140346290803164,
+      "learning_rate": 1.9098890330249405e-06,
+      "loss": 0.6895,
+      "step": 5434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4187799986356722,
+      "learning_rate": 1.909848950821209e-06,
+      "loss": 0.6569,
+      "step": 5435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.2591301474942895,
+      "learning_rate": 1.9098088601257766e-06,
+      "loss": 0.6829,
+      "step": 5436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.6451428066093845,
+      "learning_rate": 1.9097687609390174e-06,
+      "loss": 0.6732,
+      "step": 5437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5786572909423557,
+      "learning_rate": 1.9097286532613065e-06,
+      "loss": 0.6875,
+      "step": 5438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2131650221200236,
+      "learning_rate": 1.9096885370930173e-06,
+      "loss": 0.6914,
+      "step": 5439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.574565004786006,
+      "learning_rate": 1.9096484124345243e-06,
+      "loss": 0.6947,
+      "step": 5440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.195612088441409,
+      "learning_rate": 1.9096082792862027e-06,
+      "loss": 0.6777,
+      "step": 5441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.483404293811685,
+      "learning_rate": 1.9095681376484264e-06,
+      "loss": 0.6862,
+      "step": 5442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.033055000413567,
+      "learning_rate": 1.9095279875215704e-06,
+      "loss": 0.6875,
+      "step": 5443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0127968042269733,
+      "learning_rate": 1.909487828906009e-06,
+      "loss": 0.6849,
+      "step": 5444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6976555580714874,
+      "learning_rate": 1.909447661802117e-06,
+      "loss": 0.6751,
+      "step": 5445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1213807824219213,
+      "learning_rate": 1.90940748621027e-06,
+      "loss": 0.6758,
+      "step": 5446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6034238017014153,
+      "learning_rate": 1.9093673021308425e-06,
+      "loss": 0.6751,
+      "step": 5447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.75919847411345,
+      "learning_rate": 1.9093271095642096e-06,
+      "loss": 0.6712,
+      "step": 5448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9659816763235773,
+      "learning_rate": 1.909286908510746e-06,
+      "loss": 0.6764,
+      "step": 5449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5304222834784398,
+      "learning_rate": 1.9092466989708273e-06,
+      "loss": 0.6875,
+      "step": 5450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4893162266451367,
+      "learning_rate": 1.909206480944829e-06,
+      "loss": 0.6823,
+      "step": 5451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5157187869858134,
+      "learning_rate": 1.909166254433126e-06,
+      "loss": 0.6745,
+      "step": 5452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.245056060820955,
+      "learning_rate": 1.909126019436094e-06,
+      "loss": 0.6797,
+      "step": 5453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.419531017761062,
+      "learning_rate": 1.9090857759541087e-06,
+      "loss": 0.668,
+      "step": 5454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0155505750481986,
+      "learning_rate": 1.9090455239875455e-06,
+      "loss": 0.6764,
+      "step": 5455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1002494346397511,
+      "learning_rate": 1.9090052635367796e-06,
+      "loss": 0.6934,
+      "step": 5456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8368075602841207,
+      "learning_rate": 1.9089649946021876e-06,
+      "loss": 0.666,
+      "step": 5457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4973382378709329,
+      "learning_rate": 1.9089247171841448e-06,
+      "loss": 0.6803,
+      "step": 5458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5197707097464869,
+      "learning_rate": 1.908884431283027e-06,
+      "loss": 0.6823,
+      "step": 5459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.210714597457763,
+      "learning_rate": 1.9088441368992104e-06,
+      "loss": 0.6986,
+      "step": 5460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.0234567090809092,
+      "learning_rate": 1.9088038340330715e-06,
+      "loss": 0.6901,
+      "step": 5461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.530196872891378,
+      "learning_rate": 1.9087635226849857e-06,
+      "loss": 0.6953,
+      "step": 5462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.568936466158479,
+      "learning_rate": 1.9087232028553298e-06,
+      "loss": 0.7057,
+      "step": 5463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.367526944736571,
+      "learning_rate": 1.9086828745444795e-06,
+      "loss": 0.6452,
+      "step": 5464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6787337315248867,
+      "learning_rate": 1.9086425377528122e-06,
+      "loss": 0.6712,
+      "step": 5465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.083053758443006,
+      "learning_rate": 1.9086021924807036e-06,
+      "loss": 0.6706,
+      "step": 5466
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.127501774504163,
+      "learning_rate": 1.9085618387285297e-06,
+      "loss": 0.6764,
+      "step": 5467
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1875166386616898,
+      "learning_rate": 1.9085214764966688e-06,
+      "loss": 0.666,
+      "step": 5468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1779443526238365,
+      "learning_rate": 1.908481105785496e-06,
+      "loss": 0.6693,
+      "step": 5469
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.201222501312542,
+      "learning_rate": 1.9084407265953887e-06,
+      "loss": 0.6986,
+      "step": 5470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2077790508883879,
+      "learning_rate": 1.908400338926724e-06,
+      "loss": 0.6901,
+      "step": 5471
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5832507215831164,
+      "learning_rate": 1.9083599427798785e-06,
+      "loss": 0.6836,
+      "step": 5472
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3914376823415122,
+      "learning_rate": 1.9083195381552293e-06,
+      "loss": 0.6908,
+      "step": 5473
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.899767592830175,
+      "learning_rate": 1.9082791250531534e-06,
+      "loss": 0.7018,
+      "step": 5474
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.214299218729961,
+      "learning_rate": 1.908238703474028e-06,
+      "loss": 0.6875,
+      "step": 5475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.810100046663468,
+      "learning_rate": 1.9081982734182308e-06,
+      "loss": 0.6908,
+      "step": 5476
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0585482000393385,
+      "learning_rate": 1.9081578348861387e-06,
+      "loss": 0.7168,
+      "step": 5477
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5606487313478286,
+      "learning_rate": 1.9081173878781293e-06,
+      "loss": 0.6914,
+      "step": 5478
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.392562032238874,
+      "learning_rate": 1.90807693239458e-06,
+      "loss": 0.6771,
+      "step": 5479
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6865232715480841,
+      "learning_rate": 1.9080364684358683e-06,
+      "loss": 0.7025,
+      "step": 5480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6170390365593836,
+      "learning_rate": 1.9079959960023718e-06,
+      "loss": 0.6751,
+      "step": 5481
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9726649503527964,
+      "learning_rate": 1.907955515094468e-06,
+      "loss": 0.6777,
+      "step": 5482
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.2448436159621483,
+      "learning_rate": 1.907915025712536e-06,
+      "loss": 0.679,
+      "step": 5483
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.838720190070574,
+      "learning_rate": 1.9078745278569526e-06,
+      "loss": 0.6953,
+      "step": 5484
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.292577110669105,
+      "learning_rate": 1.9078340215280956e-06,
+      "loss": 0.6745,
+      "step": 5485
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9722802901473545,
+      "learning_rate": 1.907793506726343e-06,
+      "loss": 0.7038,
+      "step": 5486
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.192141468498595,
+      "learning_rate": 1.907752983452074e-06,
+      "loss": 0.7018,
+      "step": 5487
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.527450883436419,
+      "learning_rate": 1.907712451705666e-06,
+      "loss": 0.6855,
+      "step": 5488
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.048926085471254,
+      "learning_rate": 1.9076719114874974e-06,
+      "loss": 0.6823,
+      "step": 5489
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1854030253616425,
+      "learning_rate": 1.9076313627979468e-06,
+      "loss": 0.6836,
+      "step": 5490
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8601122331909363,
+      "learning_rate": 1.9075908056373916e-06,
+      "loss": 0.7051,
+      "step": 5491
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7418189526086105,
+      "learning_rate": 1.907550240006212e-06,
+      "loss": 0.6862,
+      "step": 5492
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.923063926991367,
+      "learning_rate": 1.9075096659047855e-06,
+      "loss": 0.6914,
+      "step": 5493
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9930556726369295,
+      "learning_rate": 1.907469083333491e-06,
+      "loss": 0.7064,
+      "step": 5494
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3706675597206615,
+      "learning_rate": 1.9074284922927072e-06,
+      "loss": 0.6738,
+      "step": 5495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4865047555726787,
+      "learning_rate": 1.907387892782813e-06,
+      "loss": 0.6836,
+      "step": 5496
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9638886750769846,
+      "learning_rate": 1.9073472848041874e-06,
+      "loss": 0.679,
+      "step": 5497
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5084467427604431,
+      "learning_rate": 1.9073066683572097e-06,
+      "loss": 0.6882,
+      "step": 5498
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.923361927211544,
+      "learning_rate": 1.907266043442258e-06,
+      "loss": 0.6875,
+      "step": 5499
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0933380003287518,
+      "learning_rate": 1.9072254100597124e-06,
+      "loss": 0.6693,
+      "step": 5500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.153848342356767,
+      "learning_rate": 1.907184768209952e-06,
+      "loss": 0.6868,
+      "step": 5501
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8926452488264023,
+      "learning_rate": 1.9071441178933557e-06,
+      "loss": 0.6647,
+      "step": 5502
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0496568133996464,
+      "learning_rate": 1.9071034591103032e-06,
+      "loss": 0.6836,
+      "step": 5503
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7173697014190534,
+      "learning_rate": 1.9070627918611739e-06,
+      "loss": 0.6836,
+      "step": 5504
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7879396707748789,
+      "learning_rate": 1.9070221161463473e-06,
+      "loss": 0.6634,
+      "step": 5505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.672777599451549,
+      "learning_rate": 1.9069814319662033e-06,
+      "loss": 0.6986,
+      "step": 5506
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.980769777113529,
+      "learning_rate": 1.906940739321121e-06,
+      "loss": 0.6797,
+      "step": 5507
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.1697154073411653,
+      "learning_rate": 1.9069000382114809e-06,
+      "loss": 0.6732,
+      "step": 5508
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.38316977616134,
+      "learning_rate": 1.9068593286376622e-06,
+      "loss": 0.7057,
+      "step": 5509
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1103146584092167,
+      "learning_rate": 1.9068186106000454e-06,
+      "loss": 0.6829,
+      "step": 5510
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.310400243416692,
+      "learning_rate": 1.9067778840990102e-06,
+      "loss": 0.6816,
+      "step": 5511
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.392394439902363,
+      "learning_rate": 1.906737149134937e-06,
+      "loss": 0.6855,
+      "step": 5512
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4131561576491556,
+      "learning_rate": 1.9066964057082058e-06,
+      "loss": 0.6895,
+      "step": 5513
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8291043737926767,
+      "learning_rate": 1.9066556538191968e-06,
+      "loss": 0.6751,
+      "step": 5514
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5228480341394834,
+      "learning_rate": 1.9066148934682904e-06,
+      "loss": 0.6908,
+      "step": 5515
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.0042488976515052,
+      "learning_rate": 1.906574124655867e-06,
+      "loss": 0.6745,
+      "step": 5516
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.466600628882615,
+      "learning_rate": 1.9065333473823073e-06,
+      "loss": 0.694,
+      "step": 5517
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.235110594055505,
+      "learning_rate": 1.9064925616479913e-06,
+      "loss": 0.6816,
+      "step": 5518
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5911316528705001,
+      "learning_rate": 1.9064517674533004e-06,
+      "loss": 0.6523,
+      "step": 5519
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.656923950951761,
+      "learning_rate": 1.9064109647986149e-06,
+      "loss": 0.6641,
+      "step": 5520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9541338671619948,
+      "learning_rate": 1.9063701536843157e-06,
+      "loss": 0.6862,
+      "step": 5521
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1524806527486637,
+      "learning_rate": 1.9063293341107838e-06,
+      "loss": 0.6589,
+      "step": 5522
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.075029383816859,
+      "learning_rate": 1.9062885060783998e-06,
+      "loss": 0.694,
+      "step": 5523
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4309766803121466,
+      "learning_rate": 1.9062476695875451e-06,
+      "loss": 0.6934,
+      "step": 5524
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.266341590520029,
+      "learning_rate": 1.9062068246386008e-06,
+      "loss": 0.6667,
+      "step": 5525
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.349668617910969,
+      "learning_rate": 1.9061659712319483e-06,
+      "loss": 0.6699,
+      "step": 5526
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8623251883377696,
+      "learning_rate": 1.9061251093679682e-06,
+      "loss": 0.6836,
+      "step": 5527
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9611527222895082,
+      "learning_rate": 1.9060842390470429e-06,
+      "loss": 0.6921,
+      "step": 5528
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8719238644133777,
+      "learning_rate": 1.9060433602695528e-06,
+      "loss": 0.6979,
+      "step": 5529
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.346726851977964,
+      "learning_rate": 1.9060024730358801e-06,
+      "loss": 0.6699,
+      "step": 5530
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.461802456751439,
+      "learning_rate": 1.905961577346406e-06,
+      "loss": 0.6849,
+      "step": 5531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1045062609887557,
+      "learning_rate": 1.9059206732015125e-06,
+      "loss": 0.6934,
+      "step": 5532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8535395331046938,
+      "learning_rate": 1.9058797606015813e-06,
+      "loss": 0.6973,
+      "step": 5533
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.509742372142129,
+      "learning_rate": 1.9058388395469942e-06,
+      "loss": 0.6921,
+      "step": 5534
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6754059615603349,
+      "learning_rate": 1.9057979100381329e-06,
+      "loss": 0.6738,
+      "step": 5535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2091607398730064,
+      "learning_rate": 1.9057569720753798e-06,
+      "loss": 0.6777,
+      "step": 5536
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6992908401957099,
+      "learning_rate": 1.9057160256591168e-06,
+      "loss": 0.6784,
+      "step": 5537
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.501900314119416,
+      "learning_rate": 1.905675070789726e-06,
+      "loss": 0.7122,
+      "step": 5538
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.716357896633141,
+      "learning_rate": 1.9056341074675895e-06,
+      "loss": 0.7051,
+      "step": 5539
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3457222333882157,
+      "learning_rate": 1.9055931356930898e-06,
+      "loss": 0.679,
+      "step": 5540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4058457797311832,
+      "learning_rate": 1.9055521554666095e-06,
+      "loss": 0.7083,
+      "step": 5541
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8939512833677639,
+      "learning_rate": 1.9055111667885306e-06,
+      "loss": 0.6875,
+      "step": 5542
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6144577402911784,
+      "learning_rate": 1.9054701696592361e-06,
+      "loss": 0.6927,
+      "step": 5543
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.432947596930979,
+      "learning_rate": 1.9054291640791086e-06,
+      "loss": 0.6816,
+      "step": 5544
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.516394563642583,
+      "learning_rate": 1.9053881500485303e-06,
+      "loss": 0.6836,
+      "step": 5545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3286123520151523,
+      "learning_rate": 1.9053471275678845e-06,
+      "loss": 0.7005,
+      "step": 5546
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.433038387176593,
+      "learning_rate": 1.9053060966375537e-06,
+      "loss": 0.6816,
+      "step": 5547
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3560453839975755,
+      "learning_rate": 1.9052650572579213e-06,
+      "loss": 0.6602,
+      "step": 5548
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4569986654387582,
+      "learning_rate": 1.9052240094293696e-06,
+      "loss": 0.6979,
+      "step": 5549
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.488576314997142,
+      "learning_rate": 1.9051829531522826e-06,
+      "loss": 0.6693,
+      "step": 5550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9611763885294521,
+      "learning_rate": 1.905141888427043e-06,
+      "loss": 0.6823,
+      "step": 5551
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9498372537585407,
+      "learning_rate": 1.905100815254034e-06,
+      "loss": 0.6816,
+      "step": 5552
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.008252723955906,
+      "learning_rate": 1.9050597336336392e-06,
+      "loss": 0.6829,
+      "step": 5553
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9034304034041019,
+      "learning_rate": 1.9050186435662417e-06,
+      "loss": 0.6777,
+      "step": 5554
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9399570766670484,
+      "learning_rate": 1.9049775450522252e-06,
+      "loss": 0.6771,
+      "step": 5555
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.37099021733811455,
+      "learning_rate": 1.9049364380919734e-06,
+      "loss": 0.6849,
+      "step": 5556
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.40486151236802925,
+      "learning_rate": 1.9048953226858696e-06,
+      "loss": 0.6751,
+      "step": 5557
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.239667488169959,
+      "learning_rate": 1.9048541988342976e-06,
+      "loss": 0.7005,
+      "step": 5558
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.340978068999794,
+      "learning_rate": 1.9048130665376418e-06,
+      "loss": 0.6992,
+      "step": 5559
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.332801192347401,
+      "learning_rate": 1.9047719257962854e-06,
+      "loss": 0.6901,
+      "step": 5560
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5013617200877036,
+      "learning_rate": 1.9047307766106125e-06,
+      "loss": 0.6712,
+      "step": 5561
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6822386088326282,
+      "learning_rate": 1.9046896189810073e-06,
+      "loss": 0.6895,
+      "step": 5562
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6510129512975134,
+      "learning_rate": 1.9046484529078539e-06,
+      "loss": 0.6745,
+      "step": 5563
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.430354723579045,
+      "learning_rate": 1.9046072783915366e-06,
+      "loss": 0.679,
+      "step": 5564
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.942884948437109,
+      "learning_rate": 1.9045660954324394e-06,
+      "loss": 0.7122,
+      "step": 5565
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4561347428470864,
+      "learning_rate": 1.9045249040309468e-06,
+      "loss": 0.668,
+      "step": 5566
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8533093115338986,
+      "learning_rate": 1.9044837041874432e-06,
+      "loss": 0.6803,
+      "step": 5567
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9691231415096038,
+      "learning_rate": 1.9044424959023137e-06,
+      "loss": 0.6986,
+      "step": 5568
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.244948029067535,
+      "learning_rate": 1.9044012791759419e-06,
+      "loss": 0.6882,
+      "step": 5569
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.452168806051766,
+      "learning_rate": 1.9043600540087133e-06,
+      "loss": 0.6823,
+      "step": 5570
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.48430303310187,
+      "learning_rate": 1.904318820401012e-06,
+      "loss": 0.7012,
+      "step": 5571
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.521539043906775,
+      "learning_rate": 1.9042775783532234e-06,
+      "loss": 0.6777,
+      "step": 5572
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.563345467283827,
+      "learning_rate": 1.9042363278657323e-06,
+      "loss": 0.6992,
+      "step": 5573
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5468957896889424,
+      "learning_rate": 1.9041950689389235e-06,
+      "loss": 0.6921,
+      "step": 5574
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.269298626370951,
+      "learning_rate": 1.904153801573182e-06,
+      "loss": 0.6654,
+      "step": 5575
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8760341906137397,
+      "learning_rate": 1.9041125257688932e-06,
+      "loss": 0.6992,
+      "step": 5576
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8323690228524838,
+      "learning_rate": 1.9040712415264423e-06,
+      "loss": 0.6751,
+      "step": 5577
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.403663640539477,
+      "learning_rate": 1.9040299488462144e-06,
+      "loss": 0.6868,
+      "step": 5578
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.300935966256576,
+      "learning_rate": 1.9039886477285952e-06,
+      "loss": 0.6758,
+      "step": 5579
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3230239139010567,
+      "learning_rate": 1.90394733817397e-06,
+      "loss": 0.6973,
+      "step": 5580
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.580814893298534,
+      "learning_rate": 1.9039060201827244e-06,
+      "loss": 0.7077,
+      "step": 5581
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.842883049786627,
+      "learning_rate": 1.9038646937552438e-06,
+      "loss": 0.696,
+      "step": 5582
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.191514461587872,
+      "learning_rate": 1.9038233588919142e-06,
+      "loss": 0.6914,
+      "step": 5583
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.231342903077598,
+      "learning_rate": 1.903782015593121e-06,
+      "loss": 0.694,
+      "step": 5584
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.380486649447479,
+      "learning_rate": 1.9037406638592507e-06,
+      "loss": 0.6953,
+      "step": 5585
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.150370148568734,
+      "learning_rate": 1.9036993036906884e-06,
+      "loss": 0.6823,
+      "step": 5586
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5452880380155316,
+      "learning_rate": 1.9036579350878213e-06,
+      "loss": 0.6947,
+      "step": 5587
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8786276789392509,
+      "learning_rate": 1.9036165580510342e-06,
+      "loss": 0.6882,
+      "step": 5588
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8833083630213898,
+      "learning_rate": 1.9035751725807142e-06,
+      "loss": 0.6777,
+      "step": 5589
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.206979991748056,
+      "learning_rate": 1.9035337786772469e-06,
+      "loss": 0.6823,
+      "step": 5590
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.878651980265916,
+      "learning_rate": 1.903492376341019e-06,
+      "loss": 0.6908,
+      "step": 5591
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9663714503514953,
+      "learning_rate": 1.903450965572417e-06,
+      "loss": 0.6764,
+      "step": 5592
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48835185226363453,
+      "learning_rate": 1.903409546371827e-06,
+      "loss": 0.6875,
+      "step": 5593
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.602337492842773,
+      "learning_rate": 1.9033681187396362e-06,
+      "loss": 0.6556,
+      "step": 5594
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.110398669312392,
+      "learning_rate": 1.9033266826762305e-06,
+      "loss": 0.6868,
+      "step": 5595
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.562402089130054,
+      "learning_rate": 1.903285238181997e-06,
+      "loss": 0.6725,
+      "step": 5596
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.931575463062654,
+      "learning_rate": 1.9032437852573226e-06,
+      "loss": 0.696,
+      "step": 5597
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9534272532505077,
+      "learning_rate": 1.9032023239025942e-06,
+      "loss": 0.6829,
+      "step": 5598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4356155227780331,
+      "learning_rate": 1.9031608541181985e-06,
+      "loss": 0.6699,
+      "step": 5599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7873276854621354,
+      "learning_rate": 1.9031193759045227e-06,
+      "loss": 0.694,
+      "step": 5600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8829814252582762,
+      "learning_rate": 1.9030778892619538e-06,
+      "loss": 0.6751,
+      "step": 5601
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.798677522154087,
+      "learning_rate": 1.9030363941908792e-06,
+      "loss": 0.6647,
+      "step": 5602
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5138221677469433,
+      "learning_rate": 1.9029948906916862e-06,
+      "loss": 0.6758,
+      "step": 5603
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.466451203212498,
+      "learning_rate": 1.9029533787647618e-06,
+      "loss": 0.6999,
+      "step": 5604
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.212157115313361,
+      "learning_rate": 1.9029118584104938e-06,
+      "loss": 0.6706,
+      "step": 5605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2818718778166838,
+      "learning_rate": 1.9028703296292691e-06,
+      "loss": 0.6855,
+      "step": 5606
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.667039711939353,
+      "learning_rate": 1.9028287924214762e-06,
+      "loss": 0.7057,
+      "step": 5607
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8327088386838178,
+      "learning_rate": 1.9027872467875025e-06,
+      "loss": 0.681,
+      "step": 5608
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1723759375396887,
+      "learning_rate": 1.9027456927277355e-06,
+      "loss": 0.6654,
+      "step": 5609
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9202725400245598,
+      "learning_rate": 1.9027041302425629e-06,
+      "loss": 0.6855,
+      "step": 5610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2735595024550885,
+      "learning_rate": 1.902662559332373e-06,
+      "loss": 0.6947,
+      "step": 5611
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8415535711084681,
+      "learning_rate": 1.9026209799975534e-06,
+      "loss": 0.6719,
+      "step": 5612
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.411468942860319,
+      "learning_rate": 1.9025793922384927e-06,
+      "loss": 0.6654,
+      "step": 5613
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.195490451419684,
+      "learning_rate": 1.9025377960555784e-06,
+      "loss": 0.6855,
+      "step": 5614
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7242749983763039,
+      "learning_rate": 1.9024961914491992e-06,
+      "loss": 0.6823,
+      "step": 5615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9947161117837513,
+      "learning_rate": 1.9024545784197431e-06,
+      "loss": 0.6797,
+      "step": 5616
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7771540212277601,
+      "learning_rate": 1.9024129569675987e-06,
+      "loss": 0.6888,
+      "step": 5617
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3340339921474116,
+      "learning_rate": 1.9023713270931548e-06,
+      "loss": 0.6888,
+      "step": 5618
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8666583528263951,
+      "learning_rate": 1.9023296887967987e-06,
+      "loss": 0.6868,
+      "step": 5619
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3747849869845643,
+      "learning_rate": 1.9022880420789202e-06,
+      "loss": 0.6999,
+      "step": 5620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9209564921938944,
+      "learning_rate": 1.9022463869399077e-06,
+      "loss": 0.6855,
+      "step": 5621
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3106768035126835,
+      "learning_rate": 1.9022047233801497e-06,
+      "loss": 0.6667,
+      "step": 5622
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9586487126789836,
+      "learning_rate": 1.9021630514000355e-06,
+      "loss": 0.6816,
+      "step": 5623
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.16192282235105,
+      "learning_rate": 1.9021213709999535e-06,
+      "loss": 0.681,
+      "step": 5624
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7369812568023882,
+      "learning_rate": 1.902079682180293e-06,
+      "loss": 0.6712,
+      "step": 5625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0804726017346984,
+      "learning_rate": 1.902037984941443e-06,
+      "loss": 0.6777,
+      "step": 5626
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.467095060668563,
+      "learning_rate": 1.901996279283793e-06,
+      "loss": 0.6849,
+      "step": 5627
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0884931465249004,
+      "learning_rate": 1.9019545652077317e-06,
+      "loss": 0.7057,
+      "step": 5628
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8913512300780436,
+      "learning_rate": 1.901912842713649e-06,
+      "loss": 0.6934,
+      "step": 5629
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5115748136588367,
+      "learning_rate": 1.9018711118019332e-06,
+      "loss": 0.6849,
+      "step": 5630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8678658689964105,
+      "learning_rate": 1.9018293724729753e-06,
+      "loss": 0.6738,
+      "step": 5631
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.466789935749207,
+      "learning_rate": 1.9017876247271638e-06,
+      "loss": 0.6823,
+      "step": 5632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.555571829713764,
+      "learning_rate": 1.901745868564889e-06,
+      "loss": 0.696,
+      "step": 5633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3945097179950103,
+      "learning_rate": 1.9017041039865398e-06,
+      "loss": 0.6764,
+      "step": 5634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.498458678144041,
+      "learning_rate": 1.9016623309925067e-06,
+      "loss": 0.6888,
+      "step": 5635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7875264296671552,
+      "learning_rate": 1.9016205495831793e-06,
+      "loss": 0.6797,
+      "step": 5636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4539165680269443,
+      "learning_rate": 1.9015787597589475e-06,
+      "loss": 0.6602,
+      "step": 5637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4877404641015688,
+      "learning_rate": 1.9015369615202013e-06,
+      "loss": 0.6758,
+      "step": 5638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.447523559393441,
+      "learning_rate": 1.901495154867331e-06,
+      "loss": 0.6947,
+      "step": 5639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.24603456146483,
+      "learning_rate": 1.9014533398007269e-06,
+      "loss": 0.7005,
+      "step": 5640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.9774354686953249,
+      "learning_rate": 1.901411516320779e-06,
+      "loss": 0.6829,
+      "step": 5641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8100876478671406,
+      "learning_rate": 1.9013696844278775e-06,
+      "loss": 0.681,
+      "step": 5642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.6299978435474936,
+      "learning_rate": 1.901327844122413e-06,
+      "loss": 0.6836,
+      "step": 5643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7054331390989067,
+      "learning_rate": 1.9012859954047763e-06,
+      "loss": 0.6862,
+      "step": 5644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.47364210208110985,
+      "learning_rate": 1.9012441382753576e-06,
+      "loss": 0.6745,
+      "step": 5645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9266122350726373,
+      "learning_rate": 1.901202272734548e-06,
+      "loss": 0.6914,
+      "step": 5646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.606600163849815,
+      "learning_rate": 1.9011603987827372e-06,
+      "loss": 0.6641,
+      "step": 5647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7917842462473805,
+      "learning_rate": 1.901118516420317e-06,
+      "loss": 0.679,
+      "step": 5648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.546589532309159,
+      "learning_rate": 1.9010766256476777e-06,
+      "loss": 0.6823,
+      "step": 5649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7628491356510734,
+      "learning_rate": 1.901034726465211e-06,
+      "loss": 0.6732,
+      "step": 5650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4612057823070486,
+      "learning_rate": 1.9009928188733076e-06,
+      "loss": 0.6823,
+      "step": 5651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.90974246344727,
+      "learning_rate": 1.9009509028723583e-06,
+      "loss": 0.6797,
+      "step": 5652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3321216712494932,
+      "learning_rate": 1.9009089784627545e-06,
+      "loss": 0.6562,
+      "step": 5653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.663047754354477,
+      "learning_rate": 1.9008670456448876e-06,
+      "loss": 0.6803,
+      "step": 5654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.795307518788447,
+      "learning_rate": 1.900825104419149e-06,
+      "loss": 0.6816,
+      "step": 5655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6479373717707054,
+      "learning_rate": 1.9007831547859299e-06,
+      "loss": 0.6882,
+      "step": 5656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2747659057648204,
+      "learning_rate": 1.900741196745622e-06,
+      "loss": 0.6927,
+      "step": 5657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1489486306586663,
+      "learning_rate": 1.9006992302986168e-06,
+      "loss": 0.6888,
+      "step": 5658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0631560679075096,
+      "learning_rate": 1.9006572554453059e-06,
+      "loss": 0.6934,
+      "step": 5659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.292291383592662,
+      "learning_rate": 1.9006152721860815e-06,
+      "loss": 0.668,
+      "step": 5660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.886055996123708,
+      "learning_rate": 1.900573280521335e-06,
+      "loss": 0.7174,
+      "step": 5661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9485468562059807,
+      "learning_rate": 1.9005312804514584e-06,
+      "loss": 0.6862,
+      "step": 5662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.633852948573134,
+      "learning_rate": 1.9004892719768437e-06,
+      "loss": 0.6901,
+      "step": 5663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4864091909031532,
+      "learning_rate": 1.900447255097883e-06,
+      "loss": 0.6921,
+      "step": 5664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3226959674003864,
+      "learning_rate": 1.9004052298149684e-06,
+      "loss": 0.6745,
+      "step": 5665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.609281726305633,
+      "learning_rate": 1.9003631961284925e-06,
+      "loss": 0.6771,
+      "step": 5666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2314401198019638,
+      "learning_rate": 1.9003211540388467e-06,
+      "loss": 0.6849,
+      "step": 5667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.268316054100911,
+      "learning_rate": 1.9002791035464245e-06,
+      "loss": 0.679,
+      "step": 5668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.818254798224212,
+      "learning_rate": 1.9002370446516175e-06,
+      "loss": 0.6797,
+      "step": 5669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1032159842152875,
+      "learning_rate": 1.9001949773548184e-06,
+      "loss": 0.6979,
+      "step": 5670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7182901217918238,
+      "learning_rate": 1.9001529016564204e-06,
+      "loss": 0.6836,
+      "step": 5671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3622848907564897,
+      "learning_rate": 1.9001108175568153e-06,
+      "loss": 0.6927,
+      "step": 5672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.0556311899896045,
+      "learning_rate": 1.9000687250563967e-06,
+      "loss": 0.6921,
+      "step": 5673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8013458967099603,
+      "learning_rate": 1.900026624155557e-06,
+      "loss": 0.6842,
+      "step": 5674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1382900199681067,
+      "learning_rate": 1.8999845148546892e-06,
+      "loss": 0.6797,
+      "step": 5675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2795776092559787,
+      "learning_rate": 1.8999423971541862e-06,
+      "loss": 0.6712,
+      "step": 5676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1939675239421175,
+      "learning_rate": 1.8999002710544412e-06,
+      "loss": 0.6719,
+      "step": 5677
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.16970920615337,
+      "learning_rate": 1.8998581365558478e-06,
+      "loss": 0.6888,
+      "step": 5678
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3312513650899354,
+      "learning_rate": 1.8998159936587986e-06,
+      "loss": 0.6921,
+      "step": 5679
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4192313603325126,
+      "learning_rate": 1.899773842363687e-06,
+      "loss": 0.6953,
+      "step": 5680
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7602624522432158,
+      "learning_rate": 1.8997316826709064e-06,
+      "loss": 0.6667,
+      "step": 5681
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.162450537788318,
+      "learning_rate": 1.8996895145808506e-06,
+      "loss": 0.6823,
+      "step": 5682
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.501587496565855,
+      "learning_rate": 1.8996473380939133e-06,
+      "loss": 0.6836,
+      "step": 5683
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6496453754647847,
+      "learning_rate": 1.8996051532104878e-06,
+      "loss": 0.6771,
+      "step": 5684
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0876229669716295,
+      "learning_rate": 1.8995629599309675e-06,
+      "loss": 0.6751,
+      "step": 5685
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6179285748744595,
+      "learning_rate": 1.8995207582557469e-06,
+      "loss": 0.6855,
+      "step": 5686
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.045979277780798,
+      "learning_rate": 1.899478548185219e-06,
+      "loss": 0.707,
+      "step": 5687
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1419130575104504,
+      "learning_rate": 1.8994363297197789e-06,
+      "loss": 0.6829,
+      "step": 5688
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7241946836669815,
+      "learning_rate": 1.8993941028598197e-06,
+      "loss": 0.6895,
+      "step": 5689
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.165313805286444,
+      "learning_rate": 1.8993518676057357e-06,
+      "loss": 0.6712,
+      "step": 5690
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.671142936661056,
+      "learning_rate": 1.8993096239579215e-06,
+      "loss": 0.6966,
+      "step": 5691
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6938901904841335,
+      "learning_rate": 1.8992673719167707e-06,
+      "loss": 0.6784,
+      "step": 5692
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.976364326847813,
+      "learning_rate": 1.899225111482678e-06,
+      "loss": 0.6914,
+      "step": 5693
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5372933123006596,
+      "learning_rate": 1.8991828426560378e-06,
+      "loss": 0.6973,
+      "step": 5694
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.520438450410901,
+      "learning_rate": 1.899140565437245e-06,
+      "loss": 0.6803,
+      "step": 5695
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.155176673604168,
+      "learning_rate": 1.8990982798266933e-06,
+      "loss": 0.6621,
+      "step": 5696
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.275211006486067,
+      "learning_rate": 1.899055985824778e-06,
+      "loss": 0.6784,
+      "step": 5697
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.697621767700265,
+      "learning_rate": 1.8990136834318938e-06,
+      "loss": 0.6784,
+      "step": 5698
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.619970370030874,
+      "learning_rate": 1.8989713726484356e-06,
+      "loss": 0.6803,
+      "step": 5699
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7955639190223,
+      "learning_rate": 1.8989290534747977e-06,
+      "loss": 0.6764,
+      "step": 5700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4822126077620442,
+      "learning_rate": 1.8988867259113754e-06,
+      "loss": 0.6895,
+      "step": 5701
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.864200902362447,
+      "learning_rate": 1.898844389958564e-06,
+      "loss": 0.6947,
+      "step": 5702
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.914384355045049,
+      "learning_rate": 1.8988020456167584e-06,
+      "loss": 0.6927,
+      "step": 5703
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3245644076642944,
+      "learning_rate": 1.8987596928863535e-06,
+      "loss": 0.6699,
+      "step": 5704
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8173784235488852,
+      "learning_rate": 1.8987173317677454e-06,
+      "loss": 0.6829,
+      "step": 5705
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.014796859110358,
+      "learning_rate": 1.8986749622613287e-06,
+      "loss": 0.6667,
+      "step": 5706
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6230632678191484,
+      "learning_rate": 1.8986325843674994e-06,
+      "loss": 0.6823,
+      "step": 5707
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.741686756394081,
+      "learning_rate": 1.8985901980866523e-06,
+      "loss": 0.6862,
+      "step": 5708
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8348524352382031,
+      "learning_rate": 1.8985478034191837e-06,
+      "loss": 0.6654,
+      "step": 5709
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0213473651315232,
+      "learning_rate": 1.8985054003654888e-06,
+      "loss": 0.6947,
+      "step": 5710
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.559204799202996,
+      "learning_rate": 1.8984629889259636e-06,
+      "loss": 0.6725,
+      "step": 5711
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5003665957501666,
+      "learning_rate": 1.8984205691010039e-06,
+      "loss": 0.6784,
+      "step": 5712
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8532288938012103,
+      "learning_rate": 1.8983781408910056e-06,
+      "loss": 0.6895,
+      "step": 5713
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8333742487654896,
+      "learning_rate": 1.8983357042963646e-06,
+      "loss": 0.6849,
+      "step": 5714
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5439757805085557,
+      "learning_rate": 1.898293259317477e-06,
+      "loss": 0.7064,
+      "step": 5715
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.8210389609459146,
+      "learning_rate": 1.898250805954739e-06,
+      "loss": 0.6868,
+      "step": 5716
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.9681914765678747,
+      "learning_rate": 1.8982083442085468e-06,
+      "loss": 0.7031,
+      "step": 5717
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.842885709181293,
+      "learning_rate": 1.8981658740792967e-06,
+      "loss": 0.6947,
+      "step": 5718
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9563318077930694,
+      "learning_rate": 1.898123395567385e-06,
+      "loss": 0.6654,
+      "step": 5719
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4257672989071417,
+      "learning_rate": 1.8980809086732083e-06,
+      "loss": 0.6947,
+      "step": 5720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.6611321303945985,
+      "learning_rate": 1.8980384133971631e-06,
+      "loss": 0.6784,
+      "step": 5721
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.926105846832941,
+      "learning_rate": 1.897995909739646e-06,
+      "loss": 0.694,
+      "step": 5722
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.602894842187153,
+      "learning_rate": 1.8979533977010535e-06,
+      "loss": 0.6797,
+      "step": 5723
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49426812100402306,
+      "learning_rate": 1.8979108772817827e-06,
+      "loss": 0.6882,
+      "step": 5724
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.1738719648178995,
+      "learning_rate": 1.89786834848223e-06,
+      "loss": 0.6947,
+      "step": 5725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6008662314574131,
+      "learning_rate": 1.8978258113027929e-06,
+      "loss": 0.6719,
+      "step": 5726
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.508465716852739,
+      "learning_rate": 1.8977832657438682e-06,
+      "loss": 0.6641,
+      "step": 5727
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.1801666278230776,
+      "learning_rate": 1.8977407118058526e-06,
+      "loss": 0.6966,
+      "step": 5728
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.327992056767213,
+      "learning_rate": 1.8976981494891433e-06,
+      "loss": 0.6862,
+      "step": 5729
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9599689774164932,
+      "learning_rate": 1.8976555787941381e-06,
+      "loss": 0.6712,
+      "step": 5730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.15644934040858,
+      "learning_rate": 1.8976129997212342e-06,
+      "loss": 0.6725,
+      "step": 5731
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1989128849456323,
+      "learning_rate": 1.8975704122708285e-06,
+      "loss": 0.679,
+      "step": 5732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3938393842617263,
+      "learning_rate": 1.8975278164433188e-06,
+      "loss": 0.6667,
+      "step": 5733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6724216429388632,
+      "learning_rate": 1.8974852122391028e-06,
+      "loss": 0.6673,
+      "step": 5734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3791972402556618,
+      "learning_rate": 1.8974425996585777e-06,
+      "loss": 0.6777,
+      "step": 5735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.502808565867802,
+      "learning_rate": 1.8973999787021416e-06,
+      "loss": 0.6842,
+      "step": 5736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8394726413444533,
+      "learning_rate": 1.8973573493701922e-06,
+      "loss": 0.6849,
+      "step": 5737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9943399301836768,
+      "learning_rate": 1.8973147116631272e-06,
+      "loss": 0.6888,
+      "step": 5738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.946187425246753,
+      "learning_rate": 1.8972720655813446e-06,
+      "loss": 0.7077,
+      "step": 5739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3827342521395303,
+      "learning_rate": 1.8972294111252424e-06,
+      "loss": 0.679,
+      "step": 5740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.704261143629431,
+      "learning_rate": 1.8971867482952192e-06,
+      "loss": 0.6621,
+      "step": 5741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6356937418647707,
+      "learning_rate": 1.897144077091672e-06,
+      "loss": 0.7005,
+      "step": 5742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8999155694635049,
+      "learning_rate": 1.8971013975150004e-06,
+      "loss": 0.6855,
+      "step": 5743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5128330486874426,
+      "learning_rate": 1.8970587095656019e-06,
+      "loss": 0.6803,
+      "step": 5744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.133493931812541,
+      "learning_rate": 1.8970160132438752e-06,
+      "loss": 0.6966,
+      "step": 5745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.666324787951686,
+      "learning_rate": 1.8969733085502186e-06,
+      "loss": 0.707,
+      "step": 5746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8917005388006758,
+      "learning_rate": 1.8969305954850307e-06,
+      "loss": 0.7031,
+      "step": 5747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.5573490739220945,
+      "learning_rate": 1.8968878740487104e-06,
+      "loss": 0.6797,
+      "step": 5748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.255828571839916,
+      "learning_rate": 1.8968451442416562e-06,
+      "loss": 0.6797,
+      "step": 5749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5190873977692851,
+      "learning_rate": 1.8968024060642672e-06,
+      "loss": 0.6868,
+      "step": 5750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3823135655857297,
+      "learning_rate": 1.8967596595169418e-06,
+      "loss": 0.6771,
+      "step": 5751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.332256324552277,
+      "learning_rate": 1.8967169046000793e-06,
+      "loss": 0.6947,
+      "step": 5752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.124390502297776,
+      "learning_rate": 1.8966741413140784e-06,
+      "loss": 0.6986,
+      "step": 5753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.568240637280728,
+      "learning_rate": 1.8966313696593387e-06,
+      "loss": 0.6901,
+      "step": 5754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.901360787174524,
+      "learning_rate": 1.8965885896362587e-06,
+      "loss": 0.6895,
+      "step": 5755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.443272921963464,
+      "learning_rate": 1.8965458012452387e-06,
+      "loss": 0.6751,
+      "step": 5756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.788279089285079,
+      "learning_rate": 1.8965030044866772e-06,
+      "loss": 0.6888,
+      "step": 5757
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1022001648949211,
+      "learning_rate": 1.8964601993609736e-06,
+      "loss": 0.6751,
+      "step": 5758
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.494154346413941,
+      "learning_rate": 1.8964173858685281e-06,
+      "loss": 0.6771,
+      "step": 5759
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4342255340163689,
+      "learning_rate": 1.8963745640097396e-06,
+      "loss": 0.6797,
+      "step": 5760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.8755156056215,
+      "learning_rate": 1.8963317337850083e-06,
+      "loss": 0.6634,
+      "step": 5761
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.031523884533938,
+      "learning_rate": 1.8962888951947333e-06,
+      "loss": 0.6758,
+      "step": 5762
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.44709465204532595,
+      "learning_rate": 1.896246048239315e-06,
+      "loss": 0.6719,
+      "step": 5763
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.702700835291541,
+      "learning_rate": 1.896203192919153e-06,
+      "loss": 0.6556,
+      "step": 5764
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2049375198457573,
+      "learning_rate": 1.8961603292346475e-06,
+      "loss": 0.6888,
+      "step": 5765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1166873011126075,
+      "learning_rate": 1.8961174571861982e-06,
+      "loss": 0.6764,
+      "step": 5766
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5150526580454737,
+      "learning_rate": 1.8960745767742055e-06,
+      "loss": 0.6882,
+      "step": 5767
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2996652037854877,
+      "learning_rate": 1.8960316879990694e-06,
+      "loss": 0.6803,
+      "step": 5768
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.310140596784811,
+      "learning_rate": 1.8959887908611906e-06,
+      "loss": 0.6966,
+      "step": 5769
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.069711691065855,
+      "learning_rate": 1.895945885360969e-06,
+      "loss": 0.7025,
+      "step": 5770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7457525986711184,
+      "learning_rate": 1.8959029714988052e-06,
+      "loss": 0.6947,
+      "step": 5771
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.371233379472939,
+      "learning_rate": 1.8958600492750997e-06,
+      "loss": 0.6771,
+      "step": 5772
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4163868494304255,
+      "learning_rate": 1.8958171186902533e-06,
+      "loss": 0.6908,
+      "step": 5773
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7168400184118793,
+      "learning_rate": 1.8957741797446664e-06,
+      "loss": 0.6784,
+      "step": 5774
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.648158237595941,
+      "learning_rate": 1.8957312324387397e-06,
+      "loss": 0.6816,
+      "step": 5775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5789768314206953,
+      "learning_rate": 1.8956882767728745e-06,
+      "loss": 0.6764,
+      "step": 5776
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7682562086457699,
+      "learning_rate": 1.8956453127474714e-06,
+      "loss": 0.6875,
+      "step": 5777
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8211506118783203,
+      "learning_rate": 1.895602340362931e-06,
+      "loss": 0.6823,
+      "step": 5778
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2107388912122157,
+      "learning_rate": 1.8955593596196554e-06,
+      "loss": 0.6888,
+      "step": 5779
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9290956476111283,
+      "learning_rate": 1.8955163705180443e-06,
+      "loss": 0.6764,
+      "step": 5780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5513879908553017,
+      "learning_rate": 1.8954733730585005e-06,
+      "loss": 0.6842,
+      "step": 5781
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.805813814807205,
+      "learning_rate": 1.8954303672414241e-06,
+      "loss": 0.6888,
+      "step": 5782
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5199415785741963,
+      "learning_rate": 1.8953873530672166e-06,
+      "loss": 0.6947,
+      "step": 5783
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8217027052960235,
+      "learning_rate": 1.8953443305362802e-06,
+      "loss": 0.7018,
+      "step": 5784
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7638888891201998,
+      "learning_rate": 1.8953012996490162e-06,
+      "loss": 0.6764,
+      "step": 5785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.106118886001696,
+      "learning_rate": 1.8952582604058254e-06,
+      "loss": 0.6862,
+      "step": 5786
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.739461829905946,
+      "learning_rate": 1.8952152128071104e-06,
+      "loss": 0.6758,
+      "step": 5787
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.9247422163003521,
+      "learning_rate": 1.8951721568532725e-06,
+      "loss": 0.6816,
+      "step": 5788
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3759019666367902,
+      "learning_rate": 1.895129092544714e-06,
+      "loss": 0.6797,
+      "step": 5789
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7740902232028155,
+      "learning_rate": 1.8950860198818361e-06,
+      "loss": 0.6634,
+      "step": 5790
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.214971342747412,
+      "learning_rate": 1.8950429388650415e-06,
+      "loss": 0.6966,
+      "step": 5791
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1881031022766677,
+      "learning_rate": 1.8949998494947319e-06,
+      "loss": 0.679,
+      "step": 5792
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.633135843959463,
+      "learning_rate": 1.8949567517713097e-06,
+      "loss": 0.6823,
+      "step": 5793
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.853101150842682,
+      "learning_rate": 1.8949136456951766e-06,
+      "loss": 0.7044,
+      "step": 5794
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7656448721831561,
+      "learning_rate": 1.894870531266736e-06,
+      "loss": 0.6797,
+      "step": 5795
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.011784662986079,
+      "learning_rate": 1.894827408486389e-06,
+      "loss": 0.6589,
+      "step": 5796
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7702292713759276,
+      "learning_rate": 1.8947842773545389e-06,
+      "loss": 0.722,
+      "step": 5797
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0258817792746755,
+      "learning_rate": 1.8947411378715879e-06,
+      "loss": 0.7031,
+      "step": 5798
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.751800823656043,
+      "learning_rate": 1.8946979900379388e-06,
+      "loss": 0.6862,
+      "step": 5799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.357393362439956,
+      "learning_rate": 1.894654833853994e-06,
+      "loss": 0.6738,
+      "step": 5800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7818072722063145,
+      "learning_rate": 1.8946116693201568e-06,
+      "loss": 0.6758,
+      "step": 5801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.800922021928231,
+      "learning_rate": 1.8945684964368298e-06,
+      "loss": 0.6966,
+      "step": 5802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.531842322887951,
+      "learning_rate": 1.8945253152044156e-06,
+      "loss": 0.6927,
+      "step": 5803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.454003294842343,
+      "learning_rate": 1.8944821256233179e-06,
+      "loss": 0.6764,
+      "step": 5804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8202808576101277,
+      "learning_rate": 1.8944389276939395e-06,
+      "loss": 0.6966,
+      "step": 5805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.606672496239756,
+      "learning_rate": 1.894395721416683e-06,
+      "loss": 0.6914,
+      "step": 5806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.236503181948538,
+      "learning_rate": 1.8943525067919524e-06,
+      "loss": 0.6895,
+      "step": 5807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9955553539231059,
+      "learning_rate": 1.8943092838201506e-06,
+      "loss": 0.6667,
+      "step": 5808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1023162167295946,
+      "learning_rate": 1.8942660525016814e-06,
+      "loss": 0.6712,
+      "step": 5809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6686722433276703,
+      "learning_rate": 1.894222812836948e-06,
+      "loss": 0.6816,
+      "step": 5810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.399511712353142,
+      "learning_rate": 1.894179564826354e-06,
+      "loss": 0.6979,
+      "step": 5811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.6391403586305575,
+      "learning_rate": 1.894136308470303e-06,
+      "loss": 0.668,
+      "step": 5812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.833517886114246,
+      "learning_rate": 1.8940930437691987e-06,
+      "loss": 0.694,
+      "step": 5813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.077642087509309,
+      "learning_rate": 1.894049770723445e-06,
+      "loss": 0.6855,
+      "step": 5814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3675848617130593,
+      "learning_rate": 1.8940064893334457e-06,
+      "loss": 0.6953,
+      "step": 5815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.208380805314563,
+      "learning_rate": 1.893963199599605e-06,
+      "loss": 0.6797,
+      "step": 5816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.690730937779325,
+      "learning_rate": 1.8939199015223263e-06,
+      "loss": 0.6738,
+      "step": 5817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.934613076297435,
+      "learning_rate": 1.8938765951020144e-06,
+      "loss": 0.6751,
+      "step": 5818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1703147519309798,
+      "learning_rate": 1.893833280339073e-06,
+      "loss": 0.6862,
+      "step": 5819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5219298205235605,
+      "learning_rate": 1.8937899572339067e-06,
+      "loss": 0.6868,
+      "step": 5820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.059285012352195,
+      "learning_rate": 1.8937466257869197e-06,
+      "loss": 0.6725,
+      "step": 5821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5137517708945837,
+      "learning_rate": 1.8937032859985162e-06,
+      "loss": 0.6842,
+      "step": 5822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9922617329216517,
+      "learning_rate": 1.8936599378691013e-06,
+      "loss": 0.6758,
+      "step": 5823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4831263613722503,
+      "learning_rate": 1.8936165813990788e-06,
+      "loss": 0.6868,
+      "step": 5824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.612737356264833,
+      "learning_rate": 1.893573216588854e-06,
+      "loss": 0.6693,
+      "step": 5825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5551833567318094,
+      "learning_rate": 1.8935298434388311e-06,
+      "loss": 0.6686,
+      "step": 5826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8010150289318558,
+      "learning_rate": 1.8934864619494154e-06,
+      "loss": 0.6628,
+      "step": 5827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3892560768271918,
+      "learning_rate": 1.8934430721210113e-06,
+      "loss": 0.6582,
+      "step": 5828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.256292750928203,
+      "learning_rate": 1.893399673954024e-06,
+      "loss": 0.6725,
+      "step": 5829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.008553588340591,
+      "learning_rate": 1.8933562674488587e-06,
+      "loss": 0.709,
+      "step": 5830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3406687320482205,
+      "learning_rate": 1.8933128526059201e-06,
+      "loss": 0.6908,
+      "step": 5831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.676525064600641,
+      "learning_rate": 1.8932694294256138e-06,
+      "loss": 0.6849,
+      "step": 5832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.775650452545772,
+      "learning_rate": 1.893225997908345e-06,
+      "loss": 0.6764,
+      "step": 5833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7644023279080867,
+      "learning_rate": 1.8931825580545188e-06,
+      "loss": 0.6667,
+      "step": 5834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.526754605958477,
+      "learning_rate": 1.893139109864541e-06,
+      "loss": 0.666,
+      "step": 5835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.819768382311779,
+      "learning_rate": 1.8930956533388165e-06,
+      "loss": 0.6732,
+      "step": 5836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5601533813945314,
+      "learning_rate": 1.8930521884777515e-06,
+      "loss": 0.6732,
+      "step": 5837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.5258274123014814,
+      "learning_rate": 1.8930087152817516e-06,
+      "loss": 0.6784,
+      "step": 5838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.330371758496266,
+      "learning_rate": 1.8929652337512222e-06,
+      "loss": 0.6803,
+      "step": 5839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4728203896536765,
+      "learning_rate": 1.8929217438865694e-06,
+      "loss": 0.6953,
+      "step": 5840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.432364209702959,
+      "learning_rate": 1.8928782456881991e-06,
+      "loss": 0.6875,
+      "step": 5841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5534916818862385,
+      "learning_rate": 1.892834739156517e-06,
+      "loss": 0.6706,
+      "step": 5842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.608751771932098,
+      "learning_rate": 1.8927912242919295e-06,
+      "loss": 0.709,
+      "step": 5843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.933560807432667,
+      "learning_rate": 1.8927477010948425e-06,
+      "loss": 0.6823,
+      "step": 5844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.781365263533117,
+      "learning_rate": 1.892704169565662e-06,
+      "loss": 0.6836,
+      "step": 5845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3718834895666472,
+      "learning_rate": 1.8926606297047948e-06,
+      "loss": 0.6628,
+      "step": 5846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6804790363111145,
+      "learning_rate": 1.892617081512647e-06,
+      "loss": 0.6862,
+      "step": 5847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.324371175776372,
+      "learning_rate": 1.892573524989625e-06,
+      "loss": 0.6914,
+      "step": 5848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.622174614262544,
+      "learning_rate": 1.8925299601361355e-06,
+      "loss": 0.6921,
+      "step": 5849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.589615376608115,
+      "learning_rate": 1.8924863869525845e-06,
+      "loss": 0.679,
+      "step": 5850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2902023712430863,
+      "learning_rate": 1.8924428054393795e-06,
+      "loss": 0.6758,
+      "step": 5851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1070032329039132,
+      "learning_rate": 1.8923992155969269e-06,
+      "loss": 0.6699,
+      "step": 5852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4635753781230223,
+      "learning_rate": 1.8923556174256333e-06,
+      "loss": 0.6875,
+      "step": 5853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9253173454152992,
+      "learning_rate": 1.892312010925906e-06,
+      "loss": 0.6699,
+      "step": 5854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.111958646948024,
+      "learning_rate": 1.8922683960981516e-06,
+      "loss": 0.6654,
+      "step": 5855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.192286610640333,
+      "learning_rate": 1.8922247729427775e-06,
+      "loss": 0.668,
+      "step": 5856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6941452017418777,
+      "learning_rate": 1.8921811414601904e-06,
+      "loss": 0.7018,
+      "step": 5857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.618936338705843,
+      "learning_rate": 1.8921375016507981e-06,
+      "loss": 0.6934,
+      "step": 5858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2670768683580413,
+      "learning_rate": 1.8920938535150076e-06,
+      "loss": 0.6823,
+      "step": 5859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.805105141163526,
+      "learning_rate": 1.8920501970532262e-06,
+      "loss": 0.7051,
+      "step": 5860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2069087195617576,
+      "learning_rate": 1.8920065322658614e-06,
+      "loss": 0.6895,
+      "step": 5861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.534131838856608,
+      "learning_rate": 1.8919628591533207e-06,
+      "loss": 0.6654,
+      "step": 5862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6444501275079458,
+      "learning_rate": 1.8919191777160118e-06,
+      "loss": 0.6693,
+      "step": 5863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1831584762964416,
+      "learning_rate": 1.8918754879543425e-06,
+      "loss": 0.6738,
+      "step": 5864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.600372815109467,
+      "learning_rate": 1.8918317898687202e-06,
+      "loss": 0.7005,
+      "step": 5865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9924697311166937,
+      "learning_rate": 1.891788083459553e-06,
+      "loss": 0.6758,
+      "step": 5866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5143550842509055,
+      "learning_rate": 1.8917443687272488e-06,
+      "loss": 0.6764,
+      "step": 5867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.907324040742769,
+      "learning_rate": 1.8917006456722152e-06,
+      "loss": 0.7031,
+      "step": 5868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0125717732973536,
+      "learning_rate": 1.891656914294861e-06,
+      "loss": 0.6927,
+      "step": 5869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.734529918561692,
+      "learning_rate": 1.891613174595594e-06,
+      "loss": 0.6602,
+      "step": 5870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9217786698436028,
+      "learning_rate": 1.891569426574822e-06,
+      "loss": 0.6712,
+      "step": 5871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9558600033936074,
+      "learning_rate": 1.8915256702329541e-06,
+      "loss": 0.6842,
+      "step": 5872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.318633759390756,
+      "learning_rate": 1.8914819055703983e-06,
+      "loss": 0.7155,
+      "step": 5873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.0878083081310888,
+      "learning_rate": 1.8914381325875629e-06,
+      "loss": 0.6934,
+      "step": 5874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4704283982135102,
+      "learning_rate": 1.8913943512848563e-06,
+      "loss": 0.7044,
+      "step": 5875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6186704503992715,
+      "learning_rate": 1.8913505616626878e-06,
+      "loss": 0.6927,
+      "step": 5876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.485747614842573,
+      "learning_rate": 1.8913067637214655e-06,
+      "loss": 0.6725,
+      "step": 5877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7166016379946732,
+      "learning_rate": 1.8912629574615984e-06,
+      "loss": 0.6667,
+      "step": 5878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.794814159080404,
+      "learning_rate": 1.8912191428834951e-06,
+      "loss": 0.7129,
+      "step": 5879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2210114579946,
+      "learning_rate": 1.8911753199875648e-06,
+      "loss": 0.6862,
+      "step": 5880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.1127979818657205,
+      "learning_rate": 1.8911314887742168e-06,
+      "loss": 0.6706,
+      "step": 5881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.850934110301945,
+      "learning_rate": 1.8910876492438593e-06,
+      "loss": 0.707,
+      "step": 5882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.795126670901922,
+      "learning_rate": 1.8910438013969024e-06,
+      "loss": 0.6719,
+      "step": 5883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0936531079642187,
+      "learning_rate": 1.8909999452337545e-06,
+      "loss": 0.6719,
+      "step": 5884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2772097598776853,
+      "learning_rate": 1.8909560807548255e-06,
+      "loss": 0.6927,
+      "step": 5885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.761772082556573,
+      "learning_rate": 1.8909122079605247e-06,
+      "loss": 0.6999,
+      "step": 5886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.211275611370532,
+      "learning_rate": 1.8908683268512614e-06,
+      "loss": 0.6829,
+      "step": 5887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4619287689467533,
+      "learning_rate": 1.8908244374274452e-06,
+      "loss": 0.6849,
+      "step": 5888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6767189936812046,
+      "learning_rate": 1.8907805396894859e-06,
+      "loss": 0.6868,
+      "step": 5889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43577764960217996,
+      "learning_rate": 1.8907366336377927e-06,
+      "loss": 0.6868,
+      "step": 5890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6090718795628086,
+      "learning_rate": 1.8906927192727759e-06,
+      "loss": 0.681,
+      "step": 5891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.825757279787833,
+      "learning_rate": 1.8906487965948451e-06,
+      "loss": 0.681,
+      "step": 5892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.641065283158043,
+      "learning_rate": 1.8906048656044104e-06,
+      "loss": 0.6712,
+      "step": 5893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5885401642951331,
+      "learning_rate": 1.890560926301882e-06,
+      "loss": 0.6895,
+      "step": 5894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.413559631683134,
+      "learning_rate": 1.8905169786876693e-06,
+      "loss": 0.6953,
+      "step": 5895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.860726293033074,
+      "learning_rate": 1.8904730227621832e-06,
+      "loss": 0.6732,
+      "step": 5896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5648291127917496,
+      "learning_rate": 1.8904290585258335e-06,
+      "loss": 0.6797,
+      "step": 5897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.577450353460008,
+      "learning_rate": 1.8903850859790307e-06,
+      "loss": 0.6784,
+      "step": 5898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.840325015190047,
+      "learning_rate": 1.890341105122185e-06,
+      "loss": 0.6895,
+      "step": 5899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9034068546529915,
+      "learning_rate": 1.890297115955707e-06,
+      "loss": 0.6608,
+      "step": 5900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7630881266200054,
+      "learning_rate": 1.8902531184800075e-06,
+      "loss": 0.6921,
+      "step": 5901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3625475664607728,
+      "learning_rate": 1.8902091126954968e-06,
+      "loss": 0.6849,
+      "step": 5902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4314372818807035,
+      "learning_rate": 1.8901650986025857e-06,
+      "loss": 0.694,
+      "step": 5903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5332524884365424,
+      "learning_rate": 1.890121076201685e-06,
+      "loss": 0.6686,
+      "step": 5904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5649505613036236,
+      "learning_rate": 1.8900770454932059e-06,
+      "loss": 0.6914,
+      "step": 5905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6316613864772797,
+      "learning_rate": 1.8900330064775584e-06,
+      "loss": 0.6816,
+      "step": 5906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.887035442894278,
+      "learning_rate": 1.8899889591551546e-06,
+      "loss": 0.6908,
+      "step": 5907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8189305509948062,
+      "learning_rate": 1.889944903526405e-06,
+      "loss": 0.6784,
+      "step": 5908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9793176759356355,
+      "learning_rate": 1.8899008395917208e-06,
+      "loss": 0.6732,
+      "step": 5909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.3726819635631,
+      "learning_rate": 1.8898567673515137e-06,
+      "loss": 0.6621,
+      "step": 5910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.098438036010641,
+      "learning_rate": 1.8898126868061945e-06,
+      "loss": 0.668,
+      "step": 5911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9382499594386378,
+      "learning_rate": 1.8897685979561748e-06,
+      "loss": 0.7077,
+      "step": 5912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.865377381800785,
+      "learning_rate": 1.889724500801866e-06,
+      "loss": 0.7077,
+      "step": 5913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.138046610383268,
+      "learning_rate": 1.8896803953436798e-06,
+      "loss": 0.6771,
+      "step": 5914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4453037558732833,
+      "learning_rate": 1.889636281582028e-06,
+      "loss": 0.7103,
+      "step": 5915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.051366075502716,
+      "learning_rate": 1.889592159517322e-06,
+      "loss": 0.7025,
+      "step": 5916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.216645825007843,
+      "learning_rate": 1.8895480291499735e-06,
+      "loss": 0.6745,
+      "step": 5917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5540214809603665,
+      "learning_rate": 1.8895038904803945e-06,
+      "loss": 0.6862,
+      "step": 5918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.27693836321633,
+      "learning_rate": 1.8894597435089974e-06,
+      "loss": 0.6966,
+      "step": 5919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6258670775758,
+      "learning_rate": 1.8894155882361937e-06,
+      "loss": 0.6888,
+      "step": 5920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.075423097732612,
+      "learning_rate": 1.8893714246623958e-06,
+      "loss": 0.6634,
+      "step": 5921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8817476753651244,
+      "learning_rate": 1.8893272527880155e-06,
+      "loss": 0.6934,
+      "step": 5922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.89645213567567,
+      "learning_rate": 1.8892830726134654e-06,
+      "loss": 0.6634,
+      "step": 5923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7417658063265917,
+      "learning_rate": 1.8892388841391578e-06,
+      "loss": 0.7174,
+      "step": 5924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.594037589980852,
+      "learning_rate": 1.889194687365505e-06,
+      "loss": 0.7135,
+      "step": 5925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2523053612761523,
+      "learning_rate": 1.8891504822929197e-06,
+      "loss": 0.6751,
+      "step": 5926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4811828970389134,
+      "learning_rate": 1.889106268921814e-06,
+      "loss": 0.6829,
+      "step": 5927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.041427434761512,
+      "learning_rate": 1.8890620472526008e-06,
+      "loss": 0.6823,
+      "step": 5928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.91807170467695,
+      "learning_rate": 1.8890178172856932e-06,
+      "loss": 0.6836,
+      "step": 5929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44572824626082763,
+      "learning_rate": 1.8889735790215036e-06,
+      "loss": 0.6855,
+      "step": 5930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7626532393770764,
+      "learning_rate": 1.888929332460445e-06,
+      "loss": 0.6855,
+      "step": 5931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1850451194984397,
+      "learning_rate": 1.8888850776029303e-06,
+      "loss": 0.6914,
+      "step": 5932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8370436254553395,
+      "learning_rate": 1.8888408144493726e-06,
+      "loss": 0.696,
+      "step": 5933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.53940411260967,
+      "learning_rate": 1.888796543000185e-06,
+      "loss": 0.6895,
+      "step": 5934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.955040093059553,
+      "learning_rate": 1.8887522632557804e-06,
+      "loss": 0.6973,
+      "step": 5935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9333274597250709,
+      "learning_rate": 1.8887079752165726e-06,
+      "loss": 0.6693,
+      "step": 5936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.457839301694274,
+      "learning_rate": 1.8886636788829746e-06,
+      "loss": 0.6953,
+      "step": 5937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6493551063465595,
+      "learning_rate": 1.8886193742554e-06,
+      "loss": 0.681,
+      "step": 5938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0924019833406042,
+      "learning_rate": 1.888575061334262e-06,
+      "loss": 0.7005,
+      "step": 5939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7148133088168906,
+      "learning_rate": 1.8885307401199743e-06,
+      "loss": 0.6849,
+      "step": 5940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0493661953047873,
+      "learning_rate": 1.888486410612951e-06,
+      "loss": 0.6901,
+      "step": 5941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8734875794944658,
+      "learning_rate": 1.8884420728136051e-06,
+      "loss": 0.6868,
+      "step": 5942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.766719024882909,
+      "learning_rate": 1.888397726722351e-06,
+      "loss": 0.6745,
+      "step": 5943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5105398788238382,
+      "learning_rate": 1.8883533723396024e-06,
+      "loss": 0.6836,
+      "step": 5944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1367296478964994,
+      "learning_rate": 1.888309009665773e-06,
+      "loss": 0.6725,
+      "step": 5945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9293876189110004,
+      "learning_rate": 1.888264638701277e-06,
+      "loss": 0.6797,
+      "step": 5946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.140309529467257,
+      "learning_rate": 1.8882202594465285e-06,
+      "loss": 0.6745,
+      "step": 5947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.974879032212208,
+      "learning_rate": 1.8881758719019423e-06,
+      "loss": 0.6992,
+      "step": 5948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.628904218236352,
+      "learning_rate": 1.8881314760679316e-06,
+      "loss": 0.6849,
+      "step": 5949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.217959444648854,
+      "learning_rate": 1.8880870719449118e-06,
+      "loss": 0.681,
+      "step": 5950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.354262813531418,
+      "learning_rate": 1.8880426595332964e-06,
+      "loss": 0.6895,
+      "step": 5951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.099917001343992,
+      "learning_rate": 1.8879982388335005e-06,
+      "loss": 0.6686,
+      "step": 5952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.550786764285134,
+      "learning_rate": 1.8879538098459382e-06,
+      "loss": 0.6712,
+      "step": 5953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.696528847182234,
+      "learning_rate": 1.8879093725710248e-06,
+      "loss": 0.694,
+      "step": 5954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6839875085368214,
+      "learning_rate": 1.8878649270091747e-06,
+      "loss": 0.6901,
+      "step": 5955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5739838542425084,
+      "learning_rate": 1.8878204731608024e-06,
+      "loss": 0.6855,
+      "step": 5956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.2162247976892795,
+      "learning_rate": 1.8877760110263235e-06,
+      "loss": 0.6888,
+      "step": 5957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1096659028454647,
+      "learning_rate": 1.8877315406061525e-06,
+      "loss": 0.6816,
+      "step": 5958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.210146364766292,
+      "learning_rate": 1.8876870619007047e-06,
+      "loss": 0.6934,
+      "step": 5959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.301320020345402,
+      "learning_rate": 1.8876425749103946e-06,
+      "loss": 0.6992,
+      "step": 5960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.882241065143536,
+      "learning_rate": 1.8875980796356381e-06,
+      "loss": 0.6667,
+      "step": 5961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.023078893505202,
+      "learning_rate": 1.8875535760768501e-06,
+      "loss": 0.7096,
+      "step": 5962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7222346473053234,
+      "learning_rate": 1.8875090642344462e-06,
+      "loss": 0.6816,
+      "step": 5963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5341470858919406,
+      "learning_rate": 1.8874645441088415e-06,
+      "loss": 0.6888,
+      "step": 5964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0479986040652913,
+      "learning_rate": 1.887420015700452e-06,
+      "loss": 0.6732,
+      "step": 5965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.266359507095793,
+      "learning_rate": 1.887375479009693e-06,
+      "loss": 0.6693,
+      "step": 5966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1028948965096506,
+      "learning_rate": 1.88733093403698e-06,
+      "loss": 0.6699,
+      "step": 5967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.883463622843717,
+      "learning_rate": 1.887286380782729e-06,
+      "loss": 0.7038,
+      "step": 5968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3449302022011655,
+      "learning_rate": 1.8872418192473557e-06,
+      "loss": 0.6693,
+      "step": 5969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2579460540730516,
+      "learning_rate": 1.8871972494312761e-06,
+      "loss": 0.6816,
+      "step": 5970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.61156020828246,
+      "learning_rate": 1.887152671334906e-06,
+      "loss": 0.6732,
+      "step": 5971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.065771089784037,
+      "learning_rate": 1.8871080849586617e-06,
+      "loss": 0.6758,
+      "step": 5972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.354945728155412,
+      "learning_rate": 1.8870634903029594e-06,
+      "loss": 0.6706,
+      "step": 5973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8217570628738906,
+      "learning_rate": 1.8870188873682147e-06,
+      "loss": 0.6823,
+      "step": 5974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.760994323332844,
+      "learning_rate": 1.8869742761548446e-06,
+      "loss": 0.6523,
+      "step": 5975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7635657313363198,
+      "learning_rate": 1.8869296566632648e-06,
+      "loss": 0.6693,
+      "step": 5976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48390272040181975,
+      "learning_rate": 1.8868850288938925e-06,
+      "loss": 0.681,
+      "step": 5977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.872528385455043,
+      "learning_rate": 1.8868403928471436e-06,
+      "loss": 0.6947,
+      "step": 5978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.65368588047377,
+      "learning_rate": 1.8867957485234345e-06,
+      "loss": 0.6855,
+      "step": 5979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.972599505078751,
+      "learning_rate": 1.8867510959231827e-06,
+      "loss": 0.6751,
+      "step": 5980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6433032202100978,
+      "learning_rate": 1.8867064350468045e-06,
+      "loss": 0.6888,
+      "step": 5981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6182420667879165,
+      "learning_rate": 1.8866617658947166e-06,
+      "loss": 0.7031,
+      "step": 5982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.9580094472208405,
+      "learning_rate": 1.8866170884673358e-06,
+      "loss": 0.6842,
+      "step": 5983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.502343373471326,
+      "learning_rate": 1.8865724027650797e-06,
+      "loss": 0.7018,
+      "step": 5984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.270028322515516,
+      "learning_rate": 1.8865277087883649e-06,
+      "loss": 0.6803,
+      "step": 5985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6640080874880314,
+      "learning_rate": 1.8864830065376083e-06,
+      "loss": 0.6647,
+      "step": 5986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.816141769688084,
+      "learning_rate": 1.8864382960132278e-06,
+      "loss": 0.6901,
+      "step": 5987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5171170591267734,
+      "learning_rate": 1.8863935772156401e-06,
+      "loss": 0.6699,
+      "step": 5988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6090296088861038,
+      "learning_rate": 1.8863488501452626e-06,
+      "loss": 0.679,
+      "step": 5989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4244379743871818,
+      "learning_rate": 1.8863041148025133e-06,
+      "loss": 0.6992,
+      "step": 5990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.099237869868617,
+      "learning_rate": 1.8862593711878087e-06,
+      "loss": 0.6816,
+      "step": 5991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.682223145391384,
+      "learning_rate": 1.8862146193015675e-06,
+      "loss": 0.7012,
+      "step": 5992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4458408719530503,
+      "learning_rate": 1.8861698591442068e-06,
+      "loss": 0.6771,
+      "step": 5993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.129857328436434,
+      "learning_rate": 1.8861250907161443e-06,
+      "loss": 0.6523,
+      "step": 5994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.493944845580579,
+      "learning_rate": 1.886080314017798e-06,
+      "loss": 0.6556,
+      "step": 5995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4859104721704812,
+      "learning_rate": 1.886035529049586e-06,
+      "loss": 0.6914,
+      "step": 5996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5293741387621629,
+      "learning_rate": 1.8859907358119257e-06,
+      "loss": 0.6875,
+      "step": 5997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.350066854853013,
+      "learning_rate": 1.8859459343052357e-06,
+      "loss": 0.6973,
+      "step": 5998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8626651927702126,
+      "learning_rate": 1.885901124529934e-06,
+      "loss": 0.696,
+      "step": 5999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5733025126743571,
+      "learning_rate": 1.8858563064864389e-06,
+      "loss": 0.668,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.464890044198618,
+      "learning_rate": 1.8858114801751683e-06,
+      "loss": 0.666,
+      "step": 6001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.1018674650440876,
+      "learning_rate": 1.8857666455965408e-06,
+      "loss": 0.6693,
+      "step": 6002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.0180103345636065,
+      "learning_rate": 1.8857218027509749e-06,
+      "loss": 0.7018,
+      "step": 6003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.024105126265531,
+      "learning_rate": 1.8856769516388891e-06,
+      "loss": 0.6667,
+      "step": 6004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.032977837141615,
+      "learning_rate": 1.8856320922607019e-06,
+      "loss": 0.668,
+      "step": 6005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.297719463975125,
+      "learning_rate": 1.8855872246168323e-06,
+      "loss": 0.6908,
+      "step": 6006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.735384013978389,
+      "learning_rate": 1.8855423487076987e-06,
+      "loss": 0.6836,
+      "step": 6007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0994179840775584,
+      "learning_rate": 1.8854974645337198e-06,
+      "loss": 0.6745,
+      "step": 6008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.282317948435615,
+      "learning_rate": 1.885452572095315e-06,
+      "loss": 0.6868,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9676656816860187,
+      "learning_rate": 1.8854076713929032e-06,
+      "loss": 0.6621,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6008820889491513,
+      "learning_rate": 1.8853627624269031e-06,
+      "loss": 0.6777,
+      "step": 6011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.9562639656838388,
+      "learning_rate": 1.8853178451977341e-06,
+      "loss": 0.6686,
+      "step": 6012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.8237289057665285,
+      "learning_rate": 1.885272919705815e-06,
+      "loss": 0.6882,
+      "step": 6013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.213926761254298,
+      "learning_rate": 1.885227985951566e-06,
+      "loss": 0.6706,
+      "step": 6014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.483793956601741,
+      "learning_rate": 1.8851830439354056e-06,
+      "loss": 0.6725,
+      "step": 6015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.907963826185097,
+      "learning_rate": 1.8851380936577537e-06,
+      "loss": 0.6921,
+      "step": 6016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.691494089253121,
+      "learning_rate": 1.8850931351190297e-06,
+      "loss": 0.6921,
+      "step": 6017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2814534323095372,
+      "learning_rate": 1.885048168319653e-06,
+      "loss": 0.6732,
+      "step": 6018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8851839864016313,
+      "learning_rate": 1.8850031932600436e-06,
+      "loss": 0.7005,
+      "step": 6019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.886868335372431,
+      "learning_rate": 1.8849582099406215e-06,
+      "loss": 0.6751,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4141747015703596,
+      "learning_rate": 1.8849132183618055e-06,
+      "loss": 0.6882,
+      "step": 6021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4247014842118753,
+      "learning_rate": 1.8848682185240166e-06,
+      "loss": 0.6875,
+      "step": 6022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.692255152552669,
+      "learning_rate": 1.8848232104276742e-06,
+      "loss": 0.6738,
+      "step": 6023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.825028458782002,
+      "learning_rate": 1.8847781940731985e-06,
+      "loss": 0.6764,
+      "step": 6024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2113287202557455,
+      "learning_rate": 1.8847331694610098e-06,
+      "loss": 0.7083,
+      "step": 6025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.836678885344998,
+      "learning_rate": 1.884688136591528e-06,
+      "loss": 0.6875,
+      "step": 6026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.188091833741748,
+      "learning_rate": 1.884643095465174e-06,
+      "loss": 0.6855,
+      "step": 6027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4188851253359345,
+      "learning_rate": 1.8845980460823674e-06,
+      "loss": 0.6882,
+      "step": 6028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.987620387229338,
+      "learning_rate": 1.884552988443529e-06,
+      "loss": 0.6738,
+      "step": 6029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4630166285180066,
+      "learning_rate": 1.8845079225490794e-06,
+      "loss": 0.6803,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.858384491786981,
+      "learning_rate": 1.8844628483994392e-06,
+      "loss": 0.6823,
+      "step": 6031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9890080263092869,
+      "learning_rate": 1.8844177659950292e-06,
+      "loss": 0.6797,
+      "step": 6032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.299530144488247,
+      "learning_rate": 1.8843726753362693e-06,
+      "loss": 0.679,
+      "step": 6033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3923171786144533,
+      "learning_rate": 1.8843275764235815e-06,
+      "loss": 0.6868,
+      "step": 6034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.106263975899353,
+      "learning_rate": 1.8842824692573864e-06,
+      "loss": 0.6693,
+      "step": 6035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8993429348511293,
+      "learning_rate": 1.8842373538381048e-06,
+      "loss": 0.6836,
+      "step": 6036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5801176283785634,
+      "learning_rate": 1.8841922301661572e-06,
+      "loss": 0.7038,
+      "step": 6037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5682367403428263,
+      "learning_rate": 1.8841470982419658e-06,
+      "loss": 0.6823,
+      "step": 6038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48531001128223716,
+      "learning_rate": 1.8841019580659513e-06,
+      "loss": 0.6849,
+      "step": 6039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5797631729245536,
+      "learning_rate": 1.884056809638535e-06,
+      "loss": 0.6784,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.175786152753635,
+      "learning_rate": 1.8840116529601383e-06,
+      "loss": 0.6855,
+      "step": 6041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8698695707318717,
+      "learning_rate": 1.8839664880311825e-06,
+      "loss": 0.6751,
+      "step": 6042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8673113832849126,
+      "learning_rate": 1.8839213148520896e-06,
+      "loss": 0.6732,
+      "step": 6043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.557522412613059,
+      "learning_rate": 1.8838761334232806e-06,
+      "loss": 0.6725,
+      "step": 6044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0060818851257154,
+      "learning_rate": 1.8838309437451776e-06,
+      "loss": 0.6589,
+      "step": 6045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5011743499824566,
+      "learning_rate": 1.8837857458182022e-06,
+      "loss": 0.694,
+      "step": 6046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7398474972018584,
+      "learning_rate": 1.8837405396427765e-06,
+      "loss": 0.6908,
+      "step": 6047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.005524495762091,
+      "learning_rate": 1.883695325219322e-06,
+      "loss": 0.6934,
+      "step": 6048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.833593368916091,
+      "learning_rate": 1.8836501025482612e-06,
+      "loss": 0.6934,
+      "step": 6049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.373829072090312,
+      "learning_rate": 1.8836048716300153e-06,
+      "loss": 0.6686,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8542914568335578,
+      "learning_rate": 1.8835596324650073e-06,
+      "loss": 0.6855,
+      "step": 6051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5754427868132655,
+      "learning_rate": 1.883514385053659e-06,
+      "loss": 0.6751,
+      "step": 6052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2453027302569004,
+      "learning_rate": 1.883469129396393e-06,
+      "loss": 0.6745,
+      "step": 6053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.712817185198309,
+      "learning_rate": 1.8834238654936314e-06,
+      "loss": 0.6784,
+      "step": 6054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.136373523454478,
+      "learning_rate": 1.8833785933457965e-06,
+      "loss": 0.6771,
+      "step": 6055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7011422092847055,
+      "learning_rate": 1.8833333129533114e-06,
+      "loss": 0.6725,
+      "step": 6056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.819230125584055,
+      "learning_rate": 1.8832880243165983e-06,
+      "loss": 0.6986,
+      "step": 6057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5934520530716723,
+      "learning_rate": 1.8832427274360798e-06,
+      "loss": 0.7161,
+      "step": 6058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.249379468633216,
+      "learning_rate": 1.8831974223121789e-06,
+      "loss": 0.6816,
+      "step": 6059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8795514849344822,
+      "learning_rate": 1.8831521089453184e-06,
+      "loss": 0.6823,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9291867566870287,
+      "learning_rate": 1.883106787335921e-06,
+      "loss": 0.6888,
+      "step": 6061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3753892241503238,
+      "learning_rate": 1.88306145748441e-06,
+      "loss": 0.7018,
+      "step": 6062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9561582806983127,
+      "learning_rate": 1.8830161193912085e-06,
+      "loss": 0.6875,
+      "step": 6063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.049211700488271,
+      "learning_rate": 1.8829707730567393e-06,
+      "loss": 0.6934,
+      "step": 6064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.301690386702933,
+      "learning_rate": 1.8829254184814257e-06,
+      "loss": 0.6934,
+      "step": 6065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.644302390746597,
+      "learning_rate": 1.8828800556656912e-06,
+      "loss": 0.6758,
+      "step": 6066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4000054373051016,
+      "learning_rate": 1.8828346846099588e-06,
+      "loss": 0.6745,
+      "step": 6067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.625655059547706,
+      "learning_rate": 1.8827893053146525e-06,
+      "loss": 0.6745,
+      "step": 6068
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3812251696247182,
+      "learning_rate": 1.8827439177801954e-06,
+      "loss": 0.6882,
+      "step": 6069
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.747622111789243,
+      "learning_rate": 1.8826985220070112e-06,
+      "loss": 0.6842,
+      "step": 6070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6521795223549995,
+      "learning_rate": 1.8826531179955238e-06,
+      "loss": 0.6641,
+      "step": 6071
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5038471575312717,
+      "learning_rate": 1.8826077057461567e-06,
+      "loss": 0.6947,
+      "step": 6072
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.732587895363916,
+      "learning_rate": 1.8825622852593338e-06,
+      "loss": 0.6914,
+      "step": 6073
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4037292313188967,
+      "learning_rate": 1.8825168565354791e-06,
+      "loss": 0.6849,
+      "step": 6074
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.102885447773246,
+      "learning_rate": 1.8824714195750167e-06,
+      "loss": 0.6738,
+      "step": 6075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.215545158815517,
+      "learning_rate": 1.8824259743783704e-06,
+      "loss": 0.6901,
+      "step": 6076
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5760456386281643,
+      "learning_rate": 1.8823805209459645e-06,
+      "loss": 0.6777,
+      "step": 6077
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6581935610562029,
+      "learning_rate": 1.882335059278223e-06,
+      "loss": 0.681,
+      "step": 6078
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0802602734461058,
+      "learning_rate": 1.8822895893755704e-06,
+      "loss": 0.6764,
+      "step": 6079
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5928051125234637,
+      "learning_rate": 1.8822441112384309e-06,
+      "loss": 0.6712,
+      "step": 6080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5896587694785094,
+      "learning_rate": 1.8821986248672292e-06,
+      "loss": 0.6953,
+      "step": 6081
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.306732847694067,
+      "learning_rate": 1.8821531302623898e-06,
+      "loss": 0.6725,
+      "step": 6082
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0119654174088866,
+      "learning_rate": 1.8821076274243372e-06,
+      "loss": 0.6738,
+      "step": 6083
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.439744356028769,
+      "learning_rate": 1.8820621163534962e-06,
+      "loss": 0.694,
+      "step": 6084
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4374615239910034,
+      "learning_rate": 1.8820165970502912e-06,
+      "loss": 0.7012,
+      "step": 6085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6399016218744307,
+      "learning_rate": 1.8819710695151474e-06,
+      "loss": 0.6758,
+      "step": 6086
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5691375358758142,
+      "learning_rate": 1.8819255337484899e-06,
+      "loss": 0.6849,
+      "step": 6087
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5235024591318997,
+      "learning_rate": 1.8818799897507432e-06,
+      "loss": 0.6823,
+      "step": 6088
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4741234495134394,
+      "learning_rate": 1.8818344375223322e-06,
+      "loss": 0.6803,
+      "step": 6089
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.573905421543656,
+      "learning_rate": 1.8817888770636828e-06,
+      "loss": 0.6751,
+      "step": 6090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9823815099891096,
+      "learning_rate": 1.8817433083752198e-06,
+      "loss": 0.6934,
+      "step": 6091
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8530606579148706,
+      "learning_rate": 1.8816977314573686e-06,
+      "loss": 0.6732,
+      "step": 6092
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.747748897878654,
+      "learning_rate": 1.8816521463105542e-06,
+      "loss": 0.6719,
+      "step": 6093
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.036632743830533,
+      "learning_rate": 1.8816065529352025e-06,
+      "loss": 0.6823,
+      "step": 6094
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6975322767156973,
+      "learning_rate": 1.881560951331739e-06,
+      "loss": 0.6706,
+      "step": 6095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.968407483522623,
+      "learning_rate": 1.881515341500589e-06,
+      "loss": 0.681,
+      "step": 6096
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0612025723979013,
+      "learning_rate": 1.8814697234421786e-06,
+      "loss": 0.6576,
+      "step": 6097
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.163307171707782,
+      "learning_rate": 1.8814240971569331e-06,
+      "loss": 0.681,
+      "step": 6098
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.830706652031418,
+      "learning_rate": 1.8813784626452787e-06,
+      "loss": 0.6868,
+      "step": 6099
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8388815101065354,
+      "learning_rate": 1.8813328199076407e-06,
+      "loss": 0.6914,
+      "step": 6100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7948955102673124,
+      "learning_rate": 1.8812871689444463e-06,
+      "loss": 0.6875,
+      "step": 6101
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1503160342313863,
+      "learning_rate": 1.8812415097561205e-06,
+      "loss": 0.6862,
+      "step": 6102
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7518234846080198,
+      "learning_rate": 1.8811958423430897e-06,
+      "loss": 0.7031,
+      "step": 6103
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9546062676564278,
+      "learning_rate": 1.88115016670578e-06,
+      "loss": 0.6888,
+      "step": 6104
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8461077758234157,
+      "learning_rate": 1.881104482844618e-06,
+      "loss": 0.681,
+      "step": 6105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7311730123993399,
+      "learning_rate": 1.88105879076003e-06,
+      "loss": 0.6921,
+      "step": 6106
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3936392127326522,
+      "learning_rate": 1.8810130904524426e-06,
+      "loss": 0.6777,
+      "step": 6107
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2089286827172594,
+      "learning_rate": 1.8809673819222817e-06,
+      "loss": 0.6725,
+      "step": 6108
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.239619932410184,
+      "learning_rate": 1.8809216651699745e-06,
+      "loss": 0.6764,
+      "step": 6109
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5862825706348201,
+      "learning_rate": 1.8808759401959476e-06,
+      "loss": 0.6816,
+      "step": 6110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.721777406138191,
+      "learning_rate": 1.8808302070006275e-06,
+      "loss": 0.653,
+      "step": 6111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.865296966448042,
+      "learning_rate": 1.8807844655844414e-06,
+      "loss": 0.6849,
+      "step": 6112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4061529564410282,
+      "learning_rate": 1.8807387159478157e-06,
+      "loss": 0.6712,
+      "step": 6113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5545552404940826,
+      "learning_rate": 1.8806929580911777e-06,
+      "loss": 0.6784,
+      "step": 6114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.125158433099163,
+      "learning_rate": 1.8806471920149547e-06,
+      "loss": 0.6523,
+      "step": 6115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3051107475380646,
+      "learning_rate": 1.8806014177195733e-06,
+      "loss": 0.6784,
+      "step": 6116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.886778035529972,
+      "learning_rate": 1.8805556352054613e-06,
+      "loss": 0.6608,
+      "step": 6117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.071038101137393,
+      "learning_rate": 1.8805098444730452e-06,
+      "loss": 0.6576,
+      "step": 6118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.091279952869818,
+      "learning_rate": 1.8804640455227531e-06,
+      "loss": 0.6829,
+      "step": 6119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7177715068237671,
+      "learning_rate": 1.8804182383550122e-06,
+      "loss": 0.6842,
+      "step": 6120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.9655902614752728,
+      "learning_rate": 1.8803724229702501e-06,
+      "loss": 0.6452,
+      "step": 6121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0890980844881497,
+      "learning_rate": 1.8803265993688943e-06,
+      "loss": 0.6842,
+      "step": 6122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6961885699445306,
+      "learning_rate": 1.8802807675513725e-06,
+      "loss": 0.6797,
+      "step": 6123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8874562142894775,
+      "learning_rate": 1.8802349275181125e-06,
+      "loss": 0.6934,
+      "step": 6124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.515963800772158,
+      "learning_rate": 1.880189079269542e-06,
+      "loss": 0.6738,
+      "step": 6125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.368333420988293,
+      "learning_rate": 1.880143222806089e-06,
+      "loss": 0.709,
+      "step": 6126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.182270738279412,
+      "learning_rate": 1.8800973581281814e-06,
+      "loss": 0.6836,
+      "step": 6127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.637709790225781,
+      "learning_rate": 1.8800514852362475e-06,
+      "loss": 0.6895,
+      "step": 6128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5813215020100773,
+      "learning_rate": 1.8800056041307148e-06,
+      "loss": 0.6992,
+      "step": 6129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2914265030524057,
+      "learning_rate": 1.8799597148120126e-06,
+      "loss": 0.6751,
+      "step": 6130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7370414241684106,
+      "learning_rate": 1.8799138172805683e-06,
+      "loss": 0.6816,
+      "step": 6131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.831891766932348,
+      "learning_rate": 1.8798679115368103e-06,
+      "loss": 0.6595,
+      "step": 6132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.875448559305072,
+      "learning_rate": 1.8798219975811676e-06,
+      "loss": 0.6777,
+      "step": 6133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.344723877938696,
+      "learning_rate": 1.8797760754140684e-06,
+      "loss": 0.6947,
+      "step": 6134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7203431243155316,
+      "learning_rate": 1.8797301450359408e-06,
+      "loss": 0.6699,
+      "step": 6135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.511045571179299,
+      "learning_rate": 1.8796842064472144e-06,
+      "loss": 0.6829,
+      "step": 6136
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.906644915568238,
+      "learning_rate": 1.8796382596483173e-06,
+      "loss": 0.6751,
+      "step": 6137
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1582376234805443,
+      "learning_rate": 1.8795923046396789e-06,
+      "loss": 0.6947,
+      "step": 6138
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.826278594034733,
+      "learning_rate": 1.8795463414217272e-06,
+      "loss": 0.6797,
+      "step": 6139
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5220173386385696,
+      "learning_rate": 1.8795003699948922e-06,
+      "loss": 0.6725,
+      "step": 6140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.174973510640079,
+      "learning_rate": 1.8794543903596022e-06,
+      "loss": 0.7057,
+      "step": 6141
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.311719883047002,
+      "learning_rate": 1.8794084025162866e-06,
+      "loss": 0.6803,
+      "step": 6142
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5769429967538163,
+      "learning_rate": 1.8793624064653747e-06,
+      "loss": 0.6927,
+      "step": 6143
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4678049200116807,
+      "learning_rate": 1.8793164022072956e-06,
+      "loss": 0.6758,
+      "step": 6144
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7472165687822417,
+      "learning_rate": 1.8792703897424788e-06,
+      "loss": 0.6771,
+      "step": 6145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.230286841998463,
+      "learning_rate": 1.8792243690713536e-06,
+      "loss": 0.6862,
+      "step": 6146
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7084804649065791,
+      "learning_rate": 1.87917834019435e-06,
+      "loss": 0.6855,
+      "step": 6147
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.263563409104156,
+      "learning_rate": 1.8791323031118966e-06,
+      "loss": 0.6979,
+      "step": 6148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7830303290470717,
+      "learning_rate": 1.8790862578244239e-06,
+      "loss": 0.696,
+      "step": 6149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.784585536310086,
+      "learning_rate": 1.8790402043323618e-06,
+      "loss": 0.6862,
+      "step": 6150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5987106435521583,
+      "learning_rate": 1.8789941426361394e-06,
+      "loss": 0.6764,
+      "step": 6151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2000544226692056,
+      "learning_rate": 1.878948072736187e-06,
+      "loss": 0.6999,
+      "step": 6152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.005173350914075,
+      "learning_rate": 1.8789019946329345e-06,
+      "loss": 0.6816,
+      "step": 6153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7949680110606188,
+      "learning_rate": 1.8788559083268118e-06,
+      "loss": 0.681,
+      "step": 6154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0443842233319707,
+      "learning_rate": 1.8788098138182493e-06,
+      "loss": 0.6823,
+      "step": 6155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7691718212915561,
+      "learning_rate": 1.878763711107677e-06,
+      "loss": 0.7038,
+      "step": 6156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8000935614452913,
+      "learning_rate": 1.8787176001955257e-06,
+      "loss": 0.6868,
+      "step": 6157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7336689690708518,
+      "learning_rate": 1.8786714810822248e-06,
+      "loss": 0.6953,
+      "step": 6158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.442379505638934,
+      "learning_rate": 1.8786253537682054e-06,
+      "loss": 0.6745,
+      "step": 6159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.927132345459361,
+      "learning_rate": 1.878579218253898e-06,
+      "loss": 0.6875,
+      "step": 6160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6588504172178866,
+      "learning_rate": 1.878533074539733e-06,
+      "loss": 0.7109,
+      "step": 6161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1138596000204593,
+      "learning_rate": 1.878486922626141e-06,
+      "loss": 0.6615,
+      "step": 6162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.887555794877082,
+      "learning_rate": 1.878440762513553e-06,
+      "loss": 0.6849,
+      "step": 6163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.33471874772784,
+      "learning_rate": 1.8783945942023998e-06,
+      "loss": 0.6849,
+      "step": 6164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.0915350581439025,
+      "learning_rate": 1.8783484176931119e-06,
+      "loss": 0.6719,
+      "step": 6165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.415383543217328,
+      "learning_rate": 1.8783022329861208e-06,
+      "loss": 0.679,
+      "step": 6166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.007995383306124,
+      "learning_rate": 1.878256040081857e-06,
+      "loss": 0.6745,
+      "step": 6167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9124652323397013,
+      "learning_rate": 1.8782098389807521e-06,
+      "loss": 0.6882,
+      "step": 6168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1793158942752007,
+      "learning_rate": 1.878163629683237e-06,
+      "loss": 0.6855,
+      "step": 6169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.1752960552591762,
+      "learning_rate": 1.8781174121897431e-06,
+      "loss": 0.6816,
+      "step": 6170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.42368932577792084,
+      "learning_rate": 1.8780711865007017e-06,
+      "loss": 0.6979,
+      "step": 6171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2523429348503428,
+      "learning_rate": 1.8780249526165446e-06,
+      "loss": 0.6816,
+      "step": 6172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3966047113780293,
+      "learning_rate": 1.8779787105377026e-06,
+      "loss": 0.6979,
+      "step": 6173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.7908563339441566,
+      "learning_rate": 1.8779324602646079e-06,
+      "loss": 0.6764,
+      "step": 6174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6067756891932039,
+      "learning_rate": 1.8778862017976917e-06,
+      "loss": 0.6927,
+      "step": 6175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.764104854925303,
+      "learning_rate": 1.877839935137386e-06,
+      "loss": 0.6797,
+      "step": 6176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5998041885855347,
+      "learning_rate": 1.8777936602841225e-06,
+      "loss": 0.6738,
+      "step": 6177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.055760726803181,
+      "learning_rate": 1.877747377238333e-06,
+      "loss": 0.694,
+      "step": 6178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2231527383597511,
+      "learning_rate": 1.8777010860004502e-06,
+      "loss": 0.6719,
+      "step": 6179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.9556769392410246,
+      "learning_rate": 1.877654786570905e-06,
+      "loss": 0.6823,
+      "step": 6180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.3567632170013075,
+      "learning_rate": 1.8776084789501303e-06,
+      "loss": 0.6654,
+      "step": 6181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.89278727280193,
+      "learning_rate": 1.877562163138558e-06,
+      "loss": 0.6764,
+      "step": 6182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.726852024726883,
+      "learning_rate": 1.8775158391366205e-06,
+      "loss": 0.7025,
+      "step": 6183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2911447818926565,
+      "learning_rate": 1.8774695069447498e-06,
+      "loss": 0.7025,
+      "step": 6184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.0384862800241805,
+      "learning_rate": 1.8774231665633787e-06,
+      "loss": 0.6699,
+      "step": 6185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.772990816042378,
+      "learning_rate": 1.8773768179929398e-06,
+      "loss": 0.6771,
+      "step": 6186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6252504806886091,
+      "learning_rate": 1.8773304612338654e-06,
+      "loss": 0.6641,
+      "step": 6187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.66257344769312,
+      "learning_rate": 1.877284096286588e-06,
+      "loss": 0.6803,
+      "step": 6188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.680656282957612,
+      "learning_rate": 1.877237723151541e-06,
+      "loss": 0.6895,
+      "step": 6189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.762970558650049,
+      "learning_rate": 1.8771913418291561e-06,
+      "loss": 0.6875,
+      "step": 6190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48959903768348756,
+      "learning_rate": 1.8771449523198672e-06,
+      "loss": 0.6641,
+      "step": 6191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.992626117754683,
+      "learning_rate": 1.8770985546241066e-06,
+      "loss": 0.6882,
+      "step": 6192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.638161507322233,
+      "learning_rate": 1.8770521487423082e-06,
+      "loss": 0.709,
+      "step": 6193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.920234515492739,
+      "learning_rate": 1.877005734674904e-06,
+      "loss": 0.6719,
+      "step": 6194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4117581499734904,
+      "learning_rate": 1.8769593124223283e-06,
+      "loss": 0.6901,
+      "step": 6195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.170472830161039,
+      "learning_rate": 1.8769128819850134e-06,
+      "loss": 0.6803,
+      "step": 6196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.43099257150497483,
+      "learning_rate": 1.8768664433633927e-06,
+      "loss": 0.6836,
+      "step": 6197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1015737566736146,
+      "learning_rate": 1.8768199965579005e-06,
+      "loss": 0.6686,
+      "step": 6198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.47372490111568766,
+      "learning_rate": 1.8767735415689694e-06,
+      "loss": 0.6823,
+      "step": 6199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.213520214559056,
+      "learning_rate": 1.8767270783970335e-06,
+      "loss": 0.7044,
+      "step": 6200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.911652983107594,
+      "learning_rate": 1.8766806070425262e-06,
+      "loss": 0.6927,
+      "step": 6201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.07928120727656,
+      "learning_rate": 1.876634127505881e-06,
+      "loss": 0.6777,
+      "step": 6202
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4741495352108123,
+      "learning_rate": 1.8765876397875322e-06,
+      "loss": 0.6895,
+      "step": 6203
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.732014873990217,
+      "learning_rate": 1.8765411438879135e-06,
+      "loss": 0.6992,
+      "step": 6204
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8560390742360773,
+      "learning_rate": 1.8764946398074586e-06,
+      "loss": 0.7044,
+      "step": 6205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.399775164065908,
+      "learning_rate": 1.876448127546602e-06,
+      "loss": 0.681,
+      "step": 6206
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7469605188153049,
+      "learning_rate": 1.8764016071057772e-06,
+      "loss": 0.7083,
+      "step": 6207
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2750115664609463,
+      "learning_rate": 1.8763550784854187e-06,
+      "loss": 0.6549,
+      "step": 6208
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0786187351003744,
+      "learning_rate": 1.876308541685961e-06,
+      "loss": 0.6882,
+      "step": 6209
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.1234799373092645,
+      "learning_rate": 1.8762619967078378e-06,
+      "loss": 0.7057,
+      "step": 6210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.089771538981556,
+      "learning_rate": 1.8762154435514845e-06,
+      "loss": 0.6986,
+      "step": 6211
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0122832079675768,
+      "learning_rate": 1.8761688822173346e-06,
+      "loss": 0.6862,
+      "step": 6212
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.891328979444926,
+      "learning_rate": 1.876122312705823e-06,
+      "loss": 0.7018,
+      "step": 6213
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7117377800899076,
+      "learning_rate": 1.8760757350173844e-06,
+      "loss": 0.7038,
+      "step": 6214
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.916992244684363,
+      "learning_rate": 1.8760291491524533e-06,
+      "loss": 0.6934,
+      "step": 6215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.789389599519961,
+      "learning_rate": 1.8759825551114651e-06,
+      "loss": 0.6725,
+      "step": 6216
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1242620647151282,
+      "learning_rate": 1.8759359528948542e-06,
+      "loss": 0.6829,
+      "step": 6217
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.511137271169675,
+      "learning_rate": 1.8758893425030554e-06,
+      "loss": 0.679,
+      "step": 6218
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.423259460439939,
+      "learning_rate": 1.8758427239365037e-06,
+      "loss": 0.6947,
+      "step": 6219
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8327346823704251,
+      "learning_rate": 1.8757960971956348e-06,
+      "loss": 0.6862,
+      "step": 6220
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1178465778429723,
+      "learning_rate": 1.8757494622808833e-06,
+      "loss": 0.7012,
+      "step": 6221
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.300442613107654,
+      "learning_rate": 1.8757028191926846e-06,
+      "loss": 0.6829,
+      "step": 6222
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.569528078995323,
+      "learning_rate": 1.8756561679314741e-06,
+      "loss": 0.6712,
+      "step": 6223
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9803021800471802,
+      "learning_rate": 1.8756095084976872e-06,
+      "loss": 0.6738,
+      "step": 6224
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.773072760404691,
+      "learning_rate": 1.8755628408917592e-06,
+      "loss": 0.681,
+      "step": 6225
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.678198647734124,
+      "learning_rate": 1.875516165114126e-06,
+      "loss": 0.6882,
+      "step": 6226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0541082316048915,
+      "learning_rate": 1.8754694811652227e-06,
+      "loss": 0.694,
+      "step": 6227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6174128407044464,
+      "learning_rate": 1.8754227890454855e-06,
+      "loss": 0.6803,
+      "step": 6228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.099745211660662,
+      "learning_rate": 1.8753760887553498e-06,
+      "loss": 0.7038,
+      "step": 6229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.596312507813074,
+      "learning_rate": 1.8753293802952519e-06,
+      "loss": 0.6973,
+      "step": 6230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.709070635868949,
+      "learning_rate": 1.8752826636656275e-06,
+      "loss": 0.6888,
+      "step": 6231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3548520113515385,
+      "learning_rate": 1.8752359388669124e-06,
+      "loss": 0.6829,
+      "step": 6232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.122192036229687,
+      "learning_rate": 1.8751892058995428e-06,
+      "loss": 0.6823,
+      "step": 6233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.725494726286474,
+      "learning_rate": 1.875142464763955e-06,
+      "loss": 0.6895,
+      "step": 6234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.74234457272886,
+      "learning_rate": 1.8750957154605853e-06,
+      "loss": 0.6986,
+      "step": 6235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.665128370446475,
+      "learning_rate": 1.87504895798987e-06,
+      "loss": 0.679,
+      "step": 6236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8172416450028654,
+      "learning_rate": 1.8750021923522454e-06,
+      "loss": 0.6823,
+      "step": 6237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.770881007929116,
+      "learning_rate": 1.8749554185481477e-06,
+      "loss": 0.668,
+      "step": 6238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1257195244116043,
+      "learning_rate": 1.8749086365780138e-06,
+      "loss": 0.668,
+      "step": 6239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6992854328853659,
+      "learning_rate": 1.8748618464422803e-06,
+      "loss": 0.6764,
+      "step": 6240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.220862317281296,
+      "learning_rate": 1.8748150481413835e-06,
+      "loss": 0.7025,
+      "step": 6241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0888690261172767,
+      "learning_rate": 1.8747682416757611e-06,
+      "loss": 0.6654,
+      "step": 6242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7128149675491544,
+      "learning_rate": 1.874721427045849e-06,
+      "loss": 0.6764,
+      "step": 6243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.76593434843563,
+      "learning_rate": 1.8746746042520844e-06,
+      "loss": 0.6693,
+      "step": 6244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8636407149233667,
+      "learning_rate": 1.8746277732949043e-06,
+      "loss": 0.6992,
+      "step": 6245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.730573719157167,
+      "learning_rate": 1.874580934174746e-06,
+      "loss": 0.6803,
+      "step": 6246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5759570786694206,
+      "learning_rate": 1.8745340868920466e-06,
+      "loss": 0.6927,
+      "step": 6247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.042437754233315,
+      "learning_rate": 1.8744872314472429e-06,
+      "loss": 0.7109,
+      "step": 6248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9542384002794655,
+      "learning_rate": 1.8744403678407727e-06,
+      "loss": 0.666,
+      "step": 6249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2615086104423334,
+      "learning_rate": 1.8743934960730732e-06,
+      "loss": 0.6953,
+      "step": 6250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.931363069421727,
+      "learning_rate": 1.874346616144582e-06,
+      "loss": 0.6868,
+      "step": 6251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.19963196113797,
+      "learning_rate": 1.8742997280557365e-06,
+      "loss": 0.6647,
+      "step": 6252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0907548105702705,
+      "learning_rate": 1.874252831806974e-06,
+      "loss": 0.6823,
+      "step": 6253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6148818686446623,
+      "learning_rate": 1.874205927398733e-06,
+      "loss": 0.668,
+      "step": 6254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0457647683471298,
+      "learning_rate": 1.8741590148314503e-06,
+      "loss": 0.6667,
+      "step": 6255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.996532194238737,
+      "learning_rate": 1.8741120941055647e-06,
+      "loss": 0.6654,
+      "step": 6256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2424508840007804,
+      "learning_rate": 1.8740651652215132e-06,
+      "loss": 0.6582,
+      "step": 6257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3859099977996234,
+      "learning_rate": 1.8740182281797345e-06,
+      "loss": 0.6842,
+      "step": 6258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.220474949698748,
+      "learning_rate": 1.8739712829806661e-06,
+      "loss": 0.6966,
+      "step": 6259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4660018772786287,
+      "learning_rate": 1.8739243296247467e-06,
+      "loss": 0.7122,
+      "step": 6260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.057186452608995,
+      "learning_rate": 1.873877368112414e-06,
+      "loss": 0.6719,
+      "step": 6261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7197859060769284,
+      "learning_rate": 1.8738303984441067e-06,
+      "loss": 0.7064,
+      "step": 6262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1761484566829647,
+      "learning_rate": 1.873783420620263e-06,
+      "loss": 0.6634,
+      "step": 6263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6610769664748393,
+      "learning_rate": 1.8737364346413212e-06,
+      "loss": 0.6979,
+      "step": 6264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.369118725330069,
+      "learning_rate": 1.8736894405077202e-06,
+      "loss": 0.6751,
+      "step": 6265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.218109587547842,
+      "learning_rate": 1.8736424382198984e-06,
+      "loss": 0.6699,
+      "step": 6266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.9074205967013995,
+      "learning_rate": 1.8735954277782943e-06,
+      "loss": 0.6934,
+      "step": 6267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6348583982404096,
+      "learning_rate": 1.8735484091833467e-06,
+      "loss": 0.6732,
+      "step": 6268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6078110744219813,
+      "learning_rate": 1.8735013824354949e-06,
+      "loss": 0.6758,
+      "step": 6269
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.380131220304053,
+      "learning_rate": 1.8734543475351773e-06,
+      "loss": 0.6855,
+      "step": 6270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.494520971262866,
+      "learning_rate": 1.8734073044828329e-06,
+      "loss": 0.7077,
+      "step": 6271
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.839757802408945,
+      "learning_rate": 1.8733602532789011e-06,
+      "loss": 0.6816,
+      "step": 6272
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.625046060090498,
+      "learning_rate": 1.8733131939238205e-06,
+      "loss": 0.694,
+      "step": 6273
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.280144576894384,
+      "learning_rate": 1.8732661264180312e-06,
+      "loss": 0.6823,
+      "step": 6274
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8676595532734265,
+      "learning_rate": 1.8732190507619713e-06,
+      "loss": 0.6895,
+      "step": 6275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5719908606782107,
+      "learning_rate": 1.873171966956081e-06,
+      "loss": 0.6667,
+      "step": 6276
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.850119741079184,
+      "learning_rate": 1.8731248750007994e-06,
+      "loss": 0.7109,
+      "step": 6277
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8129121605873284,
+      "learning_rate": 1.8730777748965665e-06,
+      "loss": 0.6849,
+      "step": 6278
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7742996290679682,
+      "learning_rate": 1.8730306666438214e-06,
+      "loss": 0.6771,
+      "step": 6279
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.238361725159281,
+      "learning_rate": 1.8729835502430035e-06,
+      "loss": 0.6823,
+      "step": 6280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4325220317156653,
+      "learning_rate": 1.8729364256945535e-06,
+      "loss": 0.6842,
+      "step": 6281
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8268717321215315,
+      "learning_rate": 1.8728892929989102e-06,
+      "loss": 0.6888,
+      "step": 6282
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2387164580027976,
+      "learning_rate": 1.872842152156514e-06,
+      "loss": 0.6927,
+      "step": 6283
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4272862608445212,
+      "learning_rate": 1.8727950031678051e-06,
+      "loss": 0.6771,
+      "step": 6284
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.39555614108365,
+      "learning_rate": 1.872747846033223e-06,
+      "loss": 0.6934,
+      "step": 6285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.439729036997235,
+      "learning_rate": 1.872700680753208e-06,
+      "loss": 0.6868,
+      "step": 6286
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8308285238606361,
+      "learning_rate": 1.8726535073282004e-06,
+      "loss": 0.6569,
+      "step": 6287
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.317387553130564,
+      "learning_rate": 1.8726063257586407e-06,
+      "loss": 0.6882,
+      "step": 6288
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2035453560848635,
+      "learning_rate": 1.8725591360449686e-06,
+      "loss": 0.6706,
+      "step": 6289
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.210964923766792,
+      "learning_rate": 1.8725119381876252e-06,
+      "loss": 0.6921,
+      "step": 6290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5766066455151826,
+      "learning_rate": 1.8724647321870505e-06,
+      "loss": 0.6823,
+      "step": 6291
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.023601252861584,
+      "learning_rate": 1.8724175180436856e-06,
+      "loss": 0.6966,
+      "step": 6292
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5920727368910712,
+      "learning_rate": 1.8723702957579706e-06,
+      "loss": 0.6764,
+      "step": 6293
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.447167117391492,
+      "learning_rate": 1.8723230653303466e-06,
+      "loss": 0.6654,
+      "step": 6294
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6729492348200283,
+      "learning_rate": 1.8722758267612542e-06,
+      "loss": 0.6751,
+      "step": 6295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0004912980126912,
+      "learning_rate": 1.8722285800511343e-06,
+      "loss": 0.6934,
+      "step": 6296
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5936413208780447,
+      "learning_rate": 1.872181325200428e-06,
+      "loss": 0.6777,
+      "step": 6297
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.070956954593622,
+      "learning_rate": 1.8721340622095764e-06,
+      "loss": 0.6823,
+      "step": 6298
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7523567726323714,
+      "learning_rate": 1.8720867910790204e-06,
+      "loss": 0.6719,
+      "step": 6299
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.207954318165495,
+      "learning_rate": 1.8720395118092012e-06,
+      "loss": 0.6999,
+      "step": 6300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.746730324248295,
+      "learning_rate": 1.8719922244005602e-06,
+      "loss": 0.6914,
+      "step": 6301
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.227737950327188,
+      "learning_rate": 1.8719449288535385e-06,
+      "loss": 0.6895,
+      "step": 6302
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4923046203866352,
+      "learning_rate": 1.8718976251685778e-06,
+      "loss": 0.6595,
+      "step": 6303
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.009896339404022,
+      "learning_rate": 1.8718503133461194e-06,
+      "loss": 0.666,
+      "step": 6304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.771069584633817,
+      "learning_rate": 1.871802993386605e-06,
+      "loss": 0.6979,
+      "step": 6305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9572312615091352,
+      "learning_rate": 1.8717556652904762e-06,
+      "loss": 0.6836,
+      "step": 6306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5454868832478174,
+      "learning_rate": 1.8717083290581745e-06,
+      "loss": 0.6647,
+      "step": 6307
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1465355237270876,
+      "learning_rate": 1.871660984690142e-06,
+      "loss": 0.6738,
+      "step": 6308
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8528575399931512,
+      "learning_rate": 1.8716136321868206e-06,
+      "loss": 0.6797,
+      "step": 6309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.030579952003167,
+      "learning_rate": 1.8715662715486518e-06,
+      "loss": 0.6647,
+      "step": 6310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.900582823308894,
+      "learning_rate": 1.871518902776078e-06,
+      "loss": 0.7031,
+      "step": 6311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.120375172574917,
+      "learning_rate": 1.8714715258695413e-06,
+      "loss": 0.6777,
+      "step": 6312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.462334119493579,
+      "learning_rate": 1.8714241408294838e-06,
+      "loss": 0.6842,
+      "step": 6313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.35800051442324,
+      "learning_rate": 1.8713767476563476e-06,
+      "loss": 0.6908,
+      "step": 6314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.413083751682385,
+      "learning_rate": 1.871329346350575e-06,
+      "loss": 0.6855,
+      "step": 6315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2444652078019927,
+      "learning_rate": 1.8712819369126089e-06,
+      "loss": 0.6927,
+      "step": 6316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0191343268061455,
+      "learning_rate": 1.8712345193428912e-06,
+      "loss": 0.6921,
+      "step": 6317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.0463152051433844,
+      "learning_rate": 1.8711870936418647e-06,
+      "loss": 0.6621,
+      "step": 6318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8377718287648469,
+      "learning_rate": 1.8711396598099722e-06,
+      "loss": 0.6888,
+      "step": 6319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1419936687700452,
+      "learning_rate": 1.871092217847656e-06,
+      "loss": 0.679,
+      "step": 6320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.802783431523835,
+      "learning_rate": 1.871044767755359e-06,
+      "loss": 0.6641,
+      "step": 6321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6187657467982977,
+      "learning_rate": 1.8709973095335245e-06,
+      "loss": 0.6654,
+      "step": 6322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.036502248466032,
+      "learning_rate": 1.8709498431825947e-06,
+      "loss": 0.6777,
+      "step": 6323
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.280633613306184,
+      "learning_rate": 1.8709023687030135e-06,
+      "loss": 0.6934,
+      "step": 6324
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7229858940460838,
+      "learning_rate": 1.8708548860952232e-06,
+      "loss": 0.6699,
+      "step": 6325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4842633096475866,
+      "learning_rate": 1.8708073953596671e-06,
+      "loss": 0.6719,
+      "step": 6326
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.597433447215373,
+      "learning_rate": 1.8707598964967888e-06,
+      "loss": 0.6771,
+      "step": 6327
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.672477327729963,
+      "learning_rate": 1.870712389507031e-06,
+      "loss": 0.6758,
+      "step": 6328
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.444736119344417,
+      "learning_rate": 1.870664874390838e-06,
+      "loss": 0.6973,
+      "step": 6329
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9711076045324156,
+      "learning_rate": 1.8706173511486522e-06,
+      "loss": 0.6842,
+      "step": 6330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.11304512187729,
+      "learning_rate": 1.8705698197809178e-06,
+      "loss": 0.7096,
+      "step": 6331
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.073661286570759,
+      "learning_rate": 1.8705222802880787e-06,
+      "loss": 0.6777,
+      "step": 6332
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6190454901551754,
+      "learning_rate": 1.8704747326705777e-06,
+      "loss": 0.6706,
+      "step": 6333
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6044836753483153,
+      "learning_rate": 1.8704271769288593e-06,
+      "loss": 0.6875,
+      "step": 6334
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.518053833179852,
+      "learning_rate": 1.870379613063367e-06,
+      "loss": 0.7057,
+      "step": 6335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.926092525193877,
+      "learning_rate": 1.8703320410745448e-06,
+      "loss": 0.679,
+      "step": 6336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2826763369435925,
+      "learning_rate": 1.8702844609628368e-06,
+      "loss": 0.6745,
+      "step": 6337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.814634235917494,
+      "learning_rate": 1.8702368727286868e-06,
+      "loss": 0.6868,
+      "step": 6338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9165641780489953,
+      "learning_rate": 1.870189276372539e-06,
+      "loss": 0.6966,
+      "step": 6339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.238290712501135,
+      "learning_rate": 1.870141671894838e-06,
+      "loss": 0.6764,
+      "step": 6340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6428603310169407,
+      "learning_rate": 1.8700940592960277e-06,
+      "loss": 0.6986,
+      "step": 6341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.624972013718136,
+      "learning_rate": 1.8700464385765527e-06,
+      "loss": 0.6608,
+      "step": 6342
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3511328512295284,
+      "learning_rate": 1.8699988097368574e-06,
+      "loss": 0.6862,
+      "step": 6343
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.716909166682827,
+      "learning_rate": 1.869951172777386e-06,
+      "loss": 0.6973,
+      "step": 6344
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5084294036224327,
+      "learning_rate": 1.8699035276985833e-06,
+      "loss": 0.694,
+      "step": 6345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3746068391683752,
+      "learning_rate": 1.8698558745008943e-06,
+      "loss": 0.6777,
+      "step": 6346
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45533324359123084,
+      "learning_rate": 1.8698082131847634e-06,
+      "loss": 0.6777,
+      "step": 6347
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.176578124564198,
+      "learning_rate": 1.8697605437506357e-06,
+      "loss": 0.6719,
+      "step": 6348
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4933949889730083,
+      "learning_rate": 1.869712866198956e-06,
+      "loss": 0.6777,
+      "step": 6349
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.192171252690165,
+      "learning_rate": 1.869665180530169e-06,
+      "loss": 0.6855,
+      "step": 6350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.739042826572142,
+      "learning_rate": 1.8696174867447196e-06,
+      "loss": 0.6823,
+      "step": 6351
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0008461999517513,
+      "learning_rate": 1.8695697848430537e-06,
+      "loss": 0.6777,
+      "step": 6352
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.853204439543093,
+      "learning_rate": 1.8695220748256162e-06,
+      "loss": 0.6452,
+      "step": 6353
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6458549594395147,
+      "learning_rate": 1.869474356692852e-06,
+      "loss": 0.6751,
+      "step": 6354
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.939562448553499,
+      "learning_rate": 1.869426630445207e-06,
+      "loss": 0.6979,
+      "step": 6355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.939998288446473,
+      "learning_rate": 1.869378896083126e-06,
+      "loss": 0.6953,
+      "step": 6356
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.11341337253842,
+      "learning_rate": 1.8693311536070552e-06,
+      "loss": 0.6999,
+      "step": 6357
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.345180576275583,
+      "learning_rate": 1.8692834030174399e-06,
+      "loss": 0.6895,
+      "step": 6358
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.2903788118546355,
+      "learning_rate": 1.8692356443147254e-06,
+      "loss": 0.6816,
+      "step": 6359
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5256284356701191,
+      "learning_rate": 1.8691878774993581e-06,
+      "loss": 0.6641,
+      "step": 6360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4217666691508826,
+      "learning_rate": 1.8691401025717833e-06,
+      "loss": 0.6882,
+      "step": 6361
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3108511023037845,
+      "learning_rate": 1.869092319532447e-06,
+      "loss": 0.6914,
+      "step": 6362
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.085935680086019,
+      "learning_rate": 1.8690445283817952e-06,
+      "loss": 0.694,
+      "step": 6363
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8371678039053996,
+      "learning_rate": 1.8689967291202741e-06,
+      "loss": 0.6712,
+      "step": 6364
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.2091557343865516,
+      "learning_rate": 1.8689489217483296e-06,
+      "loss": 0.696,
+      "step": 6365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.071197562054763,
+      "learning_rate": 1.8689011062664082e-06,
+      "loss": 0.6732,
+      "step": 6366
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.386498337379059,
+      "learning_rate": 1.8688532826749556e-06,
+      "loss": 0.6836,
+      "step": 6367
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5007268505865077,
+      "learning_rate": 1.8688054509744187e-06,
+      "loss": 0.666,
+      "step": 6368
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0857825663126883,
+      "learning_rate": 1.8687576111652437e-06,
+      "loss": 0.7096,
+      "step": 6369
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.629661668696504,
+      "learning_rate": 1.868709763247877e-06,
+      "loss": 0.6895,
+      "step": 6370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4718685446535111,
+      "learning_rate": 1.8686619072227652e-06,
+      "loss": 0.6699,
+      "step": 6371
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.222120425591797,
+      "learning_rate": 1.8686140430903551e-06,
+      "loss": 0.6914,
+      "step": 6372
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4730439920787831,
+      "learning_rate": 1.8685661708510933e-06,
+      "loss": 0.6712,
+      "step": 6373
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.59522467344479,
+      "learning_rate": 1.8685182905054267e-06,
+      "loss": 0.6868,
+      "step": 6374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6628048464614543,
+      "learning_rate": 1.868470402053802e-06,
+      "loss": 0.6634,
+      "step": 6375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.47551583851560647,
+      "learning_rate": 1.868422505496666e-06,
+      "loss": 0.6628,
+      "step": 6376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.585425464656077,
+      "learning_rate": 1.8683746008344663e-06,
+      "loss": 0.6549,
+      "step": 6377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.343447437874227,
+      "learning_rate": 1.8683266880676496e-06,
+      "loss": 0.6908,
+      "step": 6378
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9803498272616835,
+      "learning_rate": 1.8682787671966632e-06,
+      "loss": 0.6836,
+      "step": 6379
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.622632566826875,
+      "learning_rate": 1.868230838221954e-06,
+      "loss": 0.7122,
+      "step": 6380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.220735935778717,
+      "learning_rate": 1.8681829011439696e-06,
+      "loss": 0.6927,
+      "step": 6381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7474905086165569,
+      "learning_rate": 1.8681349559631576e-06,
+      "loss": 0.6953,
+      "step": 6382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.290641382147447,
+      "learning_rate": 1.868087002679965e-06,
+      "loss": 0.6823,
+      "step": 6383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9603038488069617,
+      "learning_rate": 1.86803904129484e-06,
+      "loss": 0.6764,
+      "step": 6384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1236613251160716,
+      "learning_rate": 1.86799107180823e-06,
+      "loss": 0.6738,
+      "step": 6385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4607759629164634,
+      "learning_rate": 1.8679430942205822e-06,
+      "loss": 0.6628,
+      "step": 6386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3933059638614957,
+      "learning_rate": 1.8678951085323447e-06,
+      "loss": 0.6738,
+      "step": 6387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6976756339955079,
+      "learning_rate": 1.8678471147439657e-06,
+      "loss": 0.6979,
+      "step": 6388
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.899196384277235,
+      "learning_rate": 1.8677991128558926e-06,
+      "loss": 0.6836,
+      "step": 6389
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.086466876819816,
+      "learning_rate": 1.8677511028685736e-06,
+      "loss": 0.6862,
+      "step": 6390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4778593222442257,
+      "learning_rate": 1.8677030847824572e-06,
+      "loss": 0.6836,
+      "step": 6391
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1351900178871137,
+      "learning_rate": 1.8676550585979906e-06,
+      "loss": 0.6947,
+      "step": 6392
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2729538016293718,
+      "learning_rate": 1.867607024315623e-06,
+      "loss": 0.6816,
+      "step": 6393
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5566849531719532,
+      "learning_rate": 1.867558981935802e-06,
+      "loss": 0.6628,
+      "step": 6394
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5837046206963903,
+      "learning_rate": 1.8675109314589767e-06,
+      "loss": 0.6927,
+      "step": 6395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.13061587514612,
+      "learning_rate": 1.867462872885595e-06,
+      "loss": 0.7057,
+      "step": 6396
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.546731594174343,
+      "learning_rate": 1.8674148062161055e-06,
+      "loss": 0.7031,
+      "step": 6397
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6916345295396396,
+      "learning_rate": 1.8673667314509572e-06,
+      "loss": 0.6628,
+      "step": 6398
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8922350003484478,
+      "learning_rate": 1.8673186485905982e-06,
+      "loss": 0.6706,
+      "step": 6399
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.815315041634688,
+      "learning_rate": 1.8672705576354775e-06,
+      "loss": 0.6771,
+      "step": 6400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1532363334440052,
+      "learning_rate": 1.8672224585860441e-06,
+      "loss": 0.6862,
+      "step": 6401
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.163913610023156,
+      "learning_rate": 1.8671743514427468e-06,
+      "loss": 0.6973,
+      "step": 6402
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4469873055913753,
+      "learning_rate": 1.8671262362060346e-06,
+      "loss": 0.6764,
+      "step": 6403
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3363082246572333,
+      "learning_rate": 1.8670781128763565e-06,
+      "loss": 0.7148,
+      "step": 6404
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.451811659617845,
+      "learning_rate": 1.8670299814541616e-06,
+      "loss": 0.6966,
+      "step": 6405
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6326615390110235,
+      "learning_rate": 1.8669818419398995e-06,
+      "loss": 0.6842,
+      "step": 6406
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5491910773721497,
+      "learning_rate": 1.866933694334019e-06,
+      "loss": 0.6986,
+      "step": 6407
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2817817659420436,
+      "learning_rate": 1.8668855386369698e-06,
+      "loss": 0.6895,
+      "step": 6408
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9241537719026507,
+      "learning_rate": 1.866837374849201e-06,
+      "loss": 0.6973,
+      "step": 6409
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.42150719431791617,
+      "learning_rate": 1.8667892029711623e-06,
+      "loss": 0.6758,
+      "step": 6410
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6865000810355846,
+      "learning_rate": 1.8667410230033035e-06,
+      "loss": 0.6908,
+      "step": 6411
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.613917957760698,
+      "learning_rate": 1.866692834946074e-06,
+      "loss": 0.6842,
+      "step": 6412
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.106876195386032,
+      "learning_rate": 1.8666446387999237e-06,
+      "loss": 0.6862,
+      "step": 6413
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.163918517097381,
+      "learning_rate": 1.8665964345653022e-06,
+      "loss": 0.6875,
+      "step": 6414
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6732776979781585,
+      "learning_rate": 1.8665482222426595e-06,
+      "loss": 0.681,
+      "step": 6415
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.227815610393997,
+      "learning_rate": 1.866500001832446e-06,
+      "loss": 0.679,
+      "step": 6416
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.130377400824001,
+      "learning_rate": 1.8664517733351111e-06,
+      "loss": 0.6758,
+      "step": 6417
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.962267077684355,
+      "learning_rate": 1.8664035367511054e-06,
+      "loss": 0.6895,
+      "step": 6418
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.700854151168571,
+      "learning_rate": 1.8663552920808786e-06,
+      "loss": 0.7025,
+      "step": 6419
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5220790211565816,
+      "learning_rate": 1.866307039324881e-06,
+      "loss": 0.6569,
+      "step": 6420
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.827481383965972,
+      "learning_rate": 1.8662587784835635e-06,
+      "loss": 0.681,
+      "step": 6421
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.010988992430959,
+      "learning_rate": 1.8662105095573762e-06,
+      "loss": 0.6758,
+      "step": 6422
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4419541804231049,
+      "learning_rate": 1.8661622325467695e-06,
+      "loss": 0.6908,
+      "step": 6423
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.40991730881231336,
+      "learning_rate": 1.8661139474521944e-06,
+      "loss": 0.6764,
+      "step": 6424
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0202695774040063,
+      "learning_rate": 1.8660656542741009e-06,
+      "loss": 0.6634,
+      "step": 6425
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2277915006733935,
+      "learning_rate": 1.86601735301294e-06,
+      "loss": 0.6706,
+      "step": 6426
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5577591927404695,
+      "learning_rate": 1.8659690436691629e-06,
+      "loss": 0.6927,
+      "step": 6427
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.724051879227557,
+      "learning_rate": 1.86592072624322e-06,
+      "loss": 0.6725,
+      "step": 6428
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.986872745246057,
+      "learning_rate": 1.8658724007355624e-06,
+      "loss": 0.6973,
+      "step": 6429
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.705646844256053,
+      "learning_rate": 1.865824067146641e-06,
+      "loss": 0.7057,
+      "step": 6430
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.971739332316349,
+      "learning_rate": 1.865775725476907e-06,
+      "loss": 0.6764,
+      "step": 6431
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.74685527328894,
+      "learning_rate": 1.8657273757268115e-06,
+      "loss": 0.6947,
+      "step": 6432
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.832057754792157,
+      "learning_rate": 1.865679017896806e-06,
+      "loss": 0.7012,
+      "step": 6433
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.244650090206115,
+      "learning_rate": 1.8656306519873417e-06,
+      "loss": 0.6836,
+      "step": 6434
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.034400209936846,
+      "learning_rate": 1.86558227799887e-06,
+      "loss": 0.6777,
+      "step": 6435
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8959592766348585,
+      "learning_rate": 1.8655338959318423e-06,
+      "loss": 0.6862,
+      "step": 6436
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4347423139688538,
+      "learning_rate": 1.8654855057867102e-06,
+      "loss": 0.6712,
+      "step": 6437
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6693169785789351,
+      "learning_rate": 1.8654371075639253e-06,
+      "loss": 0.6667,
+      "step": 6438
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.015907136284261,
+      "learning_rate": 1.8653887012639392e-06,
+      "loss": 0.6647,
+      "step": 6439
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.31228266790519,
+      "learning_rate": 1.865340286887204e-06,
+      "loss": 0.6816,
+      "step": 6440
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9887796508574636,
+      "learning_rate": 1.8652918644341714e-06,
+      "loss": 0.7064,
+      "step": 6441
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.3851588314110295,
+      "learning_rate": 1.8652434339052934e-06,
+      "loss": 0.6882,
+      "step": 6442
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.001408423505329,
+      "learning_rate": 1.8651949953010219e-06,
+      "loss": 0.6953,
+      "step": 6443
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7902391339197112,
+      "learning_rate": 1.8651465486218088e-06,
+      "loss": 0.6712,
+      "step": 6444
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5151305964195654,
+      "learning_rate": 1.8650980938681064e-06,
+      "loss": 0.6921,
+      "step": 6445
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.367413048591461,
+      "learning_rate": 1.8650496310403671e-06,
+      "loss": 0.6797,
+      "step": 6446
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4421736775200973,
+      "learning_rate": 1.8650011601390431e-06,
+      "loss": 0.6719,
+      "step": 6447
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7839325058440494,
+      "learning_rate": 1.8649526811645869e-06,
+      "loss": 0.6654,
+      "step": 6448
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5219396190323475,
+      "learning_rate": 1.8649041941174505e-06,
+      "loss": 0.6829,
+      "step": 6449
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.1192936053754634,
+      "learning_rate": 1.8648556989980872e-06,
+      "loss": 0.6634,
+      "step": 6450
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7103247990793193,
+      "learning_rate": 1.8648071958069489e-06,
+      "loss": 0.6947,
+      "step": 6451
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9160580319929172,
+      "learning_rate": 1.8647586845444886e-06,
+      "loss": 0.6745,
+      "step": 6452
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.908656525476134,
+      "learning_rate": 1.864710165211159e-06,
+      "loss": 0.6816,
+      "step": 6453
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.7894273485312375,
+      "learning_rate": 1.864661637807413e-06,
+      "loss": 0.6992,
+      "step": 6454
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.146249696515614,
+      "learning_rate": 1.8646131023337034e-06,
+      "loss": 0.6986,
+      "step": 6455
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.61330951811665,
+      "learning_rate": 1.8645645587904834e-06,
+      "loss": 0.6816,
+      "step": 6456
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8654281424910488,
+      "learning_rate": 1.864516007178206e-06,
+      "loss": 0.6569,
+      "step": 6457
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5311728292071833,
+      "learning_rate": 1.8644674474973238e-06,
+      "loss": 0.6732,
+      "step": 6458
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.414689679016519,
+      "learning_rate": 1.864418879748291e-06,
+      "loss": 0.679,
+      "step": 6459
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5603552981069959,
+      "learning_rate": 1.86437030393156e-06,
+      "loss": 0.6947,
+      "step": 6460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.140813313138478,
+      "learning_rate": 1.8643217200475845e-06,
+      "loss": 0.668,
+      "step": 6461
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4993217203469524,
+      "learning_rate": 1.8642731280968182e-06,
+      "loss": 0.6673,
+      "step": 6462
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.854226818901376,
+      "learning_rate": 1.8642245280797142e-06,
+      "loss": 0.6758,
+      "step": 6463
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9708753324087507,
+      "learning_rate": 1.8641759199967262e-06,
+      "loss": 0.6647,
+      "step": 6464
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9485293791262137,
+      "learning_rate": 1.864127303848308e-06,
+      "loss": 0.6751,
+      "step": 6465
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2598455562122965,
+      "learning_rate": 1.8640786796349134e-06,
+      "loss": 0.7018,
+      "step": 6466
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.17488338446789,
+      "learning_rate": 1.864030047356996e-06,
+      "loss": 0.6758,
+      "step": 6467
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.991952364457256,
+      "learning_rate": 1.8639814070150095e-06,
+      "loss": 0.6784,
+      "step": 6468
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.4476865449194545,
+      "learning_rate": 1.8639327586094083e-06,
+      "loss": 0.7012,
+      "step": 6469
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8088379183298321,
+      "learning_rate": 1.8638841021406462e-06,
+      "loss": 0.6816,
+      "step": 6470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.705259790953747,
+      "learning_rate": 1.8638354376091776e-06,
+      "loss": 0.6725,
+      "step": 6471
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6911709316770736,
+      "learning_rate": 1.8637867650154562e-06,
+      "loss": 0.6914,
+      "step": 6472
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1295306972520838,
+      "learning_rate": 1.8637380843599366e-06,
+      "loss": 0.666,
+      "step": 6473
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.067461421380177,
+      "learning_rate": 1.8636893956430731e-06,
+      "loss": 0.707,
+      "step": 6474
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9202734876682023,
+      "learning_rate": 1.86364069886532e-06,
+      "loss": 0.7116,
+      "step": 6475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.037852400352243,
+      "learning_rate": 1.8635919940271322e-06,
+      "loss": 0.6732,
+      "step": 6476
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.607658140279162,
+      "learning_rate": 1.8635432811289635e-06,
+      "loss": 0.6784,
+      "step": 6477
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8953560269646634,
+      "learning_rate": 1.8634945601712692e-06,
+      "loss": 0.6849,
+      "step": 6478
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.115265978631081,
+      "learning_rate": 1.8634458311545038e-06,
+      "loss": 0.6973,
+      "step": 6479
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7233522873435755,
+      "learning_rate": 1.863397094079122e-06,
+      "loss": 0.6777,
+      "step": 6480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.865075858277898,
+      "learning_rate": 1.8633483489455792e-06,
+      "loss": 0.707,
+      "step": 6481
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4416650886711402,
+      "learning_rate": 1.8632995957543293e-06,
+      "loss": 0.6797,
+      "step": 6482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.521443264094885,
+      "learning_rate": 1.8632508345058283e-06,
+      "loss": 0.6927,
+      "step": 6483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6147457689246825,
+      "learning_rate": 1.863202065200531e-06,
+      "loss": 0.6582,
+      "step": 6484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8167173762332465,
+      "learning_rate": 1.8631532878388922e-06,
+      "loss": 0.6686,
+      "step": 6485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3013890947521987,
+      "learning_rate": 1.8631045024213676e-06,
+      "loss": 0.679,
+      "step": 6486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8329728766728385,
+      "learning_rate": 1.8630557089484125e-06,
+      "loss": 0.6947,
+      "step": 6487
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6159446220399283,
+      "learning_rate": 1.8630069074204817e-06,
+      "loss": 0.6738,
+      "step": 6488
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.864905539828197,
+      "learning_rate": 1.8629580978380315e-06,
+      "loss": 0.694,
+      "step": 6489
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.449908623039939,
+      "learning_rate": 1.862909280201517e-06,
+      "loss": 0.7103,
+      "step": 6490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.685455837716848,
+      "learning_rate": 1.8628604545113937e-06,
+      "loss": 0.6771,
+      "step": 6491
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.186574083754597,
+      "learning_rate": 1.8628116207681174e-06,
+      "loss": 0.6888,
+      "step": 6492
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.780252998917692,
+      "learning_rate": 1.8627627789721442e-06,
+      "loss": 0.6641,
+      "step": 6493
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3631368798719343,
+      "learning_rate": 1.8627139291239296e-06,
+      "loss": 0.696,
+      "step": 6494
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.9940566821054434,
+      "learning_rate": 1.8626650712239295e-06,
+      "loss": 0.6699,
+      "step": 6495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.258161102214819,
+      "learning_rate": 1.8626162052726002e-06,
+      "loss": 0.6803,
+      "step": 6496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.627750563137641,
+      "learning_rate": 1.8625673312703976e-06,
+      "loss": 0.6816,
+      "step": 6497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8121729641142985,
+      "learning_rate": 1.8625184492177777e-06,
+      "loss": 0.6875,
+      "step": 6498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.556215429904746,
+      "learning_rate": 1.8624695591151967e-06,
+      "loss": 0.6816,
+      "step": 6499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4797513590850158,
+      "learning_rate": 1.862420660963111e-06,
+      "loss": 0.6738,
+      "step": 6500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.150940912526497,
+      "learning_rate": 1.8623717547619773e-06,
+      "loss": 0.6862,
+      "step": 6501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.529064270111706,
+      "learning_rate": 1.8623228405122514e-06,
+      "loss": 0.6849,
+      "step": 6502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.915102244967386,
+      "learning_rate": 1.8622739182143908e-06,
+      "loss": 0.6901,
+      "step": 6503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.391118047256188,
+      "learning_rate": 1.8622249878688509e-06,
+      "loss": 0.6934,
+      "step": 6504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.695184602114022,
+      "learning_rate": 1.862176049476089e-06,
+      "loss": 0.6797,
+      "step": 6505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4096515255291737,
+      "learning_rate": 1.862127103036562e-06,
+      "loss": 0.6758,
+      "step": 6506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9122170326725634,
+      "learning_rate": 1.8620781485507263e-06,
+      "loss": 0.6797,
+      "step": 6507
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5943104845108107,
+      "learning_rate": 1.8620291860190389e-06,
+      "loss": 0.7031,
+      "step": 6508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7666537087704017,
+      "learning_rate": 1.861980215441957e-06,
+      "loss": 0.668,
+      "step": 6509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.004670085357831,
+      "learning_rate": 1.8619312368199376e-06,
+      "loss": 0.6868,
+      "step": 6510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2067031399889196,
+      "learning_rate": 1.861882250153438e-06,
+      "loss": 0.6836,
+      "step": 6511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.927663312147311,
+      "learning_rate": 1.8618332554429145e-06,
+      "loss": 0.6823,
+      "step": 6512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0521064042255817,
+      "learning_rate": 1.8617842526888254e-06,
+      "loss": 0.653,
+      "step": 6513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8270247258073082,
+      "learning_rate": 1.8617352418916277e-06,
+      "loss": 0.6771,
+      "step": 6514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.625115786405678,
+      "learning_rate": 1.8616862230517785e-06,
+      "loss": 0.7031,
+      "step": 6515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.340144293689997,
+      "learning_rate": 1.861637196169736e-06,
+      "loss": 0.6953,
+      "step": 6516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5785219610972918,
+      "learning_rate": 1.861588161245957e-06,
+      "loss": 0.6764,
+      "step": 6517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4856632784365987,
+      "learning_rate": 1.8615391182808997e-06,
+      "loss": 0.6797,
+      "step": 6518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7986161979252567,
+      "learning_rate": 1.8614900672750214e-06,
+      "loss": 0.6999,
+      "step": 6519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.404199554947715,
+      "learning_rate": 1.8614410082287804e-06,
+      "loss": 0.6816,
+      "step": 6520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7750257619326295,
+      "learning_rate": 1.861391941142634e-06,
+      "loss": 0.6745,
+      "step": 6521
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6574910260217555,
+      "learning_rate": 1.8613428660170408e-06,
+      "loss": 0.6875,
+      "step": 6522
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8376295139625045,
+      "learning_rate": 1.861293782852458e-06,
+      "loss": 0.6803,
+      "step": 6523
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3346456286257644,
+      "learning_rate": 1.8612446916493442e-06,
+      "loss": 0.6901,
+      "step": 6524
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5279595751096138,
+      "learning_rate": 1.861195592408158e-06,
+      "loss": 0.6797,
+      "step": 6525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4389594953042424,
+      "learning_rate": 1.8611464851293568e-06,
+      "loss": 0.6673,
+      "step": 6526
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.48641433781952215,
+      "learning_rate": 1.8610973698133993e-06,
+      "loss": 0.6693,
+      "step": 6527
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6532099510290372,
+      "learning_rate": 1.861048246460744e-06,
+      "loss": 0.6934,
+      "step": 6528
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.331790457962405,
+      "learning_rate": 1.8609991150718493e-06,
+      "loss": 0.6771,
+      "step": 6529
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.0979788911287165,
+      "learning_rate": 1.8609499756471736e-06,
+      "loss": 0.7122,
+      "step": 6530
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.499578895174726,
+      "learning_rate": 1.8609008281871757e-06,
+      "loss": 0.6647,
+      "step": 6531
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5648775651618354,
+      "learning_rate": 1.8608516726923142e-06,
+      "loss": 0.6862,
+      "step": 6532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5076083517645148,
+      "learning_rate": 1.860802509163048e-06,
+      "loss": 0.6745,
+      "step": 6533
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.360200260004777,
+      "learning_rate": 1.860753337599836e-06,
+      "loss": 0.7103,
+      "step": 6534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7789808795443838,
+      "learning_rate": 1.8607041580031366e-06,
+      "loss": 0.7012,
+      "step": 6535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.021089349907555,
+      "learning_rate": 1.8606549703734093e-06,
+      "loss": 0.6947,
+      "step": 6536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1340446019616115,
+      "learning_rate": 1.860605774711113e-06,
+      "loss": 0.6836,
+      "step": 6537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.069607612008416,
+      "learning_rate": 1.860556571016707e-06,
+      "loss": 0.6745,
+      "step": 6538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6095251627356104,
+      "learning_rate": 1.8605073592906506e-06,
+      "loss": 0.6771,
+      "step": 6539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.155409176034626,
+      "learning_rate": 1.8604581395334028e-06,
+      "loss": 0.7018,
+      "step": 6540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5795875776156305,
+      "learning_rate": 1.860408911745423e-06,
+      "loss": 0.6921,
+      "step": 6541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.735267800231069,
+      "learning_rate": 1.8603596759271707e-06,
+      "loss": 0.6732,
+      "step": 6542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5149768344540373,
+      "learning_rate": 1.8603104320791054e-06,
+      "loss": 0.6816,
+      "step": 6543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.2079978174206945,
+      "learning_rate": 1.860261180201687e-06,
+      "loss": 0.6927,
+      "step": 6544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8672946484224378,
+      "learning_rate": 1.8602119202953748e-06,
+      "loss": 0.6888,
+      "step": 6545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.906850876074825,
+      "learning_rate": 1.8601626523606286e-06,
+      "loss": 0.6953,
+      "step": 6546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.289961318004927,
+      "learning_rate": 1.8601133763979084e-06,
+      "loss": 0.6842,
+      "step": 6547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5100168389694729,
+      "learning_rate": 1.8600640924076737e-06,
+      "loss": 0.6751,
+      "step": 6548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.480536854988924,
+      "learning_rate": 1.8600148003903853e-06,
+      "loss": 0.6849,
+      "step": 6549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.103792370758966,
+      "learning_rate": 1.8599655003465025e-06,
+      "loss": 0.6934,
+      "step": 6550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.649421897395156,
+      "learning_rate": 1.8599161922764854e-06,
+      "loss": 0.6901,
+      "step": 6551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6143573991309292,
+      "learning_rate": 1.8598668761807948e-06,
+      "loss": 0.6999,
+      "step": 6552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5480858665790658,
+      "learning_rate": 1.8598175520598902e-06,
+      "loss": 0.6979,
+      "step": 6553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5781258410240753,
+      "learning_rate": 1.859768219914233e-06,
+      "loss": 0.6868,
+      "step": 6554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.637923115576538,
+      "learning_rate": 1.8597188797442823e-06,
+      "loss": 0.6875,
+      "step": 6555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.668758966057671,
+      "learning_rate": 1.8596695315504993e-06,
+      "loss": 0.6921,
+      "step": 6556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.5034688363064665,
+      "learning_rate": 1.8596201753333447e-06,
+      "loss": 0.6745,
+      "step": 6557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2528078132914384,
+      "learning_rate": 1.8595708110932793e-06,
+      "loss": 0.6901,
+      "step": 6558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.014753801975447,
+      "learning_rate": 1.8595214388307631e-06,
+      "loss": 0.6823,
+      "step": 6559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.562989455909385,
+      "learning_rate": 1.8594720585462573e-06,
+      "loss": 0.6706,
+      "step": 6560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.357388145873125,
+      "learning_rate": 1.8594226702402232e-06,
+      "loss": 0.6901,
+      "step": 6561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5372940359902505,
+      "learning_rate": 1.8593732739131206e-06,
+      "loss": 0.6882,
+      "step": 6562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.480352146935645,
+      "learning_rate": 1.8593238695654119e-06,
+      "loss": 0.6784,
+      "step": 6563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8273814816110705,
+      "learning_rate": 1.8592744571975572e-06,
+      "loss": 0.679,
+      "step": 6564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.046103293501954,
+      "learning_rate": 1.859225036810018e-06,
+      "loss": 0.6934,
+      "step": 6565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.631271749268336,
+      "learning_rate": 1.8591756084032556e-06,
+      "loss": 0.6882,
+      "step": 6566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7126056735689793,
+      "learning_rate": 1.8591261719777313e-06,
+      "loss": 0.668,
+      "step": 6567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5869595094254465,
+      "learning_rate": 1.8590767275339066e-06,
+      "loss": 0.6836,
+      "step": 6568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.288063662124137,
+      "learning_rate": 1.8590272750722427e-06,
+      "loss": 0.6979,
+      "step": 6569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2096128256718401,
+      "learning_rate": 1.8589778145932012e-06,
+      "loss": 0.6868,
+      "step": 6570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.791514354104738,
+      "learning_rate": 1.8589283460972438e-06,
+      "loss": 0.6914,
+      "step": 6571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4542403515596616,
+      "learning_rate": 1.8588788695848322e-06,
+      "loss": 0.6719,
+      "step": 6572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.11658147359153,
+      "learning_rate": 1.8588293850564281e-06,
+      "loss": 0.681,
+      "step": 6573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.983505906396281,
+      "learning_rate": 1.8587798925124935e-06,
+      "loss": 0.6979,
+      "step": 6574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5823974445573312,
+      "learning_rate": 1.8587303919534903e-06,
+      "loss": 0.6823,
+      "step": 6575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5278051171261926,
+      "learning_rate": 1.8586808833798804e-06,
+      "loss": 0.696,
+      "step": 6576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.402897681737694,
+      "learning_rate": 1.8586313667921257e-06,
+      "loss": 0.6706,
+      "step": 6577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.244199077412784,
+      "learning_rate": 1.8585818421906886e-06,
+      "loss": 0.7025,
+      "step": 6578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9012311891545943,
+      "learning_rate": 1.8585323095760312e-06,
+      "loss": 0.6719,
+      "step": 6579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.3964444735998764,
+      "learning_rate": 1.8584827689486157e-06,
+      "loss": 0.6842,
+      "step": 6580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5057728026577213,
+      "learning_rate": 1.8584332203089047e-06,
+      "loss": 0.6725,
+      "step": 6581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6211927793668222,
+      "learning_rate": 1.8583836636573609e-06,
+      "loss": 0.6797,
+      "step": 6582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.341967456958294,
+      "learning_rate": 1.858334098994446e-06,
+      "loss": 0.679,
+      "step": 6583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2553617348124797,
+      "learning_rate": 1.8582845263206232e-06,
+      "loss": 0.7012,
+      "step": 6584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9644592782751504,
+      "learning_rate": 1.8582349456363552e-06,
+      "loss": 0.6895,
+      "step": 6585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.560210909026044,
+      "learning_rate": 1.8581853569421042e-06,
+      "loss": 0.6914,
+      "step": 6586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9785200542383454,
+      "learning_rate": 1.8581357602383334e-06,
+      "loss": 0.7096,
+      "step": 6587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4615147990402858,
+      "learning_rate": 1.858086155525506e-06,
+      "loss": 0.6875,
+      "step": 6588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.0759530758384805,
+      "learning_rate": 1.8580365428040846e-06,
+      "loss": 0.6784,
+      "step": 6589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.868382791996868,
+      "learning_rate": 1.857986922074532e-06,
+      "loss": 0.6849,
+      "step": 6590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9550530177558854,
+      "learning_rate": 1.8579372933373115e-06,
+      "loss": 0.6901,
+      "step": 6591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.198385405419071,
+      "learning_rate": 1.8578876565928866e-06,
+      "loss": 0.6927,
+      "step": 6592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6939826871007115,
+      "learning_rate": 1.8578380118417203e-06,
+      "loss": 0.6615,
+      "step": 6593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1640903531856512,
+      "learning_rate": 1.8577883590842758e-06,
+      "loss": 0.6784,
+      "step": 6594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5112385047429189,
+      "learning_rate": 1.8577386983210168e-06,
+      "loss": 0.6738,
+      "step": 6595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.474669637103727,
+      "learning_rate": 1.8576890295524069e-06,
+      "loss": 0.6816,
+      "step": 6596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.673180834185709,
+      "learning_rate": 1.857639352778909e-06,
+      "loss": 0.6615,
+      "step": 6597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.865655573141371,
+      "learning_rate": 1.8575896680009877e-06,
+      "loss": 0.6875,
+      "step": 6598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0894246853448633,
+      "learning_rate": 1.8575399752191059e-06,
+      "loss": 0.6816,
+      "step": 6599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.242867241386685,
+      "learning_rate": 1.8574902744337275e-06,
+      "loss": 0.6562,
+      "step": 6600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.544371352883825,
+      "learning_rate": 1.8574405656453168e-06,
+      "loss": 0.6862,
+      "step": 6601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2615602583967744,
+      "learning_rate": 1.8573908488543376e-06,
+      "loss": 0.6973,
+      "step": 6602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0667185621977655,
+      "learning_rate": 1.8573411240612533e-06,
+      "loss": 0.6641,
+      "step": 6603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.054851608120882,
+      "learning_rate": 1.8572913912665289e-06,
+      "loss": 0.6686,
+      "step": 6604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5240045087569118,
+      "learning_rate": 1.8572416504706279e-06,
+      "loss": 0.6829,
+      "step": 6605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.840106068879677,
+      "learning_rate": 1.8571919016740148e-06,
+      "loss": 0.6947,
+      "step": 6606
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.49720404256001,
+      "learning_rate": 1.8571421448771538e-06,
+      "loss": 0.6816,
+      "step": 6607
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5263428055195744,
+      "learning_rate": 1.8570923800805097e-06,
+      "loss": 0.6777,
+      "step": 6608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7224609062243235,
+      "learning_rate": 1.8570426072845464e-06,
+      "loss": 0.6888,
+      "step": 6609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2224810402466941,
+      "learning_rate": 1.8569928264897286e-06,
+      "loss": 0.6562,
+      "step": 6610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9308882780157302,
+      "learning_rate": 1.856943037696521e-06,
+      "loss": 0.6699,
+      "step": 6611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8639326540476732,
+      "learning_rate": 1.8568932409053884e-06,
+      "loss": 0.6706,
+      "step": 6612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5615012683678361,
+      "learning_rate": 1.856843436116795e-06,
+      "loss": 0.6758,
+      "step": 6613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9408903351538997,
+      "learning_rate": 1.8567936233312062e-06,
+      "loss": 0.6595,
+      "step": 6614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.640154243178397,
+      "learning_rate": 1.856743802549087e-06,
+      "loss": 0.6966,
+      "step": 6615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8371276934852536,
+      "learning_rate": 1.856693973770902e-06,
+      "loss": 0.6777,
+      "step": 6616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6062073609712393,
+      "learning_rate": 1.8566441369971163e-06,
+      "loss": 0.6751,
+      "step": 6617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.195840305925646,
+      "learning_rate": 1.856594292228195e-06,
+      "loss": 0.6549,
+      "step": 6618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.449298558374584,
+      "learning_rate": 1.8565444394646038e-06,
+      "loss": 0.7051,
+      "step": 6619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6344704585174077,
+      "learning_rate": 1.8564945787068074e-06,
+      "loss": 0.651,
+      "step": 6620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9550805344990505,
+      "learning_rate": 1.856444709955271e-06,
+      "loss": 0.7083,
+      "step": 6621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6193653291611428,
+      "learning_rate": 1.856394833210461e-06,
+      "loss": 0.6641,
+      "step": 6622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.418866017271391,
+      "learning_rate": 1.8563449484728419e-06,
+      "loss": 0.6882,
+      "step": 6623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.067953102070022,
+      "learning_rate": 1.8562950557428797e-06,
+      "loss": 0.6992,
+      "step": 6624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.991946527817631,
+      "learning_rate": 1.85624515502104e-06,
+      "loss": 0.6732,
+      "step": 6625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.332281733353921,
+      "learning_rate": 1.8561952463077887e-06,
+      "loss": 0.6745,
+      "step": 6626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5982639518140968,
+      "learning_rate": 1.8561453296035913e-06,
+      "loss": 0.6699,
+      "step": 6627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.987178369159516,
+      "learning_rate": 1.856095404908914e-06,
+      "loss": 0.6777,
+      "step": 6628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9306820417671373,
+      "learning_rate": 1.8560454722242224e-06,
+      "loss": 0.6921,
+      "step": 6629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7133499875243237,
+      "learning_rate": 1.8559955315499826e-06,
+      "loss": 0.679,
+      "step": 6630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0009228360291886,
+      "learning_rate": 1.8559455828866612e-06,
+      "loss": 0.6842,
+      "step": 6631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.400798653350712,
+      "learning_rate": 1.8558956262347235e-06,
+      "loss": 0.6882,
+      "step": 6632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.036996349281076,
+      "learning_rate": 1.8558456615946366e-06,
+      "loss": 0.6771,
+      "step": 6633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5424438364928837,
+      "learning_rate": 1.8557956889668664e-06,
+      "loss": 0.6569,
+      "step": 6634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.424413127840211,
+      "learning_rate": 1.8557457083518792e-06,
+      "loss": 0.6901,
+      "step": 6635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3490981281537682,
+      "learning_rate": 1.855695719750142e-06,
+      "loss": 0.6745,
+      "step": 6636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.795498238583546,
+      "learning_rate": 1.8556457231621202e-06,
+      "loss": 0.6947,
+      "step": 6637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.040450839826729,
+      "learning_rate": 1.855595718588282e-06,
+      "loss": 0.6562,
+      "step": 6638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3663001290175596,
+      "learning_rate": 1.8555457060290929e-06,
+      "loss": 0.6914,
+      "step": 6639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5086286399934044,
+      "learning_rate": 1.8554956854850201e-06,
+      "loss": 0.7109,
+      "step": 6640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.116064733597686,
+      "learning_rate": 1.8554456569565305e-06,
+      "loss": 0.6875,
+      "step": 6641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.03916978610033,
+      "learning_rate": 1.8553956204440907e-06,
+      "loss": 0.6628,
+      "step": 6642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9525227747864292,
+      "learning_rate": 1.8553455759481684e-06,
+      "loss": 0.6816,
+      "step": 6643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.873851551920292,
+      "learning_rate": 1.8552955234692298e-06,
+      "loss": 0.7025,
+      "step": 6644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4889355746045236,
+      "learning_rate": 1.8552454630077426e-06,
+      "loss": 0.7201,
+      "step": 6645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1776546745445406,
+      "learning_rate": 1.8551953945641738e-06,
+      "loss": 0.6829,
+      "step": 6646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4942071925553457,
+      "learning_rate": 1.8551453181389906e-06,
+      "loss": 0.6862,
+      "step": 6647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.451260514137733,
+      "learning_rate": 1.8550952337326606e-06,
+      "loss": 0.668,
+      "step": 6648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7699360379328808,
+      "learning_rate": 1.8550451413456517e-06,
+      "loss": 0.6803,
+      "step": 6649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5106385091673709,
+      "learning_rate": 1.85499504097843e-06,
+      "loss": 0.6595,
+      "step": 6650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3858164945546978,
+      "learning_rate": 1.8549449326314645e-06,
+      "loss": 0.6836,
+      "step": 6651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.44224441975790196,
+      "learning_rate": 1.8548948163052222e-06,
+      "loss": 0.6953,
+      "step": 6652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.801687938947746,
+      "learning_rate": 1.8548446920001712e-06,
+      "loss": 0.6908,
+      "step": 6653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6913857077573615,
+      "learning_rate": 1.8547945597167786e-06,
+      "loss": 0.6725,
+      "step": 6654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7086335553375545,
+      "learning_rate": 1.8547444194555132e-06,
+      "loss": 0.6908,
+      "step": 6655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.037105814863272,
+      "learning_rate": 1.8546942712168424e-06,
+      "loss": 0.6849,
+      "step": 6656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.488230493530531,
+      "learning_rate": 1.8546441150012346e-06,
+      "loss": 0.679,
+      "step": 6657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4618132980791901,
+      "learning_rate": 1.8545939508091575e-06,
+      "loss": 0.6895,
+      "step": 6658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7537446658013136,
+      "learning_rate": 1.8545437786410794e-06,
+      "loss": 0.6712,
+      "step": 6659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2739831478639936,
+      "learning_rate": 1.8544935984974687e-06,
+      "loss": 0.6836,
+      "step": 6660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.142022932527656,
+      "learning_rate": 1.8544434103787939e-06,
+      "loss": 0.7031,
+      "step": 6661
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.223524017626008,
+      "learning_rate": 1.8543932142855227e-06,
+      "loss": 0.6849,
+      "step": 6662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7713765894803797,
+      "learning_rate": 1.8543430102181245e-06,
+      "loss": 0.7012,
+      "step": 6663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.1423801923504797,
+      "learning_rate": 1.854292798177067e-06,
+      "loss": 0.6771,
+      "step": 6664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7029727941254367,
+      "learning_rate": 1.8542425781628197e-06,
+      "loss": 0.6829,
+      "step": 6665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2024348599348444,
+      "learning_rate": 1.8541923501758508e-06,
+      "loss": 0.6751,
+      "step": 6666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9945017284871528,
+      "learning_rate": 1.8541421142166291e-06,
+      "loss": 0.6706,
+      "step": 6667
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5320563327890223,
+      "learning_rate": 1.8540918702856234e-06,
+      "loss": 0.6764,
+      "step": 6668
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.878222847483,
+      "learning_rate": 1.8540416183833027e-06,
+      "loss": 0.6999,
+      "step": 6669
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7383066578296853,
+      "learning_rate": 1.8539913585101363e-06,
+      "loss": 0.6725,
+      "step": 6670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5661593154181657,
+      "learning_rate": 1.853941090666593e-06,
+      "loss": 0.6647,
+      "step": 6671
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.40965815823185214,
+      "learning_rate": 1.8538908148531418e-06,
+      "loss": 0.6751,
+      "step": 6672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46381971202293953,
+      "learning_rate": 1.853840531070252e-06,
+      "loss": 0.6725,
+      "step": 6673
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.4992269712789925,
+      "learning_rate": 1.8537902393183935e-06,
+      "loss": 0.6712,
+      "step": 6674
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9171604285525736,
+      "learning_rate": 1.853739939598035e-06,
+      "loss": 0.681,
+      "step": 6675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4751883412307195,
+      "learning_rate": 1.853689631909646e-06,
+      "loss": 0.6602,
+      "step": 6676
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8786727352684027,
+      "learning_rate": 1.8536393162536963e-06,
+      "loss": 0.6921,
+      "step": 6677
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6142197022294877,
+      "learning_rate": 1.8535889926306552e-06,
+      "loss": 0.6901,
+      "step": 6678
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8995436956808702,
+      "learning_rate": 1.8535386610409925e-06,
+      "loss": 0.6855,
+      "step": 6679
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6129631250119167,
+      "learning_rate": 1.8534883214851781e-06,
+      "loss": 0.707,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.220523500475032,
+      "learning_rate": 1.8534379739636818e-06,
+      "loss": 0.6797,
+      "step": 6681
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.67938625309235,
+      "learning_rate": 1.8533876184769735e-06,
+      "loss": 0.6882,
+      "step": 6682
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.274947275043748,
+      "learning_rate": 1.8533372550255228e-06,
+      "loss": 0.709,
+      "step": 6683
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48727245797344587,
+      "learning_rate": 1.8532868836098e-06,
+      "loss": 0.6901,
+      "step": 6684
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7634126293764205,
+      "learning_rate": 1.8532365042302756e-06,
+      "loss": 0.6751,
+      "step": 6685
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.164177434757322,
+      "learning_rate": 1.8531861168874192e-06,
+      "loss": 0.6947,
+      "step": 6686
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.244135252581009,
+      "learning_rate": 1.8531357215817015e-06,
+      "loss": 0.6777,
+      "step": 6687
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.48730153262053794,
+      "learning_rate": 1.8530853183135925e-06,
+      "loss": 0.6764,
+      "step": 6688
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.64185547126021,
+      "learning_rate": 1.853034907083563e-06,
+      "loss": 0.6608,
+      "step": 6689
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7217067731860382,
+      "learning_rate": 1.852984487892083e-06,
+      "loss": 0.6634,
+      "step": 6690
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7133663960426704,
+      "learning_rate": 1.8529340607396236e-06,
+      "loss": 0.6934,
+      "step": 6691
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.864273730652383,
+      "learning_rate": 1.8528836256266552e-06,
+      "loss": 0.6797,
+      "step": 6692
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2994075298404586,
+      "learning_rate": 1.852833182553648e-06,
+      "loss": 0.6862,
+      "step": 6693
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5191310130089788,
+      "learning_rate": 1.8527827315210737e-06,
+      "loss": 0.6842,
+      "step": 6694
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.973775898472121,
+      "learning_rate": 1.8527322725294028e-06,
+      "loss": 0.6686,
+      "step": 6695
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6388192113420652,
+      "learning_rate": 1.852681805579106e-06,
+      "loss": 0.6745,
+      "step": 6696
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7884944810996695,
+      "learning_rate": 1.8526313306706544e-06,
+      "loss": 0.6855,
+      "step": 6697
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.716976365409744,
+      "learning_rate": 1.8525808478045194e-06,
+      "loss": 0.6751,
+      "step": 6698
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1371545334006539,
+      "learning_rate": 1.8525303569811719e-06,
+      "loss": 0.6771,
+      "step": 6699
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.020738827440076,
+      "learning_rate": 1.8524798582010831e-06,
+      "loss": 0.6868,
+      "step": 6700
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.05472135232009,
+      "learning_rate": 1.8524293514647247e-06,
+      "loss": 0.6738,
+      "step": 6701
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3671468363030728,
+      "learning_rate": 1.8523788367725675e-06,
+      "loss": 0.6842,
+      "step": 6702
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.7057062988042238,
+      "learning_rate": 1.852328314125083e-06,
+      "loss": 0.6706,
+      "step": 6703
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.718673217343224,
+      "learning_rate": 1.8522777835227434e-06,
+      "loss": 0.6934,
+      "step": 6704
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.637707467072178,
+      "learning_rate": 1.8522272449660199e-06,
+      "loss": 0.6803,
+      "step": 6705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.816426554616545,
+      "learning_rate": 1.852176698455384e-06,
+      "loss": 0.6556,
+      "step": 6706
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6390331347292056,
+      "learning_rate": 1.8521261439913075e-06,
+      "loss": 0.6738,
+      "step": 6707
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9787469629136438,
+      "learning_rate": 1.8520755815742626e-06,
+      "loss": 0.6797,
+      "step": 6708
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9478654962043869,
+      "learning_rate": 1.8520250112047209e-06,
+      "loss": 0.6699,
+      "step": 6709
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3241198080804524,
+      "learning_rate": 1.851974432883154e-06,
+      "loss": 0.6901,
+      "step": 6710
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.876353607340394,
+      "learning_rate": 1.851923846610035e-06,
+      "loss": 0.6602,
+      "step": 6711
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.277558455076857,
+      "learning_rate": 1.851873252385835e-06,
+      "loss": 0.6719,
+      "step": 6712
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7360719693991303,
+      "learning_rate": 1.8518226502110266e-06,
+      "loss": 0.6654,
+      "step": 6713
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.511294108832639,
+      "learning_rate": 1.851772040086082e-06,
+      "loss": 0.6855,
+      "step": 6714
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.002974574498549,
+      "learning_rate": 1.851721422011474e-06,
+      "loss": 0.6667,
+      "step": 6715
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3600523646879643,
+      "learning_rate": 1.8516707959876743e-06,
+      "loss": 0.681,
+      "step": 6716
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8120373159762168,
+      "learning_rate": 1.8516201620151558e-06,
+      "loss": 0.6536,
+      "step": 6717
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6770606664083743,
+      "learning_rate": 1.851569520094391e-06,
+      "loss": 0.6797,
+      "step": 6718
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5771173178695845,
+      "learning_rate": 1.8515188702258529e-06,
+      "loss": 0.6576,
+      "step": 6719
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9903552812354621,
+      "learning_rate": 1.8514682124100136e-06,
+      "loss": 0.7025,
+      "step": 6720
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.377402397193773,
+      "learning_rate": 1.851417546647346e-06,
+      "loss": 0.681,
+      "step": 6721
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.150047670111178,
+      "learning_rate": 1.8513668729383235e-06,
+      "loss": 0.6719,
+      "step": 6722
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.246290968722913,
+      "learning_rate": 1.8513161912834186e-06,
+      "loss": 0.6673,
+      "step": 6723
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4138079748270416,
+      "learning_rate": 1.8512655016831042e-06,
+      "loss": 0.6732,
+      "step": 6724
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.343296935360941,
+      "learning_rate": 1.8512148041378536e-06,
+      "loss": 0.6771,
+      "step": 6725
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.769514021095052,
+      "learning_rate": 1.8511640986481402e-06,
+      "loss": 0.6647,
+      "step": 6726
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4894220056178296,
+      "learning_rate": 1.8511133852144369e-06,
+      "loss": 0.6973,
+      "step": 6727
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.435300556584951,
+      "learning_rate": 1.851062663837217e-06,
+      "loss": 0.6654,
+      "step": 6728
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0519547394250592,
+      "learning_rate": 1.8510119345169542e-06,
+      "loss": 0.6758,
+      "step": 6729
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8635415782863484,
+      "learning_rate": 1.8509611972541214e-06,
+      "loss": 0.6849,
+      "step": 6730
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.752279199469599,
+      "learning_rate": 1.8509104520491929e-06,
+      "loss": 0.6927,
+      "step": 6731
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.598727998330249,
+      "learning_rate": 1.8508596989026418e-06,
+      "loss": 0.7018,
+      "step": 6732
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9670847674620684,
+      "learning_rate": 1.8508089378149417e-06,
+      "loss": 0.6836,
+      "step": 6733
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8969656455640703,
+      "learning_rate": 1.8507581687865665e-06,
+      "loss": 0.6706,
+      "step": 6734
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7653807459713625,
+      "learning_rate": 1.8507073918179902e-06,
+      "loss": 0.6595,
+      "step": 6735
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.901754647266317,
+      "learning_rate": 1.8506566069096868e-06,
+      "loss": 0.6517,
+      "step": 6736
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1623412040532268,
+      "learning_rate": 1.8506058140621297e-06,
+      "loss": 0.666,
+      "step": 6737
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.226294731660979,
+      "learning_rate": 1.8505550132757934e-06,
+      "loss": 0.6921,
+      "step": 6738
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5805963236388242,
+      "learning_rate": 1.8505042045511522e-06,
+      "loss": 0.6621,
+      "step": 6739
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5986910524191317,
+      "learning_rate": 1.8504533878886798e-06,
+      "loss": 0.6816,
+      "step": 6740
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7388439311335336,
+      "learning_rate": 1.8504025632888507e-06,
+      "loss": 0.6634,
+      "step": 6741
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.283468511101873,
+      "learning_rate": 1.8503517307521396e-06,
+      "loss": 0.6784,
+      "step": 6742
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7981056347414509,
+      "learning_rate": 1.85030089027902e-06,
+      "loss": 0.6673,
+      "step": 6743
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.8614589987564436,
+      "learning_rate": 1.8502500418699675e-06,
+      "loss": 0.6973,
+      "step": 6744
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.4928893732392075,
+      "learning_rate": 1.850199185525456e-06,
+      "loss": 0.6803,
+      "step": 6745
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 5.076279403684931,
+      "learning_rate": 1.8501483212459603e-06,
+      "loss": 0.6914,
+      "step": 6746
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9875889973875243,
+      "learning_rate": 1.850097449031955e-06,
+      "loss": 0.6732,
+      "step": 6747
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9541449811965922,
+      "learning_rate": 1.8500465688839153e-06,
+      "loss": 0.6999,
+      "step": 6748
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.428990584058827,
+      "learning_rate": 1.8499956808023156e-06,
+      "loss": 0.6921,
+      "step": 6749
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.0720778019924604,
+      "learning_rate": 1.849944784787631e-06,
+      "loss": 0.6745,
+      "step": 6750
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.23513168856775,
+      "learning_rate": 1.8498938808403368e-06,
+      "loss": 0.6738,
+      "step": 6751
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8691727099860858,
+      "learning_rate": 1.8498429689609077e-06,
+      "loss": 0.6986,
+      "step": 6752
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9868938937371229,
+      "learning_rate": 1.849792049149819e-06,
+      "loss": 0.6608,
+      "step": 6753
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.24219265277345,
+      "learning_rate": 1.849741121407546e-06,
+      "loss": 0.6829,
+      "step": 6754
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6468302482115987,
+      "learning_rate": 1.8496901857345638e-06,
+      "loss": 0.6855,
+      "step": 6755
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0079755807610695,
+      "learning_rate": 1.8496392421313482e-06,
+      "loss": 0.6634,
+      "step": 6756
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6824658922518377,
+      "learning_rate": 1.8495882905983745e-06,
+      "loss": 0.6908,
+      "step": 6757
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.7156191020573557,
+      "learning_rate": 1.8495373311361178e-06,
+      "loss": 0.681,
+      "step": 6758
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.50217442669622,
+      "learning_rate": 1.8494863637450542e-06,
+      "loss": 0.6934,
+      "step": 6759
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.563188533605097,
+      "learning_rate": 1.8494353884256593e-06,
+      "loss": 0.7012,
+      "step": 6760
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6401980665392026,
+      "learning_rate": 1.849384405178409e-06,
+      "loss": 0.6569,
+      "step": 6761
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7588296227375738,
+      "learning_rate": 1.8493334140037787e-06,
+      "loss": 0.6725,
+      "step": 6762
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3643372236103,
+      "learning_rate": 1.8492824149022446e-06,
+      "loss": 0.6497,
+      "step": 6763
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6473290324342449,
+      "learning_rate": 1.8492314078742823e-06,
+      "loss": 0.653,
+      "step": 6764
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.849080898445266,
+      "learning_rate": 1.8491803929203687e-06,
+      "loss": 0.6497,
+      "step": 6765
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9333701705194225,
+      "learning_rate": 1.8491293700409789e-06,
+      "loss": 0.6615,
+      "step": 6766
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6253975661730835,
+      "learning_rate": 1.8490783392365897e-06,
+      "loss": 0.6732,
+      "step": 6767
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.500039846260997,
+      "learning_rate": 1.8490273005076776e-06,
+      "loss": 0.6523,
+      "step": 6768
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0349895059843996,
+      "learning_rate": 1.8489762538547184e-06,
+      "loss": 0.6888,
+      "step": 6769
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 6.232445569354863,
+      "learning_rate": 1.8489251992781887e-06,
+      "loss": 0.6797,
+      "step": 6770
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9427093178274546,
+      "learning_rate": 1.8488741367785651e-06,
+      "loss": 0.6921,
+      "step": 6771
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.2103804010433064,
+      "learning_rate": 1.8488230663563241e-06,
+      "loss": 0.6725,
+      "step": 6772
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0624014580042362,
+      "learning_rate": 1.8487719880119423e-06,
+      "loss": 0.6719,
+      "step": 6773
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.783392885548068,
+      "learning_rate": 1.8487209017458967e-06,
+      "loss": 0.6882,
+      "step": 6774
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9607285016975315,
+      "learning_rate": 1.8486698075586636e-06,
+      "loss": 0.666,
+      "step": 6775
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.859620780855016,
+      "learning_rate": 1.8486187054507202e-06,
+      "loss": 0.6934,
+      "step": 6776
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6127144290083153,
+      "learning_rate": 1.8485675954225433e-06,
+      "loss": 0.6973,
+      "step": 6777
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 7.535248662080887,
+      "learning_rate": 1.8485164774746103e-06,
+      "loss": 0.6745,
+      "step": 6778
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.263475649774124,
+      "learning_rate": 1.8484653516073977e-06,
+      "loss": 0.7188,
+      "step": 6779
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9928857678271354,
+      "learning_rate": 1.848414217821383e-06,
+      "loss": 0.7279,
+      "step": 6780
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.348088253370447,
+      "learning_rate": 1.8483630761170433e-06,
+      "loss": 0.6901,
+      "step": 6781
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.7003022291404304,
+      "learning_rate": 1.8483119264948561e-06,
+      "loss": 0.6771,
+      "step": 6782
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7689156347858895,
+      "learning_rate": 1.8482607689552987e-06,
+      "loss": 0.6536,
+      "step": 6783
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.359722194302439,
+      "learning_rate": 1.8482096034988483e-06,
+      "loss": 0.6686,
+      "step": 6784
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.920130466515206,
+      "learning_rate": 1.848158430125983e-06,
+      "loss": 0.6927,
+      "step": 6785
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8162601658576183,
+      "learning_rate": 1.8481072488371798e-06,
+      "loss": 0.6823,
+      "step": 6786
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.1297573839471067,
+      "learning_rate": 1.8480560596329169e-06,
+      "loss": 0.7057,
+      "step": 6787
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9166980454613854,
+      "learning_rate": 1.8480048625136715e-06,
+      "loss": 0.681,
+      "step": 6788
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.886696251920872,
+      "learning_rate": 1.847953657479922e-06,
+      "loss": 0.6589,
+      "step": 6789
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5002320147457997,
+      "learning_rate": 1.8479024445321458e-06,
+      "loss": 0.668,
+      "step": 6790
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 4.245867714994213,
+      "learning_rate": 1.8478512236708212e-06,
+      "loss": 0.6842,
+      "step": 6791
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.892279196622061,
+      "learning_rate": 1.8477999948964263e-06,
+      "loss": 0.6712,
+      "step": 6792
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.8260073805473119,
+      "learning_rate": 1.8477487582094386e-06,
+      "loss": 0.6751,
+      "step": 6793
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.681289355190913,
+      "learning_rate": 1.8476975136103373e-06,
+      "loss": 0.6862,
+      "step": 6794
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3691621522813286,
+      "learning_rate": 1.8476462610995997e-06,
+      "loss": 0.6855,
+      "step": 6795
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.4499993852873923,
+      "learning_rate": 1.847595000677705e-06,
+      "loss": 0.6934,
+      "step": 6796
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.6242882453956032,
+      "learning_rate": 1.847543732345131e-06,
+      "loss": 0.6966,
+      "step": 6797
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.152548475449875,
+      "learning_rate": 1.8474924561023561e-06,
+      "loss": 0.6927,
+      "step": 6798
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7881780141028234,
+      "learning_rate": 1.8474411719498596e-06,
+      "loss": 0.681,
+      "step": 6799
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5134501397178901,
+      "learning_rate": 1.8473898798881195e-06,
+      "loss": 0.6862,
+      "step": 6800
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.946344838914507,
+      "learning_rate": 1.8473385799176149e-06,
+      "loss": 0.6901,
+      "step": 6801
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.9832002540945646,
+      "learning_rate": 1.8472872720388243e-06,
+      "loss": 0.6771,
+      "step": 6802
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.0176979328212648,
+      "learning_rate": 1.8472359562522266e-06,
+      "loss": 0.6712,
+      "step": 6803
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.22598011632839,
+      "learning_rate": 1.8471846325583006e-06,
+      "loss": 0.6927,
+      "step": 6804
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3068081165127663,
+      "learning_rate": 1.8471333009575258e-06,
+      "loss": 0.7051,
+      "step": 6805
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.3202310620423963,
+      "learning_rate": 1.847081961450381e-06,
+      "loss": 0.6777,
+      "step": 6806
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8876538262049494,
+      "learning_rate": 1.847030614037345e-06,
+      "loss": 0.6908,
+      "step": 6807
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.4747738475843284,
+      "learning_rate": 1.8469792587188978e-06,
+      "loss": 0.6745,
+      "step": 6808
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7015275967678047,
+      "learning_rate": 1.846927895495518e-06,
+      "loss": 0.6829,
+      "step": 6809
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.6840199879397773,
+      "learning_rate": 1.8468765243676854e-06,
+      "loss": 0.6908,
+      "step": 6810
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.113101731279391,
+      "learning_rate": 1.8468251453358792e-06,
+      "loss": 0.6816,
+      "step": 6811
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1883197005175115,
+      "learning_rate": 1.8467737584005792e-06,
+      "loss": 0.6595,
+      "step": 6812
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.6054219936871439,
+      "learning_rate": 1.8467223635622647e-06,
+      "loss": 0.6803,
+      "step": 6813
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.591804572484437,
+      "learning_rate": 1.8466709608214154e-06,
+      "loss": 0.681,
+      "step": 6814
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.534461086297991,
+      "learning_rate": 1.8466195501785111e-06,
+      "loss": 0.6615,
+      "step": 6815
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.223742389347049,
+      "learning_rate": 1.846568131634032e-06,
+      "loss": 0.6608,
+      "step": 6816
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.341380339532581,
+      "learning_rate": 1.8465167051884576e-06,
+      "loss": 0.6576,
+      "step": 6817
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.0096767939913205,
+      "learning_rate": 1.8464652708422676e-06,
+      "loss": 0.6654,
+      "step": 6818
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.5056742475524838,
+      "learning_rate": 1.8464138285959425e-06,
+      "loss": 0.6719,
+      "step": 6819
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7821469951168276,
+      "learning_rate": 1.8463623784499625e-06,
+      "loss": 0.6738,
+      "step": 6820
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8871329757752502,
+      "learning_rate": 1.8463109204048072e-06,
+      "loss": 0.6602,
+      "step": 6821
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.370536852511089,
+      "learning_rate": 1.8462594544609575e-06,
+      "loss": 0.6797,
+      "step": 6822
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.793526410445219,
+      "learning_rate": 1.8462079806188937e-06,
+      "loss": 0.7018,
+      "step": 6823
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.54626507929667,
+      "learning_rate": 1.8461564988790959e-06,
+      "loss": 0.6784,
+      "step": 6824
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.146513123512455,
+      "learning_rate": 1.8461050092420442e-06,
+      "loss": 0.7233,
+      "step": 6825
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.5671859259583245,
+      "learning_rate": 1.8460535117082201e-06,
+      "loss": 0.7148,
+      "step": 6826
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.743574757752061,
+      "learning_rate": 1.8460020062781034e-06,
+      "loss": 0.681,
+      "step": 6827
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.26795205504625,
+      "learning_rate": 1.8459504929521753e-06,
+      "loss": 0.6875,
+      "step": 6828
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.5567677934238149,
+      "learning_rate": 1.8458989717309165e-06,
+      "loss": 0.6992,
+      "step": 6829
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.83705093921404,
+      "learning_rate": 1.8458474426148078e-06,
+      "loss": 0.6836,
+      "step": 6830
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.5240174576706944,
+      "learning_rate": 1.8457959056043299e-06,
+      "loss": 0.6908,
+      "step": 6831
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.5703641450804815,
+      "learning_rate": 1.8457443606999643e-06,
+      "loss": 0.6999,
+      "step": 6832
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 8.486355979846582,
+      "learning_rate": 1.8456928079021918e-06,
+      "loss": 0.7135,
+      "step": 6833
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 9.784229076067088,
+      "learning_rate": 1.8456412472114935e-06,
+      "loss": 0.7129,
+      "step": 6834
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.392795874577879,
+      "learning_rate": 1.8455896786283502e-06,
+      "loss": 0.7272,
+      "step": 6835
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7605321019874602,
+      "learning_rate": 1.8455381021532445e-06,
+      "loss": 0.6855,
+      "step": 6836
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.8535966518411713,
+      "learning_rate": 1.845486517786656e-06,
+      "loss": 0.6888,
+      "step": 6837
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5845611958135027,
+      "learning_rate": 1.8454349255290675e-06,
+      "loss": 0.6862,
+      "step": 6838
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.8199569647728384,
+      "learning_rate": 1.8453833253809604e-06,
+      "loss": 0.6686,
+      "step": 6839
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.602976990987608,
+      "learning_rate": 1.8453317173428153e-06,
+      "loss": 0.6764,
+      "step": 6840
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.38774768559328,
+      "learning_rate": 1.8452801014151152e-06,
+      "loss": 0.6569,
+      "step": 6841
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6857888398945486,
+      "learning_rate": 1.8452284775983405e-06,
+      "loss": 0.6966,
+      "step": 6842
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.1336735946827443,
+      "learning_rate": 1.8451768458929742e-06,
+      "loss": 0.6992,
+      "step": 6843
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.726299019781025,
+      "learning_rate": 1.8451252062994974e-06,
+      "loss": 0.6569,
+      "step": 6844
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.48094486264641,
+      "learning_rate": 1.8450735588183921e-06,
+      "loss": 0.696,
+      "step": 6845
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3898627905067968,
+      "learning_rate": 1.845021903450141e-06,
+      "loss": 0.6719,
+      "step": 6846
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.459733471666695,
+      "learning_rate": 1.8449702401952254e-06,
+      "loss": 0.7025,
+      "step": 6847
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.7739061747570144,
+      "learning_rate": 1.844918569054128e-06,
+      "loss": 0.6966,
+      "step": 6848
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.276687393683117,
+      "learning_rate": 1.844866890027331e-06,
+      "loss": 0.6875,
+      "step": 6849
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.344548947354527,
+      "learning_rate": 1.8448152031153162e-06,
+      "loss": 0.679,
+      "step": 6850
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.5871306742574416,
+      "learning_rate": 1.8447635083185666e-06,
+      "loss": 0.696,
+      "step": 6851
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5492273559652694,
+      "learning_rate": 1.8447118056375645e-06,
+      "loss": 0.6803,
+      "step": 6852
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.5077825969968197,
+      "learning_rate": 1.8446600950727924e-06,
+      "loss": 0.6947,
+      "step": 6853
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 5.111806280403876,
+      "learning_rate": 1.8446083766247332e-06,
+      "loss": 0.681,
+      "step": 6854
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.519155202004483,
+      "learning_rate": 1.8445566502938687e-06,
+      "loss": 0.6745,
+      "step": 6855
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.493245256524762,
+      "learning_rate": 1.844504916080683e-06,
+      "loss": 0.6758,
+      "step": 6856
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7329255473633075,
+      "learning_rate": 1.8444531739856579e-06,
+      "loss": 0.6816,
+      "step": 6857
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.636959941096662,
+      "learning_rate": 1.8444014240092764e-06,
+      "loss": 0.6777,
+      "step": 6858
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 6.147229976036517,
+      "learning_rate": 1.8443496661520221e-06,
+      "loss": 0.6549,
+      "step": 6859
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.825591211529964,
+      "learning_rate": 1.8442979004143775e-06,
+      "loss": 0.6849,
+      "step": 6860
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.06005844833952,
+      "learning_rate": 1.8442461267968258e-06,
+      "loss": 0.6751,
+      "step": 6861
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1582273944328443,
+      "learning_rate": 1.8441943452998507e-06,
+      "loss": 0.6875,
+      "step": 6862
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 7.3410479180595525,
+      "learning_rate": 1.844142555923935e-06,
+      "loss": 0.6738,
+      "step": 6863
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.84869403029646,
+      "learning_rate": 1.8440907586695622e-06,
+      "loss": 0.681,
+      "step": 6864
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.5669986497674095,
+      "learning_rate": 1.8440389535372156e-06,
+      "loss": 0.6712,
+      "step": 6865
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.48592874865449,
+      "learning_rate": 1.8439871405273788e-06,
+      "loss": 0.6458,
+      "step": 6866
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0473770861850884,
+      "learning_rate": 1.8439353196405354e-06,
+      "loss": 0.6621,
+      "step": 6867
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.5760914234304084,
+      "learning_rate": 1.843883490877169e-06,
+      "loss": 0.653,
+      "step": 6868
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.0546314329624293,
+      "learning_rate": 1.8438316542377637e-06,
+      "loss": 0.6816,
+      "step": 6869
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 4.006537285587918,
+      "learning_rate": 1.8437798097228026e-06,
+      "loss": 0.6699,
+      "step": 6870
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.3865864664023255,
+      "learning_rate": 1.8437279573327697e-06,
+      "loss": 0.679,
+      "step": 6871
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9350968381060383,
+      "learning_rate": 1.8436760970681496e-06,
+      "loss": 0.7005,
+      "step": 6872
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8416284279218665,
+      "learning_rate": 1.8436242289294258e-06,
+      "loss": 0.6777,
+      "step": 6873
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9135277141268533,
+      "learning_rate": 1.843572352917082e-06,
+      "loss": 0.6712,
+      "step": 6874
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.8085342249713965,
+      "learning_rate": 1.8435204690316035e-06,
+      "loss": 0.6927,
+      "step": 6875
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 7.298518372166924,
+      "learning_rate": 1.8434685772734735e-06,
+      "loss": 0.681,
+      "step": 6876
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8455938241087184,
+      "learning_rate": 1.8434166776431765e-06,
+      "loss": 0.6868,
+      "step": 6877
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8962717913764853,
+      "learning_rate": 1.8433647701411973e-06,
+      "loss": 0.7005,
+      "step": 6878
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.732344821712517,
+      "learning_rate": 1.8433128547680203e-06,
+      "loss": 0.6706,
+      "step": 6879
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.037623592887603,
+      "learning_rate": 1.8432609315241296e-06,
+      "loss": 0.6901,
+      "step": 6880
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5777038483573511,
+      "learning_rate": 1.8432090004100101e-06,
+      "loss": 0.666,
+      "step": 6881
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.1971842647950766,
+      "learning_rate": 1.8431570614261465e-06,
+      "loss": 0.6979,
+      "step": 6882
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.1466900255531325,
+      "learning_rate": 1.8431051145730234e-06,
+      "loss": 0.7044,
+      "step": 6883
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.3437441223671833,
+      "learning_rate": 1.8430531598511256e-06,
+      "loss": 0.7038,
+      "step": 6884
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9387235880536544,
+      "learning_rate": 1.8430011972609382e-06,
+      "loss": 0.6758,
+      "step": 6885
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7270623785115673,
+      "learning_rate": 1.842949226802946e-06,
+      "loss": 0.6719,
+      "step": 6886
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.447833666845079,
+      "learning_rate": 1.8428972484776344e-06,
+      "loss": 0.6829,
+      "step": 6887
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.959904541970192,
+      "learning_rate": 1.842845262285488e-06,
+      "loss": 0.6719,
+      "step": 6888
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.186131468049194,
+      "learning_rate": 1.8427932682269924e-06,
+      "loss": 0.6953,
+      "step": 6889
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7545612609435365,
+      "learning_rate": 1.8427412663026325e-06,
+      "loss": 0.6842,
+      "step": 6890
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6868139495211089,
+      "learning_rate": 1.842689256512894e-06,
+      "loss": 0.6803,
+      "step": 6891
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9244783126988934,
+      "learning_rate": 1.8426372388582618e-06,
+      "loss": 0.7064,
+      "step": 6892
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4867175131117807,
+      "learning_rate": 1.842585213339222e-06,
+      "loss": 0.6725,
+      "step": 6893
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8925272278774259,
+      "learning_rate": 1.84253317995626e-06,
+      "loss": 0.6458,
+      "step": 6894
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5775900129256245,
+      "learning_rate": 1.8424811387098612e-06,
+      "loss": 0.6784,
+      "step": 6895
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8617580796913886,
+      "learning_rate": 1.8424290896005115e-06,
+      "loss": 0.6738,
+      "step": 6896
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.2349797975031005,
+      "learning_rate": 1.8423770326286965e-06,
+      "loss": 0.6836,
+      "step": 6897
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.46622570146322917,
+      "learning_rate": 1.8423249677949023e-06,
+      "loss": 0.681,
+      "step": 6898
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9244993868671156,
+      "learning_rate": 1.8422728950996144e-06,
+      "loss": 0.6536,
+      "step": 6899
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.828594571400317,
+      "learning_rate": 1.8422208145433194e-06,
+      "loss": 0.6589,
+      "step": 6900
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6155074860157153,
+      "learning_rate": 1.8421687261265029e-06,
+      "loss": 0.6706,
+      "step": 6901
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0112566083522554,
+      "learning_rate": 1.8421166298496513e-06,
+      "loss": 0.6745,
+      "step": 6902
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7183687606726806,
+      "learning_rate": 1.8420645257132505e-06,
+      "loss": 0.6836,
+      "step": 6903
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5455898485403436,
+      "learning_rate": 1.842012413717787e-06,
+      "loss": 0.6738,
+      "step": 6904
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.010068654465921,
+      "learning_rate": 1.8419602938637474e-06,
+      "loss": 0.6576,
+      "step": 6905
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.3982428409225247,
+      "learning_rate": 1.841908166151618e-06,
+      "loss": 0.6829,
+      "step": 6906
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.092633869139048,
+      "learning_rate": 1.841856030581885e-06,
+      "loss": 0.6725,
+      "step": 6907
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.446780600503024,
+      "learning_rate": 1.8418038871550353e-06,
+      "loss": 0.668,
+      "step": 6908
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7290841500291892,
+      "learning_rate": 1.8417517358715554e-06,
+      "loss": 0.7018,
+      "step": 6909
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.482568988596345,
+      "learning_rate": 1.8416995767319324e-06,
+      "loss": 0.6641,
+      "step": 6910
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.736489909309705,
+      "learning_rate": 1.8416474097366525e-06,
+      "loss": 0.6745,
+      "step": 6911
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5301430403258154,
+      "learning_rate": 1.8415952348862029e-06,
+      "loss": 0.6628,
+      "step": 6912
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8495765900201016,
+      "learning_rate": 1.8415430521810706e-06,
+      "loss": 0.679,
+      "step": 6913
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7872846333844105,
+      "learning_rate": 1.8414908616217426e-06,
+      "loss": 0.6816,
+      "step": 6914
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.7907305768924149,
+      "learning_rate": 1.8414386632087062e-06,
+      "loss": 0.7057,
+      "step": 6915
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.441456922668186,
+      "learning_rate": 1.8413864569424479e-06,
+      "loss": 0.6634,
+      "step": 6916
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.721025016187932,
+      "learning_rate": 1.8413342428234558e-06,
+      "loss": 0.6686,
+      "step": 6917
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6620451435087231,
+      "learning_rate": 1.8412820208522164e-06,
+      "loss": 0.6771,
+      "step": 6918
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.4200338184336925,
+      "learning_rate": 1.8412297910292178e-06,
+      "loss": 0.6888,
+      "step": 6919
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.0996543466124225,
+      "learning_rate": 1.841177553354947e-06,
+      "loss": 0.6582,
+      "step": 6920
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.4004504670486972,
+      "learning_rate": 1.8411253078298918e-06,
+      "loss": 0.6745,
+      "step": 6921
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.746607041329221,
+      "learning_rate": 1.8410730544545398e-06,
+      "loss": 0.6602,
+      "step": 6922
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.9535937351691386,
+      "learning_rate": 1.8410207932293784e-06,
+      "loss": 0.6484,
+      "step": 6923
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.879216796409624,
+      "learning_rate": 1.8409685241548958e-06,
+      "loss": 0.6745,
+      "step": 6924
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2062075838555837,
+      "learning_rate": 1.8409162472315796e-06,
+      "loss": 0.7025,
+      "step": 6925
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0314565775203697,
+      "learning_rate": 1.8408639624599175e-06,
+      "loss": 0.6855,
+      "step": 6926
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5363197379065352,
+      "learning_rate": 1.8408116698403976e-06,
+      "loss": 0.6953,
+      "step": 6927
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.1033920491683755,
+      "learning_rate": 1.8407593693735082e-06,
+      "loss": 0.6816,
+      "step": 6928
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 5.2546460950721,
+      "learning_rate": 1.8407070610597373e-06,
+      "loss": 0.7005,
+      "step": 6929
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.3052991027655403,
+      "learning_rate": 1.8406547448995729e-06,
+      "loss": 0.6751,
+      "step": 6930
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.0937849344584176,
+      "learning_rate": 1.8406024208935036e-06,
+      "loss": 0.6602,
+      "step": 6931
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.2059286828470344,
+      "learning_rate": 1.8405500890420174e-06,
+      "loss": 0.6777,
+      "step": 6932
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.0978159782057326,
+      "learning_rate": 1.8404977493456028e-06,
+      "loss": 0.6882,
+      "step": 6933
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5986559514920609,
+      "learning_rate": 1.8404454018047485e-06,
+      "loss": 0.6745,
+      "step": 6934
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 6.45814365090293,
+      "learning_rate": 1.840393046419943e-06,
+      "loss": 0.6999,
+      "step": 6935
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.7208103549842235,
+      "learning_rate": 1.840340683191675e-06,
+      "loss": 0.6875,
+      "step": 6936
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 4.751942688760206,
+      "learning_rate": 1.8402883121204328e-06,
+      "loss": 0.6803,
+      "step": 6937
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.4289532171646524,
+      "learning_rate": 1.8402359332067055e-06,
+      "loss": 0.6615,
+      "step": 6938
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.733425970956521,
+      "learning_rate": 1.8401835464509822e-06,
+      "loss": 0.6842,
+      "step": 6939
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.270432978966753,
+      "learning_rate": 1.8401311518537512e-06,
+      "loss": 0.679,
+      "step": 6940
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.3152279964866898,
+      "learning_rate": 1.8400787494155021e-06,
+      "loss": 0.6764,
+      "step": 6941
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6473599619457052,
+      "learning_rate": 1.8400263391367237e-06,
+      "loss": 0.6777,
+      "step": 6942
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.364327521105375,
+      "learning_rate": 1.8399739210179051e-06,
+      "loss": 0.6986,
+      "step": 6943
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0384858046276955,
+      "learning_rate": 1.839921495059536e-06,
+      "loss": 0.6712,
+      "step": 6944
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.5241656680404483,
+      "learning_rate": 1.8398690612621049e-06,
+      "loss": 0.6868,
+      "step": 6945
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1875306119866975,
+      "learning_rate": 1.8398166196261013e-06,
+      "loss": 0.6921,
+      "step": 6946
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9611383852443112,
+      "learning_rate": 1.8397641701520155e-06,
+      "loss": 0.6732,
+      "step": 6947
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6943692075029656,
+      "learning_rate": 1.8397117128403364e-06,
+      "loss": 0.6823,
+      "step": 6948
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.043145169576151,
+      "learning_rate": 1.839659247691553e-06,
+      "loss": 0.6888,
+      "step": 6949
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.203874719225912,
+      "learning_rate": 1.8396067747061563e-06,
+      "loss": 0.6771,
+      "step": 6950
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9771288780094791,
+      "learning_rate": 1.839554293884635e-06,
+      "loss": 0.666,
+      "step": 6951
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.76779142078125,
+      "learning_rate": 1.8395018052274794e-06,
+      "loss": 0.6842,
+      "step": 6952
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.2628157201539942,
+      "learning_rate": 1.839449308735179e-06,
+      "loss": 0.6908,
+      "step": 6953
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.653751092667478,
+      "learning_rate": 1.839396804408224e-06,
+      "loss": 0.668,
+      "step": 6954
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.8261520290584325,
+      "learning_rate": 1.8393442922471046e-06,
+      "loss": 0.6712,
+      "step": 6955
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.384425379013841,
+      "learning_rate": 1.8392917722523105e-06,
+      "loss": 0.6771,
+      "step": 6956
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.2142774320123755,
+      "learning_rate": 1.839239244424332e-06,
+      "loss": 0.6725,
+      "step": 6957
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.769118311425131,
+      "learning_rate": 1.8391867087636595e-06,
+      "loss": 0.6829,
+      "step": 6958
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9249647672719445,
+      "learning_rate": 1.8391341652707831e-06,
+      "loss": 0.6771,
+      "step": 6959
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.3117360219265475,
+      "learning_rate": 1.8390816139461936e-06,
+      "loss": 0.6589,
+      "step": 6960
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.1455509043009917,
+      "learning_rate": 1.839029054790381e-06,
+      "loss": 0.6738,
+      "step": 6961
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.1397213511152007,
+      "learning_rate": 1.838976487803836e-06,
+      "loss": 0.6699,
+      "step": 6962
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8229949503959654,
+      "learning_rate": 1.8389239129870496e-06,
+      "loss": 0.6895,
+      "step": 6963
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.3713543696825754,
+      "learning_rate": 1.838871330340512e-06,
+      "loss": 0.6771,
+      "step": 6964
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.7191832740174045,
+      "learning_rate": 1.8388187398647137e-06,
+      "loss": 0.6979,
+      "step": 6965
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.9399288275359434,
+      "learning_rate": 1.8387661415601462e-06,
+      "loss": 0.6862,
+      "step": 6966
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.124261985915034,
+      "learning_rate": 1.8387135354273001e-06,
+      "loss": 0.7038,
+      "step": 6967
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6263400972907583,
+      "learning_rate": 1.8386609214666663e-06,
+      "loss": 0.6895,
+      "step": 6968
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.1864622125295454,
+      "learning_rate": 1.8386082996787362e-06,
+      "loss": 0.6667,
+      "step": 6969
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.077218208657157,
+      "learning_rate": 1.8385556700640005e-06,
+      "loss": 0.6816,
+      "step": 6970
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.595069878526633,
+      "learning_rate": 1.8385030326229506e-06,
+      "loss": 0.6882,
+      "step": 6971
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.332737378531102,
+      "learning_rate": 1.8384503873560778e-06,
+      "loss": 0.668,
+      "step": 6972
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.337600198430594,
+      "learning_rate": 1.8383977342638735e-06,
+      "loss": 0.6901,
+      "step": 6973
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.134323736464271,
+      "learning_rate": 1.8383450733468286e-06,
+      "loss": 0.6803,
+      "step": 6974
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.4152988419774273,
+      "learning_rate": 1.8382924046054352e-06,
+      "loss": 0.6523,
+      "step": 6975
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.193997142127918,
+      "learning_rate": 1.8382397280401849e-06,
+      "loss": 0.6706,
+      "step": 6976
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0256347334423845,
+      "learning_rate": 1.838187043651569e-06,
+      "loss": 0.6615,
+      "step": 6977
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.039229828522437,
+      "learning_rate": 1.838134351440079e-06,
+      "loss": 0.6934,
+      "step": 6978
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.9066970600833497,
+      "learning_rate": 1.8380816514062072e-06,
+      "loss": 0.6706,
+      "step": 6979
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6945582231200902,
+      "learning_rate": 1.8380289435504453e-06,
+      "loss": 0.6758,
+      "step": 6980
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.981273892688787,
+      "learning_rate": 1.837976227873285e-06,
+      "loss": 0.6797,
+      "step": 6981
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7763977050198088,
+      "learning_rate": 1.8379235043752184e-06,
+      "loss": 0.6966,
+      "step": 6982
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 5.091701919774396,
+      "learning_rate": 1.8378707730567378e-06,
+      "loss": 0.6595,
+      "step": 6983
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5979210897815055,
+      "learning_rate": 1.8378180339183352e-06,
+      "loss": 0.6862,
+      "step": 6984
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.596255551888981,
+      "learning_rate": 1.8377652869605024e-06,
+      "loss": 0.6836,
+      "step": 6985
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6734389724006666,
+      "learning_rate": 1.8377125321837324e-06,
+      "loss": 0.6504,
+      "step": 6986
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4666357107382213,
+      "learning_rate": 1.8376597695885173e-06,
+      "loss": 0.6803,
+      "step": 6987
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.2334569889172549,
+      "learning_rate": 1.8376069991753497e-06,
+      "loss": 0.6836,
+      "step": 6988
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.547190201995633,
+      "learning_rate": 1.8375542209447214e-06,
+      "loss": 0.7292,
+      "step": 6989
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 6.344378800543547,
+      "learning_rate": 1.8375014348971257e-06,
+      "loss": 0.6979,
+      "step": 6990
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.0010954992925227,
+      "learning_rate": 1.8374486410330552e-06,
+      "loss": 0.6934,
+      "step": 6991
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.3691443768000584,
+      "learning_rate": 1.8373958393530023e-06,
+      "loss": 0.6615,
+      "step": 6992
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5633820955065174,
+      "learning_rate": 1.83734302985746e-06,
+      "loss": 0.6777,
+      "step": 6993
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.8846716164747783,
+      "learning_rate": 1.8372902125469212e-06,
+      "loss": 0.6764,
+      "step": 6994
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7361719144036136,
+      "learning_rate": 1.8372373874218785e-06,
+      "loss": 0.6686,
+      "step": 6995
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.113671541942443,
+      "learning_rate": 1.8371845544828256e-06,
+      "loss": 0.6849,
+      "step": 6996
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9394454652473033,
+      "learning_rate": 1.8371317137302552e-06,
+      "loss": 0.6582,
+      "step": 6997
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.405104783736042,
+      "learning_rate": 1.8370788651646603e-06,
+      "loss": 0.6536,
+      "step": 6998
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.9202913370748615,
+      "learning_rate": 1.8370260087865343e-06,
+      "loss": 0.6882,
+      "step": 6999
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.4779870369863732,
+      "learning_rate": 1.8369731445963708e-06,
+      "loss": 0.6771,
+      "step": 7000
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7515071177596435,
+      "learning_rate": 1.8369202725946629e-06,
+      "loss": 0.7129,
+      "step": 7001
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5627329832001353,
+      "learning_rate": 1.8368673927819038e-06,
+      "loss": 0.6647,
+      "step": 7002
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.317286729163837,
+      "learning_rate": 1.8368145051585876e-06,
+      "loss": 0.6745,
+      "step": 7003
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.977568786908681,
+      "learning_rate": 1.8367616097252075e-06,
+      "loss": 0.7083,
+      "step": 7004
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 4.853377848605319,
+      "learning_rate": 1.8367087064822573e-06,
+      "loss": 0.6595,
+      "step": 7005
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2250366702646063,
+      "learning_rate": 1.8366557954302307e-06,
+      "loss": 0.6836,
+      "step": 7006
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.0299108540073463,
+      "learning_rate": 1.8366028765696214e-06,
+      "loss": 0.6732,
+      "step": 7007
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.888239416443442,
+      "learning_rate": 1.8365499499009239e-06,
+      "loss": 0.6927,
+      "step": 7008
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.433977631771741,
+      "learning_rate": 1.8364970154246314e-06,
+      "loss": 0.6634,
+      "step": 7009
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3760901383260116,
+      "learning_rate": 1.8364440731412385e-06,
+      "loss": 0.6947,
+      "step": 7010
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.64617304957815,
+      "learning_rate": 1.8363911230512387e-06,
+      "loss": 0.6966,
+      "step": 7011
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4230207030324622,
+      "learning_rate": 1.8363381651551268e-06,
+      "loss": 0.6842,
+      "step": 7012
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5803123828959674,
+      "learning_rate": 1.836285199453397e-06,
+      "loss": 0.6621,
+      "step": 7013
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.392937707968575,
+      "learning_rate": 1.8362322259465434e-06,
+      "loss": 0.681,
+      "step": 7014
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7102843290360398,
+      "learning_rate": 1.8361792446350606e-06,
+      "loss": 0.6497,
+      "step": 7015
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.272671658516419,
+      "learning_rate": 1.8361262555194427e-06,
+      "loss": 0.6615,
+      "step": 7016
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.0379674039137496,
+      "learning_rate": 1.8360732586001847e-06,
+      "loss": 0.6803,
+      "step": 7017
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.765639151578921,
+      "learning_rate": 1.8360202538777808e-06,
+      "loss": 0.6829,
+      "step": 7018
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1939243796631627,
+      "learning_rate": 1.8359672413527263e-06,
+      "loss": 0.6686,
+      "step": 7019
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.1868010972083478,
+      "learning_rate": 1.8359142210255155e-06,
+      "loss": 0.668,
+      "step": 7020
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7580533381631278,
+      "learning_rate": 1.8358611928966434e-06,
+      "loss": 0.6836,
+      "step": 7021
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.5991755028224198,
+      "learning_rate": 1.8358081569666046e-06,
+      "loss": 0.6849,
+      "step": 7022
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.5383124683610108,
+      "learning_rate": 1.8357551132358946e-06,
+      "loss": 0.6686,
+      "step": 7023
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9656800938949044,
+      "learning_rate": 1.835702061705008e-06,
+      "loss": 0.6823,
+      "step": 7024
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.4750192350790954,
+      "learning_rate": 1.8356490023744402e-06,
+      "loss": 0.6764,
+      "step": 7025
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.382053122914745,
+      "learning_rate": 1.8355959352446867e-06,
+      "loss": 0.6823,
+      "step": 7026
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.8408335095165076,
+      "learning_rate": 1.835542860316242e-06,
+      "loss": 0.6908,
+      "step": 7027
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.343764557749404,
+      "learning_rate": 1.835489777589602e-06,
+      "loss": 0.6803,
+      "step": 7028
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3736603994956686,
+      "learning_rate": 1.835436687065262e-06,
+      "loss": 0.6556,
+      "step": 7029
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4261664402564007,
+      "learning_rate": 1.8353835887437176e-06,
+      "loss": 0.6803,
+      "step": 7030
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.862417439171769,
+      "learning_rate": 1.8353304826254642e-06,
+      "loss": 0.6569,
+      "step": 7031
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.0600305148018583,
+      "learning_rate": 1.8352773687109976e-06,
+      "loss": 0.6823,
+      "step": 7032
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.5791622268553,
+      "learning_rate": 1.8352242470008138e-06,
+      "loss": 0.6934,
+      "step": 7033
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.967024117235244,
+      "learning_rate": 1.835171117495408e-06,
+      "loss": 0.6908,
+      "step": 7034
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.714683802174773,
+      "learning_rate": 1.835117980195276e-06,
+      "loss": 0.6947,
+      "step": 7035
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.6159900552154034,
+      "learning_rate": 1.835064835100914e-06,
+      "loss": 0.7005,
+      "step": 7036
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.350253435669832,
+      "learning_rate": 1.8350116822128185e-06,
+      "loss": 0.7109,
+      "step": 7037
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8063935848577485,
+      "learning_rate": 1.8349585215314848e-06,
+      "loss": 0.6836,
+      "step": 7038
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6775123943310326,
+      "learning_rate": 1.8349053530574096e-06,
+      "loss": 0.7083,
+      "step": 7039
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.356741164768443,
+      "learning_rate": 1.834852176791089e-06,
+      "loss": 0.6868,
+      "step": 7040
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.834046022470828,
+      "learning_rate": 1.834798992733019e-06,
+      "loss": 0.6673,
+      "step": 7041
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.1969754518600044,
+      "learning_rate": 1.834745800883696e-06,
+      "loss": 0.6595,
+      "step": 7042
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9415839029212515,
+      "learning_rate": 1.834692601243617e-06,
+      "loss": 0.6673,
+      "step": 7043
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.2085031179095451,
+      "learning_rate": 1.8346393938132779e-06,
+      "loss": 0.6901,
+      "step": 7044
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9042916590903297,
+      "learning_rate": 1.8345861785931755e-06,
+      "loss": 0.6667,
+      "step": 7045
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.1537465296068696,
+      "learning_rate": 1.8345329555838064e-06,
+      "loss": 0.6478,
+      "step": 7046
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.548910073912032,
+      "learning_rate": 1.8344797247856678e-06,
+      "loss": 0.6764,
+      "step": 7047
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 6.414736194592172,
+      "learning_rate": 1.8344264861992557e-06,
+      "loss": 0.6947,
+      "step": 7048
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.370108422160381,
+      "learning_rate": 1.8343732398250677e-06,
+      "loss": 0.6921,
+      "step": 7049
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2963286140406933,
+      "learning_rate": 1.8343199856636003e-06,
+      "loss": 0.6875,
+      "step": 7050
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.868637484090843,
+      "learning_rate": 1.834266723715351e-06,
+      "loss": 0.6823,
+      "step": 7051
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.983246709335134,
+      "learning_rate": 1.834213453980816e-06,
+      "loss": 0.6751,
+      "step": 7052
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.159870374616478,
+      "learning_rate": 1.8341601764604934e-06,
+      "loss": 0.6745,
+      "step": 7053
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.096160772447296,
+      "learning_rate": 1.8341068911548802e-06,
+      "loss": 0.6927,
+      "step": 7054
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.248613302362108,
+      "learning_rate": 1.8340535980644738e-06,
+      "loss": 0.6823,
+      "step": 7055
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.739246342192815,
+      "learning_rate": 1.834000297189771e-06,
+      "loss": 0.681,
+      "step": 7056
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.3974164951541277,
+      "learning_rate": 1.8339469885312698e-06,
+      "loss": 0.6823,
+      "step": 7057
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.9973841852427423,
+      "learning_rate": 1.8338936720894677e-06,
+      "loss": 0.6888,
+      "step": 7058
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.703986942988171,
+      "learning_rate": 1.833840347864862e-06,
+      "loss": 0.6699,
+      "step": 7059
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.167274639556455,
+      "learning_rate": 1.833787015857951e-06,
+      "loss": 0.6764,
+      "step": 7060
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 4.098702202560682,
+      "learning_rate": 1.8337336760692317e-06,
+      "loss": 0.6888,
+      "step": 7061
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7047720948018918,
+      "learning_rate": 1.8336803284992022e-06,
+      "loss": 0.6478,
+      "step": 7062
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.6449284414239116,
+      "learning_rate": 1.8336269731483606e-06,
+      "loss": 0.6549,
+      "step": 7063
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.83699967647789,
+      "learning_rate": 1.833573610017205e-06,
+      "loss": 0.6836,
+      "step": 7064
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.6733760496729064,
+      "learning_rate": 1.8335202391062327e-06,
+      "loss": 0.6419,
+      "step": 7065
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4561065467831493,
+      "learning_rate": 1.8334668604159425e-06,
+      "loss": 0.6745,
+      "step": 7066
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.6063076575877604,
+      "learning_rate": 1.8334134739468326e-06,
+      "loss": 0.6842,
+      "step": 7067
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.3693995719979304,
+      "learning_rate": 1.8333600796994006e-06,
+      "loss": 0.6686,
+      "step": 7068
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.414226897615028,
+      "learning_rate": 1.8333066776741458e-06,
+      "loss": 0.6999,
+      "step": 7069
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 3.8130650576760945,
+      "learning_rate": 1.8332532678715659e-06,
+      "loss": 0.6569,
+      "step": 7070
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.043659346053398,
+      "learning_rate": 1.8331998502921592e-06,
+      "loss": 0.6803,
+      "step": 7071
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7739993252773123,
+      "learning_rate": 1.833146424936425e-06,
+      "loss": 0.6667,
+      "step": 7072
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.5353832494233224,
+      "learning_rate": 1.8330929918048615e-06,
+      "loss": 0.6823,
+      "step": 7073
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0520049406899967,
+      "learning_rate": 1.8330395508979673e-06,
+      "loss": 0.6686,
+      "step": 7074
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.9255779063344547,
+      "learning_rate": 1.8329861022162414e-06,
+      "loss": 0.6816,
+      "step": 7075
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.241967781676939,
+      "learning_rate": 1.8329326457601826e-06,
+      "loss": 0.666,
+      "step": 7076
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.3355874517721813,
+      "learning_rate": 1.8328791815302895e-06,
+      "loss": 0.6738,
+      "step": 7077
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7544157387399582,
+      "learning_rate": 1.8328257095270616e-06,
+      "loss": 0.7005,
+      "step": 7078
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.4829915035580905,
+      "learning_rate": 1.8327722297509976e-06,
+      "loss": 0.6836,
+      "step": 7079
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0371745250605926,
+      "learning_rate": 1.8327187422025968e-06,
+      "loss": 0.6751,
+      "step": 7080
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.613898731058296,
+      "learning_rate": 1.832665246882358e-06,
+      "loss": 0.6693,
+      "step": 7081
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.421058742754259,
+      "learning_rate": 1.8326117437907812e-06,
+      "loss": 0.6849,
+      "step": 7082
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.145287160892493,
+      "learning_rate": 1.832558232928365e-06,
+      "loss": 0.7109,
+      "step": 7083
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9418527697637538,
+      "learning_rate": 1.8325047142956098e-06,
+      "loss": 0.6641,
+      "step": 7084
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.164228164370004,
+      "learning_rate": 1.832451187893014e-06,
+      "loss": 0.6738,
+      "step": 7085
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2808652913065455,
+      "learning_rate": 1.832397653721078e-06,
+      "loss": 0.6908,
+      "step": 7086
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9065155679666836,
+      "learning_rate": 1.8323441117803008e-06,
+      "loss": 0.6725,
+      "step": 7087
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.9176741998491389,
+      "learning_rate": 1.8322905620711824e-06,
+      "loss": 0.7096,
+      "step": 7088
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.438834112762495,
+      "learning_rate": 1.8322370045942224e-06,
+      "loss": 0.6999,
+      "step": 7089
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0637657947511272,
+      "learning_rate": 1.8321834393499211e-06,
+      "loss": 0.6576,
+      "step": 7090
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.1275629763338513,
+      "learning_rate": 1.8321298663387782e-06,
+      "loss": 0.6712,
+      "step": 7091
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.8867842510907793,
+      "learning_rate": 1.8320762855612934e-06,
+      "loss": 0.6777,
+      "step": 7092
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.829661851413002,
+      "learning_rate": 1.8320226970179675e-06,
+      "loss": 0.6628,
+      "step": 7093
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.433348019740995,
+      "learning_rate": 1.8319691007092998e-06,
+      "loss": 0.6374,
+      "step": 7094
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6073094696283741,
+      "learning_rate": 1.831915496635791e-06,
+      "loss": 0.6986,
+      "step": 7095
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.5758318576645767,
+      "learning_rate": 1.8318618847979415e-06,
+      "loss": 0.6797,
+      "step": 7096
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.8607217185433162,
+      "learning_rate": 1.8318082651962512e-06,
+      "loss": 0.6862,
+      "step": 7097
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6558368760628744,
+      "learning_rate": 1.8317546378312208e-06,
+      "loss": 0.681,
+      "step": 7098
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5878091754615791,
+      "learning_rate": 1.831701002703351e-06,
+      "loss": 0.6706,
+      "step": 7099
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.780955245189495,
+      "learning_rate": 1.831647359813142e-06,
+      "loss": 0.7012,
+      "step": 7100
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.3831449689587536,
+      "learning_rate": 1.831593709161095e-06,
+      "loss": 0.6706,
+      "step": 7101
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6485505473150068,
+      "learning_rate": 1.8315400507477098e-06,
+      "loss": 0.6816,
+      "step": 7102
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6992079315930506,
+      "learning_rate": 1.831486384573488e-06,
+      "loss": 0.6947,
+      "step": 7103
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.315835060746847,
+      "learning_rate": 1.8314327106389305e-06,
+      "loss": 0.696,
+      "step": 7104
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.068222775989483,
+      "learning_rate": 1.831379028944538e-06,
+      "loss": 0.6647,
+      "step": 7105
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7831143212259006,
+      "learning_rate": 1.831325339490811e-06,
+      "loss": 0.6595,
+      "step": 7106
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.3409327893443241,
+      "learning_rate": 1.8312716422782516e-06,
+      "loss": 0.6654,
+      "step": 7107
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.3547332150329514,
+      "learning_rate": 1.8312179373073602e-06,
+      "loss": 0.7025,
+      "step": 7108
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5616662622358014,
+      "learning_rate": 1.831164224578638e-06,
+      "loss": 0.679,
+      "step": 7109
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.4894623158981564,
+      "learning_rate": 1.831110504092587e-06,
+      "loss": 0.6829,
+      "step": 7110
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.591774693859202,
+      "learning_rate": 1.8310567758497082e-06,
+      "loss": 0.6979,
+      "step": 7111
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0140090189874584,
+      "learning_rate": 1.8310030398505029e-06,
+      "loss": 0.6647,
+      "step": 7112
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8090386604577016,
+      "learning_rate": 1.8309492960954727e-06,
+      "loss": 0.6719,
+      "step": 7113
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5246803069542283,
+      "learning_rate": 1.830895544585119e-06,
+      "loss": 0.6901,
+      "step": 7114
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.214566598921988,
+      "learning_rate": 1.830841785319944e-06,
+      "loss": 0.6934,
+      "step": 7115
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.719441105514477,
+      "learning_rate": 1.8307880183004492e-06,
+      "loss": 0.6576,
+      "step": 7116
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.549942026861599,
+      "learning_rate": 1.8307342435271363e-06,
+      "loss": 0.6738,
+      "step": 7117
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.0030333206415847,
+      "learning_rate": 1.830680461000507e-06,
+      "loss": 0.6784,
+      "step": 7118
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 6.211669397183332,
+      "learning_rate": 1.8306266707210638e-06,
+      "loss": 0.696,
+      "step": 7119
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.1889396266005487,
+      "learning_rate": 1.8305728726893084e-06,
+      "loss": 0.6868,
+      "step": 7120
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.2982743871531153,
+      "learning_rate": 1.830519066905743e-06,
+      "loss": 0.6589,
+      "step": 7121
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.4667908707828778,
+      "learning_rate": 1.8304652533708693e-06,
+      "loss": 0.696,
+      "step": 7122
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9845752403432624,
+      "learning_rate": 1.8304114320851902e-06,
+      "loss": 0.6862,
+      "step": 7123
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.8500376316686333,
+      "learning_rate": 1.8303576030492079e-06,
+      "loss": 0.668,
+      "step": 7124
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.7197156305823906,
+      "learning_rate": 1.8303037662634245e-06,
+      "loss": 0.6693,
+      "step": 7125
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.2185052699093195,
+      "learning_rate": 1.8302499217283427e-06,
+      "loss": 0.6621,
+      "step": 7126
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5947070808856378,
+      "learning_rate": 1.8301960694444648e-06,
+      "loss": 0.6777,
+      "step": 7127
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.8400563742313745,
+      "learning_rate": 1.8301422094122937e-06,
+      "loss": 0.6667,
+      "step": 7128
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.68660014272437,
+      "learning_rate": 1.830088341632332e-06,
+      "loss": 0.6882,
+      "step": 7129
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7481005985080551,
+      "learning_rate": 1.8300344661050822e-06,
+      "loss": 0.6654,
+      "step": 7130
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7023636061915226,
+      "learning_rate": 1.8299805828310472e-06,
+      "loss": 0.6953,
+      "step": 7131
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.018536929551939,
+      "learning_rate": 1.8299266918107304e-06,
+      "loss": 0.6764,
+      "step": 7132
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.664367336900198,
+      "learning_rate": 1.8298727930446343e-06,
+      "loss": 0.6439,
+      "step": 7133
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.0819923613714475,
+      "learning_rate": 1.8298188865332618e-06,
+      "loss": 0.6595,
+      "step": 7134
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.0883833333567643,
+      "learning_rate": 1.8297649722771164e-06,
+      "loss": 0.6797,
+      "step": 7135
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6726029025847993,
+      "learning_rate": 1.8297110502767012e-06,
+      "loss": 0.6992,
+      "step": 7136
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 4.98475117547987,
+      "learning_rate": 1.8296571205325196e-06,
+      "loss": 0.6536,
+      "step": 7137
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.064657017056068,
+      "learning_rate": 1.8296031830450743e-06,
+      "loss": 0.6947,
+      "step": 7138
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 5.930640029110142,
+      "learning_rate": 1.8295492378148692e-06,
+      "loss": 0.679,
+      "step": 7139
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 7.895001787762332,
+      "learning_rate": 1.829495284842408e-06,
+      "loss": 0.6882,
+      "step": 7140
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8725393644923981,
+      "learning_rate": 1.8294413241281936e-06,
+      "loss": 0.6973,
+      "step": 7141
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.122493286024838,
+      "learning_rate": 1.8293873556727299e-06,
+      "loss": 0.668,
+      "step": 7142
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2553465094330782,
+      "learning_rate": 1.829333379476521e-06,
+      "loss": 0.6888,
+      "step": 7143
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.393837768572665,
+      "learning_rate": 1.82927939554007e-06,
+      "loss": 0.6549,
+      "step": 7144
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7231988229694165,
+      "learning_rate": 1.8292254038638814e-06,
+      "loss": 0.6543,
+      "step": 7145
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.811848895045366,
+      "learning_rate": 1.8291714044484587e-06,
+      "loss": 0.6751,
+      "step": 7146
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9250192571570942,
+      "learning_rate": 1.829117397294306e-06,
+      "loss": 0.6445,
+      "step": 7147
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.780615651245685,
+      "learning_rate": 1.829063382401927e-06,
+      "loss": 0.6888,
+      "step": 7148
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.308465232190144,
+      "learning_rate": 1.8290093597718265e-06,
+      "loss": 0.653,
+      "step": 7149
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.1299893450139793,
+      "learning_rate": 1.8289553294045082e-06,
+      "loss": 0.6667,
+      "step": 7150
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.592146774197155,
+      "learning_rate": 1.8289012913004761e-06,
+      "loss": 0.681,
+      "step": 7151
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.0895376568318818,
+      "learning_rate": 1.8288472454602357e-06,
+      "loss": 0.6966,
+      "step": 7152
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.448997777537957,
+      "learning_rate": 1.8287931918842903e-06,
+      "loss": 0.6725,
+      "step": 7153
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.490837770199781,
+      "learning_rate": 1.8287391305731447e-06,
+      "loss": 0.6764,
+      "step": 7154
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7827388865572432,
+      "learning_rate": 1.8286850615273033e-06,
+      "loss": 0.6576,
+      "step": 7155
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.507032510163258,
+      "learning_rate": 1.8286309847472712e-06,
+      "loss": 0.6797,
+      "step": 7156
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7288834014024398,
+      "learning_rate": 1.8285769002335528e-06,
+      "loss": 0.6536,
+      "step": 7157
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8462799790474498,
+      "learning_rate": 1.828522807986653e-06,
+      "loss": 0.6777,
+      "step": 7158
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0347034584825534,
+      "learning_rate": 1.8284687080070765e-06,
+      "loss": 0.6901,
+      "step": 7159
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.8856747078086675,
+      "learning_rate": 1.828414600295328e-06,
+      "loss": 0.681,
+      "step": 7160
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0938659917803564,
+      "learning_rate": 1.8283604848519133e-06,
+      "loss": 0.7109,
+      "step": 7161
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.313004380069963,
+      "learning_rate": 1.8283063616773365e-06,
+      "loss": 0.6647,
+      "step": 7162
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 6.399306801123899,
+      "learning_rate": 1.8282522307721036e-06,
+      "loss": 0.6908,
+      "step": 7163
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8794713219587187,
+      "learning_rate": 1.8281980921367187e-06,
+      "loss": 0.707,
+      "step": 7164
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7911568987651573,
+      "learning_rate": 1.8281439457716883e-06,
+      "loss": 0.6842,
+      "step": 7165
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.367730568527076,
+      "learning_rate": 1.8280897916775171e-06,
+      "loss": 0.7012,
+      "step": 7166
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.079620023988491,
+      "learning_rate": 1.8280356298547107e-06,
+      "loss": 0.6751,
+      "step": 7167
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2288528031954034,
+      "learning_rate": 1.8279814603037744e-06,
+      "loss": 0.6569,
+      "step": 7168
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.191628126923919,
+      "learning_rate": 1.8279272830252141e-06,
+      "loss": 0.6719,
+      "step": 7169
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.3774048713310256,
+      "learning_rate": 1.827873098019535e-06,
+      "loss": 0.6751,
+      "step": 7170
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9328467261113137,
+      "learning_rate": 1.8278189052872434e-06,
+      "loss": 0.681,
+      "step": 7171
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.011964023855743,
+      "learning_rate": 1.8277647048288446e-06,
+      "loss": 0.694,
+      "step": 7172
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4626937807322165,
+      "learning_rate": 1.8277104966448445e-06,
+      "loss": 0.6621,
+      "step": 7173
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2838270588967253,
+      "learning_rate": 1.8276562807357492e-06,
+      "loss": 0.6829,
+      "step": 7174
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.493320112359465,
+      "learning_rate": 1.8276020571020645e-06,
+      "loss": 0.6953,
+      "step": 7175
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 8.378856429860503,
+      "learning_rate": 1.827547825744297e-06,
+      "loss": 0.7246,
+      "step": 7176
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9105295387153779,
+      "learning_rate": 1.827493586662952e-06,
+      "loss": 0.6908,
+      "step": 7177
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5124779298081124,
+      "learning_rate": 1.8274393398585364e-06,
+      "loss": 0.6771,
+      "step": 7178
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.3978326503351686,
+      "learning_rate": 1.8273850853315563e-06,
+      "loss": 0.6784,
+      "step": 7179
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.628863490098975,
+      "learning_rate": 1.827330823082518e-06,
+      "loss": 0.6816,
+      "step": 7180
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.820554613126918,
+      "learning_rate": 1.8272765531119278e-06,
+      "loss": 0.6855,
+      "step": 7181
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2747043616282467,
+      "learning_rate": 1.8272222754202925e-06,
+      "loss": 0.6901,
+      "step": 7182
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.3651217706141745,
+      "learning_rate": 1.8271679900081188e-06,
+      "loss": 0.6771,
+      "step": 7183
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.41804721840154,
+      "learning_rate": 1.8271136968759127e-06,
+      "loss": 0.6595,
+      "step": 7184
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.4776630275436415,
+      "learning_rate": 1.8270593960241815e-06,
+      "loss": 0.6751,
+      "step": 7185
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.0457857723017736,
+      "learning_rate": 1.8270050874534317e-06,
+      "loss": 0.6953,
+      "step": 7186
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0117875184710476,
+      "learning_rate": 1.8269507711641703e-06,
+      "loss": 0.6569,
+      "step": 7187
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.397715053551542,
+      "learning_rate": 1.8268964471569043e-06,
+      "loss": 0.6829,
+      "step": 7188
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 7.44298530613423,
+      "learning_rate": 1.8268421154321408e-06,
+      "loss": 0.6927,
+      "step": 7189
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.6569572557851524,
+      "learning_rate": 1.8267877759903865e-06,
+      "loss": 0.6947,
+      "step": 7190
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.821664258323817,
+      "learning_rate": 1.8267334288321488e-06,
+      "loss": 0.694,
+      "step": 7191
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0331601738282739,
+      "learning_rate": 1.8266790739579348e-06,
+      "loss": 0.6797,
+      "step": 7192
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4569526445610297,
+      "learning_rate": 1.826624711368252e-06,
+      "loss": 0.7057,
+      "step": 7193
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.6579171583529515,
+      "learning_rate": 1.8265703410636078e-06,
+      "loss": 0.6908,
+      "step": 7194
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9853346568440725,
+      "learning_rate": 1.826515963044509e-06,
+      "loss": 0.6882,
+      "step": 7195
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.452527100848695,
+      "learning_rate": 1.8264615773114642e-06,
+      "loss": 0.6895,
+      "step": 7196
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.29990569452902,
+      "learning_rate": 1.8264071838649802e-06,
+      "loss": 0.6712,
+      "step": 7197
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.9880747729484978,
+      "learning_rate": 1.826352782705565e-06,
+      "loss": 0.6751,
+      "step": 7198
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.1322188464290015,
+      "learning_rate": 1.8262983738337262e-06,
+      "loss": 0.6732,
+      "step": 7199
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.8162466579494263,
+      "learning_rate": 1.8262439572499714e-06,
+      "loss": 0.6868,
+      "step": 7200
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7012387563637887,
+      "learning_rate": 1.8261895329548085e-06,
+      "loss": 0.6758,
+      "step": 7201
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0074085419752015,
+      "learning_rate": 1.826135100948746e-06,
+      "loss": 0.6829,
+      "step": 7202
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.641593131293144,
+      "learning_rate": 1.8260806612322914e-06,
+      "loss": 0.6868,
+      "step": 7203
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.4831365053489164,
+      "learning_rate": 1.826026213805953e-06,
+      "loss": 0.6732,
+      "step": 7204
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.511460862267005,
+      "learning_rate": 1.8259717586702388e-06,
+      "loss": 0.6764,
+      "step": 7205
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 5.176765971464415,
+      "learning_rate": 1.8259172958256571e-06,
+      "loss": 0.6764,
+      "step": 7206
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.5003229581766684,
+      "learning_rate": 1.8258628252727163e-06,
+      "loss": 0.6758,
+      "step": 7207
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6432396051821214,
+      "learning_rate": 1.8258083470119248e-06,
+      "loss": 0.6719,
+      "step": 7208
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.550932963129105,
+      "learning_rate": 1.825753861043791e-06,
+      "loss": 0.6849,
+      "step": 7209
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7549553789483441,
+      "learning_rate": 1.8256993673688232e-06,
+      "loss": 0.6686,
+      "step": 7210
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.2692915580637507,
+      "learning_rate": 1.8256448659875302e-06,
+      "loss": 0.679,
+      "step": 7211
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1828980392023642,
+      "learning_rate": 1.8255903569004207e-06,
+      "loss": 0.6862,
+      "step": 7212
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.916138136774664,
+      "learning_rate": 1.8255358401080034e-06,
+      "loss": 0.6823,
+      "step": 7213
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.262551025677481,
+      "learning_rate": 1.825481315610787e-06,
+      "loss": 0.6914,
+      "step": 7214
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.762819860865685,
+      "learning_rate": 1.8254267834092806e-06,
+      "loss": 0.696,
+      "step": 7215
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.928085939323418,
+      "learning_rate": 1.8253722435039934e-06,
+      "loss": 0.6654,
+      "step": 7216
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.142880588442651,
+      "learning_rate": 1.8253176958954335e-06,
+      "loss": 0.6771,
+      "step": 7217
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8624308768891542,
+      "learning_rate": 1.8252631405841108e-06,
+      "loss": 0.6608,
+      "step": 7218
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.0004008257723553,
+      "learning_rate": 1.825208577570534e-06,
+      "loss": 0.6602,
+      "step": 7219
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.023849740757716,
+      "learning_rate": 1.8251540068552127e-06,
+      "loss": 0.6348,
+      "step": 7220
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.746580065802468,
+      "learning_rate": 1.8250994284386563e-06,
+      "loss": 0.7025,
+      "step": 7221
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.2842821155739337,
+      "learning_rate": 1.8250448423213738e-06,
+      "loss": 0.681,
+      "step": 7222
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2588069678947211,
+      "learning_rate": 1.8249902485038748e-06,
+      "loss": 0.6504,
+      "step": 7223
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8756941924562145,
+      "learning_rate": 1.8249356469866689e-06,
+      "loss": 0.6882,
+      "step": 7224
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.611007346884863,
+      "learning_rate": 1.8248810377702655e-06,
+      "loss": 0.6608,
+      "step": 7225
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6727869828445554,
+      "learning_rate": 1.8248264208551744e-06,
+      "loss": 0.6803,
+      "step": 7226
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.158336505348912,
+      "learning_rate": 1.8247717962419059e-06,
+      "loss": 0.6725,
+      "step": 7227
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8519612609706468,
+      "learning_rate": 1.8247171639309686e-06,
+      "loss": 0.6732,
+      "step": 7228
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4801638048264887,
+      "learning_rate": 1.8246625239228735e-06,
+      "loss": 0.6654,
+      "step": 7229
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1285689361748696,
+      "learning_rate": 1.8246078762181299e-06,
+      "loss": 0.6777,
+      "step": 7230
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3175490192092947,
+      "learning_rate": 1.8245532208172483e-06,
+      "loss": 0.6914,
+      "step": 7231
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.288671401336204,
+      "learning_rate": 1.8244985577207384e-06,
+      "loss": 0.6882,
+      "step": 7232
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.604287822433012,
+      "learning_rate": 1.8244438869291105e-06,
+      "loss": 0.696,
+      "step": 7233
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.7356461130024967,
+      "learning_rate": 1.8243892084428749e-06,
+      "loss": 0.6562,
+      "step": 7234
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.170473047683846,
+      "learning_rate": 1.824334522262542e-06,
+      "loss": 0.6576,
+      "step": 7235
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.035565481748346,
+      "learning_rate": 1.8242798283886222e-06,
+      "loss": 0.7005,
+      "step": 7236
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.514232294241727,
+      "learning_rate": 1.8242251268216257e-06,
+      "loss": 0.6719,
+      "step": 7237
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1134423136785534,
+      "learning_rate": 1.824170417562063e-06,
+      "loss": 0.6882,
+      "step": 7238
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.115282872323748,
+      "learning_rate": 1.8241157006104454e-06,
+      "loss": 0.6738,
+      "step": 7239
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7414741925828613,
+      "learning_rate": 1.8240609759672829e-06,
+      "loss": 0.681,
+      "step": 7240
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.390962461987225,
+      "learning_rate": 1.8240062436330863e-06,
+      "loss": 0.6803,
+      "step": 7241
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.3859121240853152,
+      "learning_rate": 1.8239515036083668e-06,
+      "loss": 0.7018,
+      "step": 7242
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8648658801407387,
+      "learning_rate": 1.8238967558936346e-06,
+      "loss": 0.6628,
+      "step": 7243
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7218604798839086,
+      "learning_rate": 1.8238420004894018e-06,
+      "loss": 0.6751,
+      "step": 7244
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6164395512634436,
+      "learning_rate": 1.8237872373961783e-06,
+      "loss": 0.6855,
+      "step": 7245
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8614405707937205,
+      "learning_rate": 1.8237324666144756e-06,
+      "loss": 0.6706,
+      "step": 7246
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.413588578963587,
+      "learning_rate": 1.8236776881448051e-06,
+      "loss": 0.6966,
+      "step": 7247
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.2265618151280435,
+      "learning_rate": 1.8236229019876779e-06,
+      "loss": 0.7246,
+      "step": 7248
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.4038043931386883,
+      "learning_rate": 1.8235681081436055e-06,
+      "loss": 0.6771,
+      "step": 7249
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.9489115739120588,
+      "learning_rate": 1.8235133066130985e-06,
+      "loss": 0.6738,
+      "step": 7250
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.43265597586846,
+      "learning_rate": 1.8234584973966697e-06,
+      "loss": 0.6914,
+      "step": 7251
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.827574654502057,
+      "learning_rate": 1.8234036804948294e-06,
+      "loss": 0.6973,
+      "step": 7252
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.688850582341655,
+      "learning_rate": 1.82334885590809e-06,
+      "loss": 0.6829,
+      "step": 7253
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.590475978032138,
+      "learning_rate": 1.8232940236369626e-06,
+      "loss": 0.6621,
+      "step": 7254
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.550958989934827,
+      "learning_rate": 1.82323918368196e-06,
+      "loss": 0.6908,
+      "step": 7255
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4356949457406927,
+      "learning_rate": 1.8231843360435922e-06,
+      "loss": 0.668,
+      "step": 7256
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.939032730613223,
+      "learning_rate": 1.8231294807223732e-06,
+      "loss": 0.6758,
+      "step": 7257
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.015201329416802,
+      "learning_rate": 1.8230746177188134e-06,
+      "loss": 0.7038,
+      "step": 7258
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.847811533322734,
+      "learning_rate": 1.8230197470334257e-06,
+      "loss": 0.6686,
+      "step": 7259
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.5890241856303056,
+      "learning_rate": 1.8229648686667219e-06,
+      "loss": 0.6797,
+      "step": 7260
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.99580732868676,
+      "learning_rate": 1.822909982619214e-06,
+      "loss": 0.6803,
+      "step": 7261
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5563389471011442,
+      "learning_rate": 1.8228550888914144e-06,
+      "loss": 0.6712,
+      "step": 7262
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8490884794071439,
+      "learning_rate": 1.8228001874838355e-06,
+      "loss": 0.6764,
+      "step": 7263
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 6.118003981602806,
+      "learning_rate": 1.82274527839699e-06,
+      "loss": 0.7038,
+      "step": 7264
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.5945271507390184,
+      "learning_rate": 1.8226903616313898e-06,
+      "loss": 0.6908,
+      "step": 7265
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 3.7151394878185116,
+      "learning_rate": 1.822635437187548e-06,
+      "loss": 0.6862,
+      "step": 7266
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.844451492519643,
+      "learning_rate": 1.8225805050659763e-06,
+      "loss": 0.6927,
+      "step": 7267
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 4.404662446899828,
+      "learning_rate": 1.8225255652671887e-06,
+      "loss": 0.6673,
+      "step": 7268
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 5.025830924132563,
+      "learning_rate": 1.8224706177916969e-06,
+      "loss": 0.6908,
+      "step": 7269
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.3792988807393225,
+      "learning_rate": 1.8224156626400143e-06,
+      "loss": 0.6921,
+      "step": 7270
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2575384476749831,
+      "learning_rate": 1.8223606998126533e-06,
+      "loss": 0.6589,
+      "step": 7271
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.8498914567682243,
+      "learning_rate": 1.8223057293101272e-06,
+      "loss": 0.6745,
+      "step": 7272
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6862255179012066,
+      "learning_rate": 1.822250751132949e-06,
+      "loss": 0.6621,
+      "step": 7273
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.885609041553421,
+      "learning_rate": 1.822195765281632e-06,
+      "loss": 0.6641,
+      "step": 7274
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.479952549554041,
+      "learning_rate": 1.8221407717566891e-06,
+      "loss": 0.6797,
+      "step": 7275
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 7.71083422965666,
+      "learning_rate": 1.8220857705586336e-06,
+      "loss": 0.7096,
+      "step": 7276
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.45208222561637,
+      "learning_rate": 1.8220307616879792e-06,
+      "loss": 0.696,
+      "step": 7277
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5181100455214666,
+      "learning_rate": 1.8219757451452387e-06,
+      "loss": 0.6771,
+      "step": 7278
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.245630340260268,
+      "learning_rate": 1.821920720930926e-06,
+      "loss": 0.6921,
+      "step": 7279
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6053442829314937,
+      "learning_rate": 1.8218656890455545e-06,
+      "loss": 0.6797,
+      "step": 7280
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.853241498775552,
+      "learning_rate": 1.8218106494896376e-06,
+      "loss": 0.6621,
+      "step": 7281
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.162077099395584,
+      "learning_rate": 1.8217556022636898e-06,
+      "loss": 0.6901,
+      "step": 7282
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8047995224498609,
+      "learning_rate": 1.821700547368224e-06,
+      "loss": 0.709,
+      "step": 7283
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7856592835133405,
+      "learning_rate": 1.821645484803754e-06,
+      "loss": 0.6699,
+      "step": 7284
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9020861139425744,
+      "learning_rate": 1.8215904145707946e-06,
+      "loss": 0.6706,
+      "step": 7285
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.9606914277161466,
+      "learning_rate": 1.821535336669859e-06,
+      "loss": 0.6751,
+      "step": 7286
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.548752009630068,
+      "learning_rate": 1.8214802511014613e-06,
+      "loss": 0.6536,
+      "step": 7287
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.010477509571199,
+      "learning_rate": 1.8214251578661157e-06,
+      "loss": 0.6686,
+      "step": 7288
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.420643835377131,
+      "learning_rate": 1.8213700569643367e-06,
+      "loss": 0.6725,
+      "step": 7289
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.193478600395327,
+      "learning_rate": 1.821314948396638e-06,
+      "loss": 0.7083,
+      "step": 7290
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.745861726908077,
+      "learning_rate": 1.8212598321635348e-06,
+      "loss": 0.6693,
+      "step": 7291
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 8.588415569582555,
+      "learning_rate": 1.8212047082655405e-06,
+      "loss": 0.7135,
+      "step": 7292
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6345385314792242,
+      "learning_rate": 1.8211495767031702e-06,
+      "loss": 0.6654,
+      "step": 7293
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.206962898158289,
+      "learning_rate": 1.821094437476938e-06,
+      "loss": 0.6784,
+      "step": 7294
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9622535647533401,
+      "learning_rate": 1.8210392905873594e-06,
+      "loss": 0.6745,
+      "step": 7295
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.1039316846726348,
+      "learning_rate": 1.8209841360349478e-06,
+      "loss": 0.6908,
+      "step": 7296
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.2427515911951277,
+      "learning_rate": 1.8209289738202195e-06,
+      "loss": 0.6615,
+      "step": 7297
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.2883550535264217,
+      "learning_rate": 1.8208738039436878e-06,
+      "loss": 0.6764,
+      "step": 7298
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.5696664602468267,
+      "learning_rate": 1.8208186264058686e-06,
+      "loss": 0.6992,
+      "step": 7299
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.446002163835734,
+      "learning_rate": 1.8207634412072764e-06,
+      "loss": 0.6895,
+      "step": 7300
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.904273261273821,
+      "learning_rate": 1.8207082483484264e-06,
+      "loss": 0.6901,
+      "step": 7301
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.158244697655491,
+      "learning_rate": 1.8206530478298338e-06,
+      "loss": 0.6934,
+      "step": 7302
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.498355288883931,
+      "learning_rate": 1.8205978396520139e-06,
+      "loss": 0.6868,
+      "step": 7303
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0112532393369063,
+      "learning_rate": 1.8205426238154815e-06,
+      "loss": 0.6797,
+      "step": 7304
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.195683819179623,
+      "learning_rate": 1.8204874003207521e-06,
+      "loss": 0.6973,
+      "step": 7305
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 6.079315500513975,
+      "learning_rate": 1.8204321691683417e-06,
+      "loss": 0.6986,
+      "step": 7306
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.452095793576954,
+      "learning_rate": 1.820376930358765e-06,
+      "loss": 0.6908,
+      "step": 7307
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.7622837628871335,
+      "learning_rate": 1.8203216838925378e-06,
+      "loss": 0.6719,
+      "step": 7308
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.703265467220527,
+      "learning_rate": 1.820266429770176e-06,
+      "loss": 0.6764,
+      "step": 7309
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.1653929371788205,
+      "learning_rate": 1.8202111679921952e-06,
+      "loss": 0.6732,
+      "step": 7310
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.6106249104559636,
+      "learning_rate": 1.8201558985591108e-06,
+      "loss": 0.6875,
+      "step": 7311
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.4644560459645546,
+      "learning_rate": 1.820100621471439e-06,
+      "loss": 0.6693,
+      "step": 7312
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.284947940980634,
+      "learning_rate": 1.8200453367296954e-06,
+      "loss": 0.6751,
+      "step": 7313
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.917079534103123,
+      "learning_rate": 1.8199900443343964e-06,
+      "loss": 0.694,
+      "step": 7314
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.446710306975443,
+      "learning_rate": 1.8199347442860576e-06,
+      "loss": 0.6829,
+      "step": 7315
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.5978505463378918,
+      "learning_rate": 1.8198794365851952e-06,
+      "loss": 0.6686,
+      "step": 7316
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7342930317697398,
+      "learning_rate": 1.8198241212323255e-06,
+      "loss": 0.6602,
+      "step": 7317
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 5.371602154211714,
+      "learning_rate": 1.8197687982279652e-06,
+      "loss": 0.7155,
+      "step": 7318
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 3.1732238280835503,
+      "learning_rate": 1.81971346757263e-06,
+      "loss": 0.6921,
+      "step": 7319
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8585707511386347,
+      "learning_rate": 1.8196581292668363e-06,
+      "loss": 0.6908,
+      "step": 7320
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.3372387155157546,
+      "learning_rate": 1.819602783311101e-06,
+      "loss": 0.666,
+      "step": 7321
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6535190569311565,
+      "learning_rate": 1.8195474297059402e-06,
+      "loss": 0.6628,
+      "step": 7322
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9156334101171626,
+      "learning_rate": 1.819492068451871e-06,
+      "loss": 0.6686,
+      "step": 7323
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.538964535660808,
+      "learning_rate": 1.8194366995494099e-06,
+      "loss": 0.6797,
+      "step": 7324
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.2253679602573568,
+      "learning_rate": 1.8193813229990736e-06,
+      "loss": 0.6777,
+      "step": 7325
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.432463853302201,
+      "learning_rate": 1.8193259388013789e-06,
+      "loss": 0.6842,
+      "step": 7326
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0414396318232162,
+      "learning_rate": 1.8192705469568427e-06,
+      "loss": 0.6849,
+      "step": 7327
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5467557681450665,
+      "learning_rate": 1.8192151474659823e-06,
+      "loss": 0.6634,
+      "step": 7328
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.943883252819132,
+      "learning_rate": 1.8191597403293139e-06,
+      "loss": 0.6738,
+      "step": 7329
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6340104576013441,
+      "learning_rate": 1.8191043255473557e-06,
+      "loss": 0.6732,
+      "step": 7330
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.535982142854951,
+      "learning_rate": 1.8190489031206244e-06,
+      "loss": 0.6556,
+      "step": 7331
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.3062476194407955,
+      "learning_rate": 1.818993473049637e-06,
+      "loss": 0.6751,
+      "step": 7332
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.577564985050159,
+      "learning_rate": 1.818938035334911e-06,
+      "loss": 0.694,
+      "step": 7333
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.491623922932199,
+      "learning_rate": 1.8188825899769641e-06,
+      "loss": 0.6829,
+      "step": 7334
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 4.298712607506904,
+      "learning_rate": 1.8188271369763135e-06,
+      "loss": 0.6803,
+      "step": 7335
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7644155072853887,
+      "learning_rate": 1.818771676333477e-06,
+      "loss": 0.6895,
+      "step": 7336
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.167106924107878,
+      "learning_rate": 1.818716208048972e-06,
+      "loss": 0.6901,
+      "step": 7337
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.5036572660005894,
+      "learning_rate": 1.818660732123316e-06,
+      "loss": 0.6491,
+      "step": 7338
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.9515413784541786,
+      "learning_rate": 1.818605248557027e-06,
+      "loss": 0.6999,
+      "step": 7339
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.714075746794432,
+      "learning_rate": 1.818549757350623e-06,
+      "loss": 0.6465,
+      "step": 7340
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.964469578102391,
+      "learning_rate": 1.818494258504622e-06,
+      "loss": 0.6751,
+      "step": 7341
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.4752674202094782,
+      "learning_rate": 1.8184387520195413e-06,
+      "loss": 0.6836,
+      "step": 7342
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.097073558943924,
+      "learning_rate": 1.8183832378958992e-06,
+      "loss": 0.6549,
+      "step": 7343
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.5708253841725948,
+      "learning_rate": 1.818327716134214e-06,
+      "loss": 0.6628,
+      "step": 7344
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0246153177139243,
+      "learning_rate": 1.818272186735004e-06,
+      "loss": 0.6367,
+      "step": 7345
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.9150234184610793,
+      "learning_rate": 1.8182166496987877e-06,
+      "loss": 0.6686,
+      "step": 7346
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.666075020224906,
+      "learning_rate": 1.8181611050260825e-06,
+      "loss": 0.6712,
+      "step": 7347
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8311192875195532,
+      "learning_rate": 1.8181055527174077e-06,
+      "loss": 0.6576,
+      "step": 7348
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5708408289516476,
+      "learning_rate": 1.818049992773281e-06,
+      "loss": 0.6901,
+      "step": 7349
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.4516844673964295,
+      "learning_rate": 1.817994425194222e-06,
+      "loss": 0.696,
+      "step": 7350
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.7551328238280623,
+      "learning_rate": 1.817938849980748e-06,
+      "loss": 0.6667,
+      "step": 7351
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.395360402281314,
+      "learning_rate": 1.817883267133379e-06,
+      "loss": 0.6712,
+      "step": 7352
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.3146416169021085,
+      "learning_rate": 1.8178276766526328e-06,
+      "loss": 0.6927,
+      "step": 7353
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.638615196137087,
+      "learning_rate": 1.8177720785390286e-06,
+      "loss": 0.6823,
+      "step": 7354
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.1186853260894107,
+      "learning_rate": 1.8177164727930854e-06,
+      "loss": 0.6576,
+      "step": 7355
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.192832104972316,
+      "learning_rate": 1.817660859415322e-06,
+      "loss": 0.6296,
+      "step": 7356
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0710196301529344,
+      "learning_rate": 1.8176052384062572e-06,
+      "loss": 0.6803,
+      "step": 7357
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0734618465058556,
+      "learning_rate": 1.8175496097664107e-06,
+      "loss": 0.6868,
+      "step": 7358
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8253790245483167,
+      "learning_rate": 1.8174939734963016e-06,
+      "loss": 0.6699,
+      "step": 7359
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.8235521334862264,
+      "learning_rate": 1.8174383295964484e-06,
+      "loss": 0.7064,
+      "step": 7360
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1513808931810323,
+      "learning_rate": 1.8173826780673713e-06,
+      "loss": 0.6569,
+      "step": 7361
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6464916350495298,
+      "learning_rate": 1.8173270189095892e-06,
+      "loss": 0.7096,
+      "step": 7362
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.657598591124363,
+      "learning_rate": 1.8172713521236221e-06,
+      "loss": 0.6549,
+      "step": 7363
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.884582368669517,
+      "learning_rate": 1.817215677709989e-06,
+      "loss": 0.6576,
+      "step": 7364
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.0030600759049453,
+      "learning_rate": 1.8171599956692096e-06,
+      "loss": 0.6732,
+      "step": 7365
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.157471651205081,
+      "learning_rate": 1.8171043060018039e-06,
+      "loss": 0.6582,
+      "step": 7366
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.1281737061516737,
+      "learning_rate": 1.8170486087082913e-06,
+      "loss": 0.6712,
+      "step": 7367
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.078446639876991,
+      "learning_rate": 1.8169929037891917e-06,
+      "loss": 0.6712,
+      "step": 7368
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7693363932531958,
+      "learning_rate": 1.8169371912450253e-06,
+      "loss": 0.6836,
+      "step": 7369
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8384220030824798,
+      "learning_rate": 1.8168814710763117e-06,
+      "loss": 0.6842,
+      "step": 7370
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.2582535172315374,
+      "learning_rate": 1.8168257432835712e-06,
+      "loss": 0.6849,
+      "step": 7371
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.567503854037778,
+      "learning_rate": 1.8167700078673237e-06,
+      "loss": 0.6712,
+      "step": 7372
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.866268917741038,
+      "learning_rate": 1.8167142648280895e-06,
+      "loss": 0.6882,
+      "step": 7373
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.558779836355342,
+      "learning_rate": 1.8166585141663888e-06,
+      "loss": 0.6543,
+      "step": 7374
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.610884515653038,
+      "learning_rate": 1.8166027558827422e-06,
+      "loss": 0.6855,
+      "step": 7375
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9037057453848019,
+      "learning_rate": 1.8165469899776698e-06,
+      "loss": 0.6855,
+      "step": 7376
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.6234508324458097,
+      "learning_rate": 1.8164912164516918e-06,
+      "loss": 0.6615,
+      "step": 7377
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5995290080951414,
+      "learning_rate": 1.8164354353053293e-06,
+      "loss": 0.668,
+      "step": 7378
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7468512962832143,
+      "learning_rate": 1.8163796465391025e-06,
+      "loss": 0.6738,
+      "step": 7379
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.638707163080237,
+      "learning_rate": 1.8163238501535325e-06,
+      "loss": 0.6816,
+      "step": 7380
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6717338891912779,
+      "learning_rate": 1.81626804614914e-06,
+      "loss": 0.6797,
+      "step": 7381
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.476124658814075,
+      "learning_rate": 1.8162122345264452e-06,
+      "loss": 0.6849,
+      "step": 7382
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.1714280601872489,
+      "learning_rate": 1.8161564152859696e-06,
+      "loss": 0.6738,
+      "step": 7383
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7211587233605898,
+      "learning_rate": 1.8161005884282341e-06,
+      "loss": 0.6738,
+      "step": 7384
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4193974005508254,
+      "learning_rate": 1.8160447539537596e-06,
+      "loss": 0.6569,
+      "step": 7385
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.8390843097495075,
+      "learning_rate": 1.815988911863067e-06,
+      "loss": 0.6693,
+      "step": 7386
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.63407455291123,
+      "learning_rate": 1.815933062156678e-06,
+      "loss": 0.6771,
+      "step": 7387
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 3.70146598519997,
+      "learning_rate": 1.8158772048351137e-06,
+      "loss": 0.6641,
+      "step": 7388
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.7910540403449544,
+      "learning_rate": 1.815821339898895e-06,
+      "loss": 0.6602,
+      "step": 7389
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.709365246314985,
+      "learning_rate": 1.8157654673485437e-06,
+      "loss": 0.6608,
+      "step": 7390
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.390079436104886,
+      "learning_rate": 1.815709587184581e-06,
+      "loss": 0.6569,
+      "step": 7391
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9941120198522088,
+      "learning_rate": 1.8156536994075286e-06,
+      "loss": 0.6406,
+      "step": 7392
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 5.312360633505529,
+      "learning_rate": 1.8155978040179083e-06,
+      "loss": 0.6947,
+      "step": 7393
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.755330680774744,
+      "learning_rate": 1.8155419010162415e-06,
+      "loss": 0.6478,
+      "step": 7394
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 11.35350342414453,
+      "learning_rate": 1.8154859904030499e-06,
+      "loss": 0.7194,
+      "step": 7395
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.4054499009692765,
+      "learning_rate": 1.8154300721788555e-06,
+      "loss": 0.6758,
+      "step": 7396
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.68690345114268,
+      "learning_rate": 1.8153741463441802e-06,
+      "loss": 0.6471,
+      "step": 7397
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.2409545132511568,
+      "learning_rate": 1.8153182128995458e-06,
+      "loss": 0.6589,
+      "step": 7398
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.209366698127717,
+      "learning_rate": 1.8152622718454747e-06,
+      "loss": 0.7051,
+      "step": 7399
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.6726476025321464,
+      "learning_rate": 1.8152063231824881e-06,
+      "loss": 0.6276,
+      "step": 7400
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.9724992978372485,
+      "learning_rate": 1.8151503669111095e-06,
+      "loss": 0.7025,
+      "step": 7401
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.246168017843598,
+      "learning_rate": 1.81509440303186e-06,
+      "loss": 0.6849,
+      "step": 7402
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.0657024843777365,
+      "learning_rate": 1.8150384315452624e-06,
+      "loss": 0.6543,
+      "step": 7403
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.2156204417498613,
+      "learning_rate": 1.8149824524518392e-06,
+      "loss": 0.7318,
+      "step": 7404
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6976943517183212,
+      "learning_rate": 1.8149264657521125e-06,
+      "loss": 0.6374,
+      "step": 7405
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 4.220025297707143,
+      "learning_rate": 1.8148704714466052e-06,
+      "loss": 0.679,
+      "step": 7406
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 7.297845838011607,
+      "learning_rate": 1.8148144695358397e-06,
+      "loss": 0.6634,
+      "step": 7407
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6374127442840511,
+      "learning_rate": 1.8147584600203386e-06,
+      "loss": 0.6816,
+      "step": 7408
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8232200118292949,
+      "learning_rate": 1.8147024429006247e-06,
+      "loss": 0.6595,
+      "step": 7409
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.071467748720222,
+      "learning_rate": 1.8146464181772207e-06,
+      "loss": 0.6771,
+      "step": 7410
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.306792533001335,
+      "learning_rate": 1.81459038585065e-06,
+      "loss": 0.6413,
+      "step": 7411
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.9861320149313504,
+      "learning_rate": 1.814534345921435e-06,
+      "loss": 0.7038,
+      "step": 7412
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 6.047620720223633,
+      "learning_rate": 1.8144782983900991e-06,
+      "loss": 0.6816,
+      "step": 7413
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9051101911509063,
+      "learning_rate": 1.814422243257165e-06,
+      "loss": 0.6784,
+      "step": 7414
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.783235014172375,
+      "learning_rate": 1.814366180523156e-06,
+      "loss": 0.6686,
+      "step": 7415
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7211469997115514,
+      "learning_rate": 1.8143101101885959e-06,
+      "loss": 0.6693,
+      "step": 7416
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.211111660702086,
+      "learning_rate": 1.814254032254007e-06,
+      "loss": 0.6615,
+      "step": 7417
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.501408604671979,
+      "learning_rate": 1.8141979467199135e-06,
+      "loss": 0.6647,
+      "step": 7418
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.7506102500558174,
+      "learning_rate": 1.8141418535868384e-06,
+      "loss": 0.6595,
+      "step": 7419
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.854502921454748,
+      "learning_rate": 1.8140857528553056e-06,
+      "loss": 0.7038,
+      "step": 7420
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.795361278944093,
+      "learning_rate": 1.8140296445258381e-06,
+      "loss": 0.6947,
+      "step": 7421
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.0318837588760323,
+      "learning_rate": 1.8139735285989603e-06,
+      "loss": 0.6751,
+      "step": 7422
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.228939163648049,
+      "learning_rate": 1.8139174050751956e-06,
+      "loss": 0.6602,
+      "step": 7423
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6799274933346187,
+      "learning_rate": 1.8138612739550675e-06,
+      "loss": 0.6764,
+      "step": 7424
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.2009593439566362,
+      "learning_rate": 1.8138051352391002e-06,
+      "loss": 0.6667,
+      "step": 7425
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.539075810977762,
+      "learning_rate": 1.8137489889278178e-06,
+      "loss": 0.7044,
+      "step": 7426
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.8631749442845145,
+      "learning_rate": 1.8136928350217438e-06,
+      "loss": 0.6992,
+      "step": 7427
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.594575398297396,
+      "learning_rate": 1.8136366735214028e-06,
+      "loss": 0.681,
+      "step": 7428
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.074182313175987,
+      "learning_rate": 1.813580504427319e-06,
+      "loss": 0.6862,
+      "step": 7429
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.874297670247067,
+      "learning_rate": 1.813524327740016e-06,
+      "loss": 0.6543,
+      "step": 7430
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7862888374147761,
+      "learning_rate": 1.8134681434600188e-06,
+      "loss": 0.6908,
+      "step": 7431
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.559298502436818,
+      "learning_rate": 1.8134119515878512e-06,
+      "loss": 0.6784,
+      "step": 7432
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.154031563479754,
+      "learning_rate": 1.813355752124038e-06,
+      "loss": 0.6595,
+      "step": 7433
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.647130323849457,
+      "learning_rate": 1.8132995450691037e-06,
+      "loss": 0.6823,
+      "step": 7434
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.044757567970662,
+      "learning_rate": 1.8132433304235731e-06,
+      "loss": 0.653,
+      "step": 7435
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.738758878940332,
+      "learning_rate": 1.8131871081879703e-06,
+      "loss": 0.6868,
+      "step": 7436
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.2970145628911456,
+      "learning_rate": 1.8131308783628203e-06,
+      "loss": 0.6706,
+      "step": 7437
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.589299039594592,
+      "learning_rate": 1.8130746409486479e-06,
+      "loss": 0.6738,
+      "step": 7438
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.887556779427573,
+      "learning_rate": 1.8130183959459782e-06,
+      "loss": 0.6608,
+      "step": 7439
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.171406718979038,
+      "learning_rate": 1.8129621433553355e-06,
+      "loss": 0.7012,
+      "step": 7440
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.0479817299387615,
+      "learning_rate": 1.8129058831772456e-06,
+      "loss": 0.6771,
+      "step": 7441
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3350695161433777,
+      "learning_rate": 1.812849615412233e-06,
+      "loss": 0.6908,
+      "step": 7442
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.109990288868045,
+      "learning_rate": 1.8127933400608231e-06,
+      "loss": 0.6712,
+      "step": 7443
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.022849046992747,
+      "learning_rate": 1.812737057123541e-06,
+      "loss": 0.6673,
+      "step": 7444
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.2017391629570933,
+      "learning_rate": 1.812680766600912e-06,
+      "loss": 0.6758,
+      "step": 7445
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 5.822878484226666,
+      "learning_rate": 1.8126244684934618e-06,
+      "loss": 0.6719,
+      "step": 7446
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.280064777244867,
+      "learning_rate": 1.8125681628017153e-06,
+      "loss": 0.6589,
+      "step": 7447
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9139561040483585,
+      "learning_rate": 1.8125118495261984e-06,
+      "loss": 0.6745,
+      "step": 7448
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.084579442532635,
+      "learning_rate": 1.8124555286674368e-06,
+      "loss": 0.6986,
+      "step": 7449
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.5210945476900135,
+      "learning_rate": 1.8123992002259556e-06,
+      "loss": 0.709,
+      "step": 7450
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.164327252449194,
+      "learning_rate": 1.8123428642022808e-06,
+      "loss": 0.7148,
+      "step": 7451
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9361740831176277,
+      "learning_rate": 1.8122865205969384e-06,
+      "loss": 0.6582,
+      "step": 7452
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.317850487711159,
+      "learning_rate": 1.812230169410454e-06,
+      "loss": 0.6921,
+      "step": 7453
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.386036967297382,
+      "learning_rate": 1.8121738106433536e-06,
+      "loss": 0.6842,
+      "step": 7454
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.063789407138934,
+      "learning_rate": 1.8121174442961631e-06,
+      "loss": 0.6693,
+      "step": 7455
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8900427268545426,
+      "learning_rate": 1.8120610703694089e-06,
+      "loss": 0.6751,
+      "step": 7456
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.5727328616383436,
+      "learning_rate": 1.8120046888636166e-06,
+      "loss": 0.6589,
+      "step": 7457
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.4590570848815785,
+      "learning_rate": 1.811948299779313e-06,
+      "loss": 0.666,
+      "step": 7458
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.3519682754377125,
+      "learning_rate": 1.8118919031170239e-06,
+      "loss": 0.6699,
+      "step": 7459
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 7.002936151289689,
+      "learning_rate": 1.8118354988772762e-06,
+      "loss": 0.6855,
+      "step": 7460
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.6817965871287766,
+      "learning_rate": 1.8117790870605955e-06,
+      "loss": 0.6732,
+      "step": 7461
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8578749115726428,
+      "learning_rate": 1.8117226676675093e-06,
+      "loss": 0.6641,
+      "step": 7462
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8394723975080728,
+      "learning_rate": 1.8116662406985433e-06,
+      "loss": 0.6699,
+      "step": 7463
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9692659444234067,
+      "learning_rate": 1.8116098061542245e-06,
+      "loss": 0.6562,
+      "step": 7464
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.6630423928603166,
+      "learning_rate": 1.8115533640350798e-06,
+      "loss": 0.6764,
+      "step": 7465
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.0906300264421864,
+      "learning_rate": 1.8114969143416357e-06,
+      "loss": 0.6797,
+      "step": 7466
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7001420991698185,
+      "learning_rate": 1.8114404570744193e-06,
+      "loss": 0.6823,
+      "step": 7467
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.140641470927915,
+      "learning_rate": 1.811383992233957e-06,
+      "loss": 0.6732,
+      "step": 7468
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6957741612706475,
+      "learning_rate": 1.8113275198207765e-06,
+      "loss": 0.6693,
+      "step": 7469
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.006250481069956,
+      "learning_rate": 1.8112710398354042e-06,
+      "loss": 0.6647,
+      "step": 7470
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.374998924345669,
+      "learning_rate": 1.8112145522783674e-06,
+      "loss": 0.679,
+      "step": 7471
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.4614748328545484,
+      "learning_rate": 1.811158057150194e-06,
+      "loss": 0.7077,
+      "step": 7472
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8188884610326317,
+      "learning_rate": 1.8111015544514104e-06,
+      "loss": 0.6797,
+      "step": 7473
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.8168562722041752,
+      "learning_rate": 1.8110450441825443e-06,
+      "loss": 0.6797,
+      "step": 7474
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.3744881897208594,
+      "learning_rate": 1.8109885263441232e-06,
+      "loss": 0.6855,
+      "step": 7475
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 3.1591363163093504,
+      "learning_rate": 1.8109320009366742e-06,
+      "loss": 0.6816,
+      "step": 7476
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.188685326346386,
+      "learning_rate": 1.8108754679607255e-06,
+      "loss": 0.6777,
+      "step": 7477
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.129483610433367,
+      "learning_rate": 1.810818927416804e-06,
+      "loss": 0.6979,
+      "step": 7478
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.2074322934501134,
+      "learning_rate": 1.8107623793054376e-06,
+      "loss": 0.6888,
+      "step": 7479
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.50012326733529,
+      "learning_rate": 1.8107058236271546e-06,
+      "loss": 0.6576,
+      "step": 7480
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6885374358271288,
+      "learning_rate": 1.8106492603824824e-06,
+      "loss": 0.6543,
+      "step": 7481
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.17685301066144,
+      "learning_rate": 1.8105926895719488e-06,
+      "loss": 0.6387,
+      "step": 7482
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7838102199623187,
+      "learning_rate": 1.8105361111960819e-06,
+      "loss": 0.6413,
+      "step": 7483
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.635143733535203,
+      "learning_rate": 1.81047952525541e-06,
+      "loss": 0.6797,
+      "step": 7484
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8647866344980731,
+      "learning_rate": 1.8104229317504612e-06,
+      "loss": 0.6862,
+      "step": 7485
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.3076419574708238,
+      "learning_rate": 1.810366330681763e-06,
+      "loss": 0.707,
+      "step": 7486
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.326463229573156,
+      "learning_rate": 1.8103097220498445e-06,
+      "loss": 0.6719,
+      "step": 7487
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.0644750783504575,
+      "learning_rate": 1.8102531058552338e-06,
+      "loss": 0.6738,
+      "step": 7488
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8610392402938268,
+      "learning_rate": 1.8101964820984587e-06,
+      "loss": 0.6621,
+      "step": 7489
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1237346262859065,
+      "learning_rate": 1.8101398507800485e-06,
+      "loss": 0.6589,
+      "step": 7490
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.84637333903065,
+      "learning_rate": 1.8100832119005314e-06,
+      "loss": 0.6947,
+      "step": 7491
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5983043534674333,
+      "learning_rate": 1.8100265654604363e-06,
+      "loss": 0.6751,
+      "step": 7492
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.770662429888964,
+      "learning_rate": 1.8099699114602913e-06,
+      "loss": 0.6901,
+      "step": 7493
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.147796203883097,
+      "learning_rate": 1.8099132499006255e-06,
+      "loss": 0.6608,
+      "step": 7494
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.5235746980898215,
+      "learning_rate": 1.809856580781968e-06,
+      "loss": 0.6966,
+      "step": 7495
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.3028747386039123,
+      "learning_rate": 1.8097999041048473e-06,
+      "loss": 0.668,
+      "step": 7496
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8327062366379931,
+      "learning_rate": 1.809743219869792e-06,
+      "loss": 0.668,
+      "step": 7497
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.0063478943933934,
+      "learning_rate": 1.809686528077332e-06,
+      "loss": 0.6667,
+      "step": 7498
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8835783369446857,
+      "learning_rate": 1.8096298287279962e-06,
+      "loss": 0.7038,
+      "step": 7499
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.1117979636322437,
+      "learning_rate": 1.8095731218223136e-06,
+      "loss": 0.696,
+      "step": 7500
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.843536497705405,
+      "learning_rate": 1.8095164073608133e-06,
+      "loss": 0.6986,
+      "step": 7501
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.8691561988045935,
+      "learning_rate": 1.809459685344025e-06,
+      "loss": 0.6693,
+      "step": 7502
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.3558886598279996,
+      "learning_rate": 1.8094029557724775e-06,
+      "loss": 0.6641,
+      "step": 7503
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.2727315181331615,
+      "learning_rate": 1.8093462186467007e-06,
+      "loss": 0.7161,
+      "step": 7504
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.002513491050286,
+      "learning_rate": 1.8092894739672242e-06,
+      "loss": 0.6764,
+      "step": 7505
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.786840256765564,
+      "learning_rate": 1.8092327217345774e-06,
+      "loss": 0.6921,
+      "step": 7506
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6146775006896171,
+      "learning_rate": 1.80917596194929e-06,
+      "loss": 0.6823,
+      "step": 7507
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.103851086018532,
+      "learning_rate": 1.809119194611892e-06,
+      "loss": 0.6875,
+      "step": 7508
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5061712704456871,
+      "learning_rate": 1.8090624197229127e-06,
+      "loss": 0.6745,
+      "step": 7509
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9693285248173811,
+      "learning_rate": 1.8090056372828827e-06,
+      "loss": 0.6927,
+      "step": 7510
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.865292122453303,
+      "learning_rate": 1.8089488472923314e-06,
+      "loss": 0.7116,
+      "step": 7511
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.0181438895280626,
+      "learning_rate": 1.8088920497517886e-06,
+      "loss": 0.6686,
+      "step": 7512
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.228773634596141,
+      "learning_rate": 1.8088352446617852e-06,
+      "loss": 0.6875,
+      "step": 7513
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.746568875911184,
+      "learning_rate": 1.8087784320228507e-06,
+      "loss": 0.679,
+      "step": 7514
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4726451854121043,
+      "learning_rate": 1.8087216118355157e-06,
+      "loss": 0.6927,
+      "step": 7515
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.2567364288555065,
+      "learning_rate": 1.8086647841003102e-06,
+      "loss": 0.6549,
+      "step": 7516
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.227911099152864,
+      "learning_rate": 1.808607948817765e-06,
+      "loss": 0.6908,
+      "step": 7517
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.174689050145776,
+      "learning_rate": 1.80855110598841e-06,
+      "loss": 0.6882,
+      "step": 7518
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.480781010628385,
+      "learning_rate": 1.8084942556127762e-06,
+      "loss": 0.6797,
+      "step": 7519
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8322838469324212,
+      "learning_rate": 1.808437397691394e-06,
+      "loss": 0.6947,
+      "step": 7520
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.14145841750839,
+      "learning_rate": 1.8083805322247943e-06,
+      "loss": 0.6797,
+      "step": 7521
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 4.690988987277324,
+      "learning_rate": 1.8083236592135074e-06,
+      "loss": 0.6888,
+      "step": 7522
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8431280863966504,
+      "learning_rate": 1.808266778658064e-06,
+      "loss": 0.6823,
+      "step": 7523
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.8229386395718508,
+      "learning_rate": 1.8082098905589956e-06,
+      "loss": 0.6771,
+      "step": 7524
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4838792494698287,
+      "learning_rate": 1.808152994916833e-06,
+      "loss": 0.6634,
+      "step": 7525
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5838489007897552,
+      "learning_rate": 1.8080960917321067e-06,
+      "loss": 0.6738,
+      "step": 7526
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7101960323037394,
+      "learning_rate": 1.8080391810053482e-06,
+      "loss": 0.679,
+      "step": 7527
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5264364370657599,
+      "learning_rate": 1.8079822627370886e-06,
+      "loss": 0.6699,
+      "step": 7528
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 6.148951034759593,
+      "learning_rate": 1.807925336927859e-06,
+      "loss": 0.6699,
+      "step": 7529
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.333525885722905,
+      "learning_rate": 1.8078684035781908e-06,
+      "loss": 0.6725,
+      "step": 7530
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 5.603991033669679,
+      "learning_rate": 1.8078114626886155e-06,
+      "loss": 0.668,
+      "step": 7531
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.437984005609779,
+      "learning_rate": 1.8077545142596641e-06,
+      "loss": 0.6758,
+      "step": 7532
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9099102754128967,
+      "learning_rate": 1.8076975582918688e-06,
+      "loss": 0.6562,
+      "step": 7533
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9225076332537238,
+      "learning_rate": 1.8076405947857604e-06,
+      "loss": 0.666,
+      "step": 7534
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.609217843738097,
+      "learning_rate": 1.8075836237418711e-06,
+      "loss": 0.6986,
+      "step": 7535
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.131980565189938,
+      "learning_rate": 1.8075266451607326e-06,
+      "loss": 0.6764,
+      "step": 7536
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.083496758177941,
+      "learning_rate": 1.807469659042876e-06,
+      "loss": 0.7181,
+      "step": 7537
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.607767993055828,
+      "learning_rate": 1.807412665388834e-06,
+      "loss": 0.6634,
+      "step": 7538
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5955013008143215,
+      "learning_rate": 1.807355664199138e-06,
+      "loss": 0.6855,
+      "step": 7539
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.929734350664788,
+      "learning_rate": 1.8072986554743204e-06,
+      "loss": 0.6712,
+      "step": 7540
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.377912137628293,
+      "learning_rate": 1.8072416392149128e-06,
+      "loss": 0.6947,
+      "step": 7541
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9136570524474887,
+      "learning_rate": 1.8071846154214478e-06,
+      "loss": 0.6693,
+      "step": 7542
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.137962851604011,
+      "learning_rate": 1.8071275840944574e-06,
+      "loss": 0.6921,
+      "step": 7543
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.97197551177776,
+      "learning_rate": 1.807070545234474e-06,
+      "loss": 0.6471,
+      "step": 7544
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.935983874036902,
+      "learning_rate": 1.8070134988420294e-06,
+      "loss": 0.6647,
+      "step": 7545
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.434261854365477,
+      "learning_rate": 1.8069564449176569e-06,
+      "loss": 0.6751,
+      "step": 7546
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.222903349115411,
+      "learning_rate": 1.8068993834618881e-06,
+      "loss": 0.6589,
+      "step": 7547
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.898999451098594,
+      "learning_rate": 1.8068423144752562e-06,
+      "loss": 0.6504,
+      "step": 7548
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.669774827320959,
+      "learning_rate": 1.8067852379582938e-06,
+      "loss": 0.6927,
+      "step": 7549
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.606608500213705,
+      "learning_rate": 1.8067281539115334e-06,
+      "loss": 0.6914,
+      "step": 7550
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.2849575857037947,
+      "learning_rate": 1.8066710623355075e-06,
+      "loss": 0.6549,
+      "step": 7551
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.195914983342289,
+      "learning_rate": 1.8066139632307493e-06,
+      "loss": 0.6751,
+      "step": 7552
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.251206976640443,
+      "learning_rate": 1.8065568565977918e-06,
+      "loss": 0.6784,
+      "step": 7553
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.6825415500575036,
+      "learning_rate": 1.8064997424371678e-06,
+      "loss": 0.6882,
+      "step": 7554
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.147696510597642,
+      "learning_rate": 1.8064426207494105e-06,
+      "loss": 0.6419,
+      "step": 7555
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.7200530733070782,
+      "learning_rate": 1.8063854915350527e-06,
+      "loss": 0.6719,
+      "step": 7556
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8209595675097672,
+      "learning_rate": 1.806328354794628e-06,
+      "loss": 0.6868,
+      "step": 7557
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.201186403860847,
+      "learning_rate": 1.8062712105286692e-06,
+      "loss": 0.6608,
+      "step": 7558
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.979132480935657,
+      "learning_rate": 1.80621405873771e-06,
+      "loss": 0.6797,
+      "step": 7559
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.088505957815975,
+      "learning_rate": 1.8061568994222838e-06,
+      "loss": 0.6589,
+      "step": 7560
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.049429410140051,
+      "learning_rate": 1.8060997325829237e-06,
+      "loss": 0.694,
+      "step": 7561
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.386192162871459,
+      "learning_rate": 1.8060425582201637e-06,
+      "loss": 0.679,
+      "step": 7562
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1252765526389015,
+      "learning_rate": 1.8059853763345374e-06,
+      "loss": 0.6673,
+      "step": 7563
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 7.273206086870948,
+      "learning_rate": 1.805928186926578e-06,
+      "loss": 0.6797,
+      "step": 7564
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5946057621225835,
+      "learning_rate": 1.8058709899968196e-06,
+      "loss": 0.6842,
+      "step": 7565
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.665616878193305,
+      "learning_rate": 1.805813785545796e-06,
+      "loss": 0.6758,
+      "step": 7566
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.6577070922644357,
+      "learning_rate": 1.8057565735740413e-06,
+      "loss": 0.7181,
+      "step": 7567
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8365990366036306,
+      "learning_rate": 1.805699354082089e-06,
+      "loss": 0.6921,
+      "step": 7568
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.25919253816712,
+      "learning_rate": 1.8056421270704733e-06,
+      "loss": 0.6745,
+      "step": 7569
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9614005393781139,
+      "learning_rate": 1.8055848925397285e-06,
+      "loss": 0.6803,
+      "step": 7570
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.8710553908608563,
+      "learning_rate": 1.8055276504903887e-06,
+      "loss": 0.681,
+      "step": 7571
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.085116015438871,
+      "learning_rate": 1.8054704009229882e-06,
+      "loss": 0.6517,
+      "step": 7572
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.9629052290813893,
+      "learning_rate": 1.805413143838061e-06,
+      "loss": 0.6797,
+      "step": 7573
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5520019394998228,
+      "learning_rate": 1.8053558792361418e-06,
+      "loss": 0.6895,
+      "step": 7574
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.25787238732106,
+      "learning_rate": 1.805298607117765e-06,
+      "loss": 0.6849,
+      "step": 7575
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.85942339000292,
+      "learning_rate": 1.805241327483465e-06,
+      "loss": 0.6771,
+      "step": 7576
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.1769526921093,
+      "learning_rate": 1.8051840403337763e-06,
+      "loss": 0.6764,
+      "step": 7577
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.1661644450370683,
+      "learning_rate": 1.8051267456692342e-06,
+      "loss": 0.6855,
+      "step": 7578
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.062424552757287,
+      "learning_rate": 1.8050694434903729e-06,
+      "loss": 0.6882,
+      "step": 7579
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.078536836145203,
+      "learning_rate": 1.8050121337977269e-06,
+      "loss": 0.7038,
+      "step": 7580
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.2800640099616403,
+      "learning_rate": 1.8049548165918317e-06,
+      "loss": 0.7103,
+      "step": 7581
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.710676898316017,
+      "learning_rate": 1.8048974918732221e-06,
+      "loss": 0.6706,
+      "step": 7582
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5475756222666774,
+      "learning_rate": 1.804840159642433e-06,
+      "loss": 0.6738,
+      "step": 7583
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5592494140925217,
+      "learning_rate": 1.8047828198999993e-06,
+      "loss": 0.6751,
+      "step": 7584
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.721883561216686,
+      "learning_rate": 1.8047254726464565e-06,
+      "loss": 0.6882,
+      "step": 7585
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.668818341949005,
+      "learning_rate": 1.8046681178823397e-06,
+      "loss": 0.6992,
+      "step": 7586
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.233536348011961,
+      "learning_rate": 1.8046107556081843e-06,
+      "loss": 0.6673,
+      "step": 7587
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.517986593718916,
+      "learning_rate": 1.8045533858245254e-06,
+      "loss": 0.6784,
+      "step": 7588
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.293488914623266,
+      "learning_rate": 1.8044960085318985e-06,
+      "loss": 0.6862,
+      "step": 7589
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.0222847671825352,
+      "learning_rate": 1.8044386237308396e-06,
+      "loss": 0.681,
+      "step": 7590
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.661788749768422,
+      "learning_rate": 1.8043812314218836e-06,
+      "loss": 0.6882,
+      "step": 7591
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 5.78896236758053,
+      "learning_rate": 1.8043238316055665e-06,
+      "loss": 0.6654,
+      "step": 7592
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.088842327490802,
+      "learning_rate": 1.8042664242824239e-06,
+      "loss": 0.6712,
+      "step": 7593
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1995852199074506,
+      "learning_rate": 1.8042090094529915e-06,
+      "loss": 0.6914,
+      "step": 7594
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 6.769872195500121,
+      "learning_rate": 1.8041515871178054e-06,
+      "loss": 0.6803,
+      "step": 7595
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8050425109111234,
+      "learning_rate": 1.8040941572774017e-06,
+      "loss": 0.6764,
+      "step": 7596
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.6555225107403007,
+      "learning_rate": 1.804036719932316e-06,
+      "loss": 0.6888,
+      "step": 7597
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8959446175575153,
+      "learning_rate": 1.8039792750830843e-06,
+      "loss": 0.6849,
+      "step": 7598
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 3.673662521236642,
+      "learning_rate": 1.803921822730243e-06,
+      "loss": 0.6849,
+      "step": 7599
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8386573954097687,
+      "learning_rate": 1.8038643628743283e-06,
+      "loss": 0.6901,
+      "step": 7600
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8105993441026929,
+      "learning_rate": 1.8038068955158762e-06,
+      "loss": 0.696,
+      "step": 7601
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9852521147196218,
+      "learning_rate": 1.8037494206554236e-06,
+      "loss": 0.6855,
+      "step": 7602
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.717775374514062,
+      "learning_rate": 1.8036919382935062e-06,
+      "loss": 0.6784,
+      "step": 7603
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7227757234534503,
+      "learning_rate": 1.8036344484306612e-06,
+      "loss": 0.6582,
+      "step": 7604
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.306418348991347,
+      "learning_rate": 1.8035769510674248e-06,
+      "loss": 0.6725,
+      "step": 7605
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.118874478120104,
+      "learning_rate": 1.8035194462043336e-06,
+      "loss": 0.6953,
+      "step": 7606
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.7554950055270702,
+      "learning_rate": 1.8034619338419241e-06,
+      "loss": 0.7181,
+      "step": 7607
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5475297103468596,
+      "learning_rate": 1.8034044139807337e-06,
+      "loss": 0.694,
+      "step": 7608
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7673883577884323,
+      "learning_rate": 1.8033468866212984e-06,
+      "loss": 0.6719,
+      "step": 7609
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 4.566406176004526,
+      "learning_rate": 1.8032893517641558e-06,
+      "loss": 0.6921,
+      "step": 7610
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.077974056531398,
+      "learning_rate": 1.8032318094098429e-06,
+      "loss": 0.6667,
+      "step": 7611
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.827401908515998,
+      "learning_rate": 1.8031742595588961e-06,
+      "loss": 0.6712,
+      "step": 7612
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3069577166083164,
+      "learning_rate": 1.8031167022118533e-06,
+      "loss": 0.6803,
+      "step": 7613
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.77605051462597,
+      "learning_rate": 1.8030591373692507e-06,
+      "loss": 0.6771,
+      "step": 7614
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.16249480122612,
+      "learning_rate": 1.8030015650316265e-06,
+      "loss": 0.6901,
+      "step": 7615
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.9233867419366804,
+      "learning_rate": 1.802943985199518e-06,
+      "loss": 0.6569,
+      "step": 7616
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.72898237860487,
+      "learning_rate": 1.8028863978734617e-06,
+      "loss": 0.6777,
+      "step": 7617
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8820677478236003,
+      "learning_rate": 1.802828803053996e-06,
+      "loss": 0.6862,
+      "step": 7618
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.873911713482087,
+      "learning_rate": 1.8027712007416579e-06,
+      "loss": 0.6829,
+      "step": 7619
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8588987857386263,
+      "learning_rate": 1.8027135909369852e-06,
+      "loss": 0.6771,
+      "step": 7620
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.4625633682853765,
+      "learning_rate": 1.8026559736405157e-06,
+      "loss": 0.6823,
+      "step": 7621
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0147163890899398,
+      "learning_rate": 1.8025983488527869e-06,
+      "loss": 0.6621,
+      "step": 7622
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.70556788710403,
+      "learning_rate": 1.8025407165743366e-06,
+      "loss": 0.6686,
+      "step": 7623
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0528672405552248,
+      "learning_rate": 1.8024830768057034e-06,
+      "loss": 0.6849,
+      "step": 7624
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 7.331674201040157,
+      "learning_rate": 1.8024254295474242e-06,
+      "loss": 0.696,
+      "step": 7625
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.3041863542790884,
+      "learning_rate": 1.8023677748000377e-06,
+      "loss": 0.6725,
+      "step": 7626
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2488717514478656,
+      "learning_rate": 1.8023101125640815e-06,
+      "loss": 0.6693,
+      "step": 7627
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.913000167363831,
+      "learning_rate": 1.8022524428400944e-06,
+      "loss": 0.7057,
+      "step": 7628
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9833764039891042,
+      "learning_rate": 1.8021947656286143e-06,
+      "loss": 0.6628,
+      "step": 7629
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6279920421410395,
+      "learning_rate": 1.802137080930179e-06,
+      "loss": 0.6576,
+      "step": 7630
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0905432897852878,
+      "learning_rate": 1.8020793887453279e-06,
+      "loss": 0.6947,
+      "step": 7631
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7284796469401006,
+      "learning_rate": 1.802021689074599e-06,
+      "loss": 0.6914,
+      "step": 7632
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6911146999000104,
+      "learning_rate": 1.8019639819185303e-06,
+      "loss": 0.6758,
+      "step": 7633
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5125293710240713,
+      "learning_rate": 1.801906267277661e-06,
+      "loss": 0.6706,
+      "step": 7634
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 5.677544889074706,
+      "learning_rate": 1.8018485451525297e-06,
+      "loss": 0.6888,
+      "step": 7635
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.4711265657761836,
+      "learning_rate": 1.8017908155436752e-06,
+      "loss": 0.6816,
+      "step": 7636
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.1638968775460339,
+      "learning_rate": 1.8017330784516358e-06,
+      "loss": 0.6686,
+      "step": 7637
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7433601287423109,
+      "learning_rate": 1.8016753338769508e-06,
+      "loss": 0.653,
+      "step": 7638
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.4943865008663848,
+      "learning_rate": 1.8016175818201589e-06,
+      "loss": 0.6732,
+      "step": 7639
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0633218252561578,
+      "learning_rate": 1.8015598222817994e-06,
+      "loss": 0.6816,
+      "step": 7640
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.1564019377545627,
+      "learning_rate": 1.801502055262411e-06,
+      "loss": 0.6849,
+      "step": 7641
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.3150928774295731,
+      "learning_rate": 1.8014442807625333e-06,
+      "loss": 0.6745,
+      "step": 7642
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8543607606667492,
+      "learning_rate": 1.8013864987827048e-06,
+      "loss": 0.6784,
+      "step": 7643
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6510217403630376,
+      "learning_rate": 1.8013287093234654e-06,
+      "loss": 0.6719,
+      "step": 7644
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.815408691078686,
+      "learning_rate": 1.8012709123853546e-06,
+      "loss": 0.6582,
+      "step": 7645
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.316483294696491,
+      "learning_rate": 1.8012131079689113e-06,
+      "loss": 0.6953,
+      "step": 7646
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.4836833338591493,
+      "learning_rate": 1.8011552960746753e-06,
+      "loss": 0.6908,
+      "step": 7647
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.7214655217295354,
+      "learning_rate": 1.801097476703186e-06,
+      "loss": 0.696,
+      "step": 7648
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.3316668660724003,
+      "learning_rate": 1.8010396498549831e-06,
+      "loss": 0.6764,
+      "step": 7649
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7377984472200811,
+      "learning_rate": 1.8009818155306062e-06,
+      "loss": 0.6647,
+      "step": 7650
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7126061357460048,
+      "learning_rate": 1.8009239737305952e-06,
+      "loss": 0.668,
+      "step": 7651
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7044008972144451,
+      "learning_rate": 1.8008661244554904e-06,
+      "loss": 0.6647,
+      "step": 7652
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.7321693127011963,
+      "learning_rate": 1.8008082677058307e-06,
+      "loss": 0.6895,
+      "step": 7653
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 4.530402431481635,
+      "learning_rate": 1.8007504034821567e-06,
+      "loss": 0.6641,
+      "step": 7654
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.0813244913599878,
+      "learning_rate": 1.8006925317850086e-06,
+      "loss": 0.6803,
+      "step": 7655
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6582225522940939,
+      "learning_rate": 1.800634652614926e-06,
+      "loss": 0.6693,
+      "step": 7656
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.3074171232590674,
+      "learning_rate": 1.8005767659724494e-06,
+      "loss": 0.679,
+      "step": 7657
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6442741725402705,
+      "learning_rate": 1.8005188718581192e-06,
+      "loss": 0.6706,
+      "step": 7658
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7202350219814713,
+      "learning_rate": 1.8004609702724755e-06,
+      "loss": 0.6751,
+      "step": 7659
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6567683269219928,
+      "learning_rate": 1.800403061216059e-06,
+      "loss": 0.6914,
+      "step": 7660
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8174121777220171,
+      "learning_rate": 1.8003451446894097e-06,
+      "loss": 0.6667,
+      "step": 7661
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1604744201226627,
+      "learning_rate": 1.8002872206930684e-06,
+      "loss": 0.668,
+      "step": 7662
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.1243566734021337,
+      "learning_rate": 1.8002292892275757e-06,
+      "loss": 0.6836,
+      "step": 7663
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8992691585918293,
+      "learning_rate": 1.8001713502934721e-06,
+      "loss": 0.668,
+      "step": 7664
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.761678079785416,
+      "learning_rate": 1.800113403891299e-06,
+      "loss": 0.6764,
+      "step": 7665
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.3645323917720185,
+      "learning_rate": 1.8000554500215963e-06,
+      "loss": 0.6901,
+      "step": 7666
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.851037064712901,
+      "learning_rate": 1.7999974886849057e-06,
+      "loss": 0.679,
+      "step": 7667
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6460324426036594,
+      "learning_rate": 1.7999395198817675e-06,
+      "loss": 0.666,
+      "step": 7668
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.536610200968378,
+      "learning_rate": 1.7998815436127233e-06,
+      "loss": 0.6712,
+      "step": 7669
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0337562998783156,
+      "learning_rate": 1.7998235598783136e-06,
+      "loss": 0.6882,
+      "step": 7670
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.4379033526136926,
+      "learning_rate": 1.79976556867908e-06,
+      "loss": 0.679,
+      "step": 7671
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7639174020770754,
+      "learning_rate": 1.7997075700155636e-06,
+      "loss": 0.6673,
+      "step": 7672
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8602320386126858,
+      "learning_rate": 1.799649563888306e-06,
+      "loss": 0.6504,
+      "step": 7673
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.2929180759072705,
+      "learning_rate": 1.7995915502978479e-06,
+      "loss": 0.6823,
+      "step": 7674
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8621906323296638,
+      "learning_rate": 1.7995335292447314e-06,
+      "loss": 0.6842,
+      "step": 7675
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.4793209083607355,
+      "learning_rate": 1.7994755007294975e-06,
+      "loss": 0.6458,
+      "step": 7676
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.240575538628971,
+      "learning_rate": 1.7994174647526882e-06,
+      "loss": 0.6803,
+      "step": 7677
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.2764855203501082,
+      "learning_rate": 1.799359421314845e-06,
+      "loss": 0.7012,
+      "step": 7678
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.058882128380825,
+      "learning_rate": 1.7993013704165096e-06,
+      "loss": 0.6914,
+      "step": 7679
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7090049233116296,
+      "learning_rate": 1.799243312058224e-06,
+      "loss": 0.6855,
+      "step": 7680
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.902497994039548,
+      "learning_rate": 1.7991852462405297e-06,
+      "loss": 0.6719,
+      "step": 7681
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4244781386860668,
+      "learning_rate": 1.799127172963969e-06,
+      "loss": 0.694,
+      "step": 7682
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8463256752231572,
+      "learning_rate": 1.7990690922290836e-06,
+      "loss": 0.6556,
+      "step": 7683
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6420198310295337,
+      "learning_rate": 1.7990110040364158e-06,
+      "loss": 0.6641,
+      "step": 7684
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9040746937414061,
+      "learning_rate": 1.7989529083865074e-06,
+      "loss": 0.6504,
+      "step": 7685
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.615386656898934,
+      "learning_rate": 1.7988948052799011e-06,
+      "loss": 0.6855,
+      "step": 7686
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5050257325620753,
+      "learning_rate": 1.7988366947171387e-06,
+      "loss": 0.6934,
+      "step": 7687
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7486647427762501,
+      "learning_rate": 1.7987785766987628e-06,
+      "loss": 0.6836,
+      "step": 7688
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8067099314376366,
+      "learning_rate": 1.798720451225316e-06,
+      "loss": 0.6855,
+      "step": 7689
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.872677996591358,
+      "learning_rate": 1.7986623182973407e-06,
+      "loss": 0.6836,
+      "step": 7690
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.5587365668727973,
+      "learning_rate": 1.7986041779153788e-06,
+      "loss": 0.6914,
+      "step": 7691
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.92027012264793,
+      "learning_rate": 1.798546030079974e-06,
+      "loss": 0.6901,
+      "step": 7692
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 8.20706187422276,
+      "learning_rate": 1.7984878747916686e-06,
+      "loss": 0.6602,
+      "step": 7693
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.709565947254308,
+      "learning_rate": 1.7984297120510047e-06,
+      "loss": 0.6751,
+      "step": 7694
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6611301700096069,
+      "learning_rate": 1.798371541858526e-06,
+      "loss": 0.6836,
+      "step": 7695
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7730624850690595,
+      "learning_rate": 1.798313364214775e-06,
+      "loss": 0.6771,
+      "step": 7696
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.2222696782636524,
+      "learning_rate": 1.798255179120295e-06,
+      "loss": 0.6641,
+      "step": 7697
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7854539244191486,
+      "learning_rate": 1.7981969865756286e-06,
+      "loss": 0.6621,
+      "step": 7698
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7641282138726972,
+      "learning_rate": 1.7981387865813192e-06,
+      "loss": 0.6758,
+      "step": 7699
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8238649219920461,
+      "learning_rate": 1.7980805791379098e-06,
+      "loss": 0.6634,
+      "step": 7700
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.914251722630238,
+      "learning_rate": 1.798022364245944e-06,
+      "loss": 0.679,
+      "step": 7701
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.8312808372179883,
+      "learning_rate": 1.7979641419059647e-06,
+      "loss": 0.6465,
+      "step": 7702
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.451683486082455,
+      "learning_rate": 1.7979059121185156e-06,
+      "loss": 0.6699,
+      "step": 7703
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4118686358917665,
+      "learning_rate": 1.79784767488414e-06,
+      "loss": 0.6647,
+      "step": 7704
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8000380976923656,
+      "learning_rate": 1.7977894302033817e-06,
+      "loss": 0.6895,
+      "step": 7705
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.792137585193667,
+      "learning_rate": 1.797731178076784e-06,
+      "loss": 0.6901,
+      "step": 7706
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.6083425535522864,
+      "learning_rate": 1.7976729185048905e-06,
+      "loss": 0.6706,
+      "step": 7707
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9007979498536047,
+      "learning_rate": 1.7976146514882453e-06,
+      "loss": 0.6855,
+      "step": 7708
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8577842398276514,
+      "learning_rate": 1.7975563770273916e-06,
+      "loss": 0.6699,
+      "step": 7709
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.729927946622599,
+      "learning_rate": 1.7974980951228741e-06,
+      "loss": 0.6699,
+      "step": 7710
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 7.381285019062765,
+      "learning_rate": 1.7974398057752365e-06,
+      "loss": 0.6868,
+      "step": 7711
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.982175070767714,
+      "learning_rate": 1.7973815089850222e-06,
+      "loss": 0.6849,
+      "step": 7712
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.486677248813751,
+      "learning_rate": 1.797323204752776e-06,
+      "loss": 0.6849,
+      "step": 7713
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7754461950899618,
+      "learning_rate": 1.7972648930790419e-06,
+      "loss": 0.6699,
+      "step": 7714
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.2826547543087938,
+      "learning_rate": 1.797206573964364e-06,
+      "loss": 0.6562,
+      "step": 7715
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.896602920353304,
+      "learning_rate": 1.7971482474092867e-06,
+      "loss": 0.627,
+      "step": 7716
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.8968878697381424,
+      "learning_rate": 1.797089913414354e-06,
+      "loss": 0.6829,
+      "step": 7717
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3746363426494526,
+      "learning_rate": 1.7970315719801111e-06,
+      "loss": 0.694,
+      "step": 7718
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.4987919495230626,
+      "learning_rate": 1.7969732231071015e-06,
+      "loss": 0.6439,
+      "step": 7719
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.020005368370342,
+      "learning_rate": 1.796914866795871e-06,
+      "loss": 0.6641,
+      "step": 7720
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.7010513166065717,
+      "learning_rate": 1.796856503046963e-06,
+      "loss": 0.6999,
+      "step": 7721
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.0333052954342214,
+      "learning_rate": 1.796798131860923e-06,
+      "loss": 0.6706,
+      "step": 7722
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.0684640918297597,
+      "learning_rate": 1.7967397532382958e-06,
+      "loss": 0.6908,
+      "step": 7723
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9179957567360484,
+      "learning_rate": 1.7966813671796258e-06,
+      "loss": 0.7018,
+      "step": 7724
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.246830860249278,
+      "learning_rate": 1.7966229736854581e-06,
+      "loss": 0.6921,
+      "step": 7725
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.253782275788872,
+      "learning_rate": 1.7965645727563379e-06,
+      "loss": 0.7201,
+      "step": 7726
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.4786647675207774,
+      "learning_rate": 1.7965061643928098e-06,
+      "loss": 0.668,
+      "step": 7727
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7278085348762666,
+      "learning_rate": 1.7964477485954195e-06,
+      "loss": 0.6745,
+      "step": 7728
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.7509335524244545,
+      "learning_rate": 1.7963893253647118e-06,
+      "loss": 0.6654,
+      "step": 7729
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.0562883463994512,
+      "learning_rate": 1.7963308947012322e-06,
+      "loss": 0.6751,
+      "step": 7730
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.2520167129031368,
+      "learning_rate": 1.7962724566055257e-06,
+      "loss": 0.6608,
+      "step": 7731
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.1775281506416535,
+      "learning_rate": 1.7962140110781383e-06,
+      "loss": 0.6927,
+      "step": 7732
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.3425833661761555,
+      "learning_rate": 1.7961555581196148e-06,
+      "loss": 0.6829,
+      "step": 7733
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.1647753398770249,
+      "learning_rate": 1.7960970977305011e-06,
+      "loss": 0.6868,
+      "step": 7734
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.2063848603523009,
+      "learning_rate": 1.796038629911343e-06,
+      "loss": 0.6816,
+      "step": 7735
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.3066458084771693,
+      "learning_rate": 1.7959801546626858e-06,
+      "loss": 0.6673,
+      "step": 7736
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.7929008774938664,
+      "learning_rate": 1.7959216719850756e-06,
+      "loss": 0.6621,
+      "step": 7737
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 5.905249937941882,
+      "learning_rate": 1.795863181879058e-06,
+      "loss": 0.6595,
+      "step": 7738
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.9678752920989715,
+      "learning_rate": 1.795804684345179e-06,
+      "loss": 0.6953,
+      "step": 7739
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5681519058718973,
+      "learning_rate": 1.7957461793839843e-06,
+      "loss": 0.6706,
+      "step": 7740
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7791911142008936,
+      "learning_rate": 1.79568766699602e-06,
+      "loss": 0.696,
+      "step": 7741
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.010687611626237,
+      "learning_rate": 1.7956291471818325e-06,
+      "loss": 0.668,
+      "step": 7742
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.5677406660249715,
+      "learning_rate": 1.795570619941968e-06,
+      "loss": 0.6855,
+      "step": 7743
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 4.809522251690051,
+      "learning_rate": 1.7955120852769725e-06,
+      "loss": 0.6829,
+      "step": 7744
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.071355047416568,
+      "learning_rate": 1.795453543187392e-06,
+      "loss": 0.6849,
+      "step": 7745
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.172592845794552,
+      "learning_rate": 1.7953949936737738e-06,
+      "loss": 0.6914,
+      "step": 7746
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9592761127688448,
+      "learning_rate": 1.7953364367366633e-06,
+      "loss": 0.6868,
+      "step": 7747
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.922126682600098,
+      "learning_rate": 1.7952778723766078e-06,
+      "loss": 0.6927,
+      "step": 7748
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.945590891629098,
+      "learning_rate": 1.7952193005941535e-06,
+      "loss": 0.6836,
+      "step": 7749
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.274677979322328,
+      "learning_rate": 1.795160721389847e-06,
+      "loss": 0.6764,
+      "step": 7750
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.3709740765293046,
+      "learning_rate": 1.7951021347642351e-06,
+      "loss": 0.6549,
+      "step": 7751
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2527401144597425,
+      "learning_rate": 1.795043540717865e-06,
+      "loss": 0.6536,
+      "step": 7752
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8720641726672703,
+      "learning_rate": 1.794984939251283e-06,
+      "loss": 0.6582,
+      "step": 7753
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5301154840030979,
+      "learning_rate": 1.7949263303650364e-06,
+      "loss": 0.6608,
+      "step": 7754
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.798644986553411,
+      "learning_rate": 1.7948677140596717e-06,
+      "loss": 0.6979,
+      "step": 7755
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.226580659670496,
+      "learning_rate": 1.7948090903357369e-06,
+      "loss": 0.694,
+      "step": 7756
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.4894040159646234,
+      "learning_rate": 1.7947504591937782e-06,
+      "loss": 0.6908,
+      "step": 7757
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.0571542564194827,
+      "learning_rate": 1.7946918206343433e-06,
+      "loss": 0.6673,
+      "step": 7758
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.96881781981754,
+      "learning_rate": 1.7946331746579792e-06,
+      "loss": 0.6738,
+      "step": 7759
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7319453711046187,
+      "learning_rate": 1.7945745212652335e-06,
+      "loss": 0.6699,
+      "step": 7760
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9835389613572618,
+      "learning_rate": 1.7945158604566537e-06,
+      "loss": 0.6868,
+      "step": 7761
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.645275000242452,
+      "learning_rate": 1.794457192232787e-06,
+      "loss": 0.696,
+      "step": 7762
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.9182084702999613,
+      "learning_rate": 1.7943985165941807e-06,
+      "loss": 0.6914,
+      "step": 7763
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.790009710521913,
+      "learning_rate": 1.7943398335413833e-06,
+      "loss": 0.6777,
+      "step": 7764
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.4789472793081817,
+      "learning_rate": 1.7942811430749419e-06,
+      "loss": 0.6732,
+      "step": 7765
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.369976394148154,
+      "learning_rate": 1.7942224451954043e-06,
+      "loss": 0.6797,
+      "step": 7766
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.494576614069379,
+      "learning_rate": 1.7941637399033183e-06,
+      "loss": 0.6842,
+      "step": 7767
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.9201558611079763,
+      "learning_rate": 1.7941050271992317e-06,
+      "loss": 0.6888,
+      "step": 7768
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5602585811720093,
+      "learning_rate": 1.794046307083693e-06,
+      "loss": 0.6836,
+      "step": 7769
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.357086977225307,
+      "learning_rate": 1.7939875795572497e-06,
+      "loss": 0.6875,
+      "step": 7770
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.0134529627358908,
+      "learning_rate": 1.79392884462045e-06,
+      "loss": 0.6901,
+      "step": 7771
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1982878722337058,
+      "learning_rate": 1.7938701022738426e-06,
+      "loss": 0.6686,
+      "step": 7772
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 6.166853113079343,
+      "learning_rate": 1.793811352517975e-06,
+      "loss": 0.6517,
+      "step": 7773
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5577454264587861,
+      "learning_rate": 1.793752595353396e-06,
+      "loss": 0.6764,
+      "step": 7774
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.067197973901514,
+      "learning_rate": 1.7936938307806537e-06,
+      "loss": 0.6992,
+      "step": 7775
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8915828727167665,
+      "learning_rate": 1.7936350588002966e-06,
+      "loss": 0.6751,
+      "step": 7776
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7963194913601637,
+      "learning_rate": 1.7935762794128736e-06,
+      "loss": 0.6823,
+      "step": 7777
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.2152593837015018,
+      "learning_rate": 1.7935174926189329e-06,
+      "loss": 0.6823,
+      "step": 7778
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.786164760128728,
+      "learning_rate": 1.7934586984190231e-06,
+      "loss": 0.6751,
+      "step": 7779
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.5140771216782394,
+      "learning_rate": 1.7933998968136934e-06,
+      "loss": 0.6803,
+      "step": 7780
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.267925931375109,
+      "learning_rate": 1.7933410878034922e-06,
+      "loss": 0.6868,
+      "step": 7781
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.2037706268299297,
+      "learning_rate": 1.7932822713889683e-06,
+      "loss": 0.696,
+      "step": 7782
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.6812708221366717,
+      "learning_rate": 1.793223447570671e-06,
+      "loss": 0.6816,
+      "step": 7783
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7999344490447936,
+      "learning_rate": 1.793164616349149e-06,
+      "loss": 0.6914,
+      "step": 7784
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.8825652094393135,
+      "learning_rate": 1.7931057777249513e-06,
+      "loss": 0.6764,
+      "step": 7785
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5822375803419783,
+      "learning_rate": 1.7930469316986274e-06,
+      "loss": 0.6732,
+      "step": 7786
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.127682883839617,
+      "learning_rate": 1.7929880782707265e-06,
+      "loss": 0.6803,
+      "step": 7787
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7193864297953111,
+      "learning_rate": 1.7929292174417976e-06,
+      "loss": 0.6725,
+      "step": 7788
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.3676507900463424,
+      "learning_rate": 1.7928703492123903e-06,
+      "loss": 0.6888,
+      "step": 7789
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.115137801443432,
+      "learning_rate": 1.7928114735830537e-06,
+      "loss": 0.7135,
+      "step": 7790
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7537600839874847,
+      "learning_rate": 1.7927525905543377e-06,
+      "loss": 0.6842,
+      "step": 7791
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0724550092619083,
+      "learning_rate": 1.7926937001267917e-06,
+      "loss": 0.666,
+      "step": 7792
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.969141832248279,
+      "learning_rate": 1.7926348023009652e-06,
+      "loss": 0.6536,
+      "step": 7793
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.721789623611519,
+      "learning_rate": 1.7925758970774081e-06,
+      "loss": 0.6706,
+      "step": 7794
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.727969651922138,
+      "learning_rate": 1.79251698445667e-06,
+      "loss": 0.668,
+      "step": 7795
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.938828465655528,
+      "learning_rate": 1.7924580644393004e-06,
+      "loss": 0.6699,
+      "step": 7796
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.132714051512522,
+      "learning_rate": 1.7923991370258503e-06,
+      "loss": 0.6842,
+      "step": 7797
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.637297053023925,
+      "learning_rate": 1.7923402022168683e-06,
+      "loss": 0.681,
+      "step": 7798
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.4657858942763267,
+      "learning_rate": 1.7922812600129056e-06,
+      "loss": 0.6999,
+      "step": 7799
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.3595864158556599,
+      "learning_rate": 1.7922223104145115e-06,
+      "loss": 0.7129,
+      "step": 7800
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6375380282976904,
+      "learning_rate": 1.7921633534222367e-06,
+      "loss": 0.6686,
+      "step": 7801
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1306572616689996,
+      "learning_rate": 1.7921043890366313e-06,
+      "loss": 0.6999,
+      "step": 7802
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.5860298359402867,
+      "learning_rate": 1.7920454172582456e-06,
+      "loss": 0.6641,
+      "step": 7803
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.618784335327758,
+      "learning_rate": 1.7919864380876299e-06,
+      "loss": 0.6908,
+      "step": 7804
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.4922627112146273,
+      "learning_rate": 1.7919274515253348e-06,
+      "loss": 0.6868,
+      "step": 7805
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 5.3399386185012965,
+      "learning_rate": 1.7918684575719107e-06,
+      "loss": 0.7051,
+      "step": 7806
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 4.770604689889783,
+      "learning_rate": 1.7918094562279085e-06,
+      "loss": 0.6777,
+      "step": 7807
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8664519438639625,
+      "learning_rate": 1.7917504474938783e-06,
+      "loss": 0.6699,
+      "step": 7808
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.8511234916020909,
+      "learning_rate": 1.7916914313703712e-06,
+      "loss": 0.6491,
+      "step": 7809
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.5366092303544405,
+      "learning_rate": 1.7916324078579381e-06,
+      "loss": 0.6628,
+      "step": 7810
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2535611596830032,
+      "learning_rate": 1.7915733769571296e-06,
+      "loss": 0.6745,
+      "step": 7811
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6989698018631837,
+      "learning_rate": 1.7915143386684969e-06,
+      "loss": 0.6549,
+      "step": 7812
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.770061012684887,
+      "learning_rate": 1.7914552929925906e-06,
+      "loss": 0.6999,
+      "step": 7813
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7031130536575814,
+      "learning_rate": 1.7913962399299623e-06,
+      "loss": 0.6836,
+      "step": 7814
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.0188497144046393,
+      "learning_rate": 1.791337179481163e-06,
+      "loss": 0.6842,
+      "step": 7815
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.633438038751276,
+      "learning_rate": 1.7912781116467435e-06,
+      "loss": 0.6888,
+      "step": 7816
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.546104309533672,
+      "learning_rate": 1.7912190364272556e-06,
+      "loss": 0.6751,
+      "step": 7817
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8107740013214264,
+      "learning_rate": 1.7911599538232504e-06,
+      "loss": 0.6901,
+      "step": 7818
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.818262716955003,
+      "learning_rate": 1.7911008638352797e-06,
+      "loss": 0.6602,
+      "step": 7819
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0083262569019058,
+      "learning_rate": 1.7910417664638943e-06,
+      "loss": 0.6615,
+      "step": 7820
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.0939511501646315,
+      "learning_rate": 1.790982661709646e-06,
+      "loss": 0.6784,
+      "step": 7821
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.1041883891571533,
+      "learning_rate": 1.790923549573087e-06,
+      "loss": 0.6862,
+      "step": 7822
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.606323764405245,
+      "learning_rate": 1.7908644300547683e-06,
+      "loss": 0.6725,
+      "step": 7823
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6777755136689322,
+      "learning_rate": 1.790805303155242e-06,
+      "loss": 0.6849,
+      "step": 7824
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0733145766109056,
+      "learning_rate": 1.79074616887506e-06,
+      "loss": 0.6868,
+      "step": 7825
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.335278430313521,
+      "learning_rate": 1.790687027214774e-06,
+      "loss": 0.681,
+      "step": 7826
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.030192285457427,
+      "learning_rate": 1.790627878174936e-06,
+      "loss": 0.696,
+      "step": 7827
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 8.420089349482867,
+      "learning_rate": 1.790568721756098e-06,
+      "loss": 0.7168,
+      "step": 7828
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.526764596879638,
+      "learning_rate": 1.7905095579588124e-06,
+      "loss": 0.7038,
+      "step": 7829
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.768692363621557,
+      "learning_rate": 1.7904503867836311e-06,
+      "loss": 0.6771,
+      "step": 7830
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.3119813015634803,
+      "learning_rate": 1.7903912082311063e-06,
+      "loss": 0.6784,
+      "step": 7831
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.1091801939447323,
+      "learning_rate": 1.7903320223017908e-06,
+      "loss": 0.6758,
+      "step": 7832
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 5.1107835722489785,
+      "learning_rate": 1.7902728289962363e-06,
+      "loss": 0.6621,
+      "step": 7833
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.8227012289882496,
+      "learning_rate": 1.7902136283149956e-06,
+      "loss": 0.6673,
+      "step": 7834
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.1607004155851004,
+      "learning_rate": 1.7901544202586215e-06,
+      "loss": 0.6745,
+      "step": 7835
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.775673700870771,
+      "learning_rate": 1.7900952048276664e-06,
+      "loss": 0.6823,
+      "step": 7836
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5296733646690788,
+      "learning_rate": 1.7900359820226822e-06,
+      "loss": 0.6758,
+      "step": 7837
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.27673254186823,
+      "learning_rate": 1.789976751844223e-06,
+      "loss": 0.6921,
+      "step": 7838
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.5395910987470143,
+      "learning_rate": 1.7899175142928407e-06,
+      "loss": 0.6862,
+      "step": 7839
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.4062493103788114,
+      "learning_rate": 1.7898582693690884e-06,
+      "loss": 0.6712,
+      "step": 7840
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.895738713371363,
+      "learning_rate": 1.7897990170735188e-06,
+      "loss": 0.6732,
+      "step": 7841
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0942953944778817,
+      "learning_rate": 1.7897397574066852e-06,
+      "loss": 0.7057,
+      "step": 7842
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.423828388713636,
+      "learning_rate": 1.789680490369141e-06,
+      "loss": 0.6895,
+      "step": 7843
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.1955719972870693,
+      "learning_rate": 1.7896212159614385e-06,
+      "loss": 0.6803,
+      "step": 7844
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.140832320569366,
+      "learning_rate": 1.7895619341841316e-06,
+      "loss": 0.6836,
+      "step": 7845
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9938169050670111,
+      "learning_rate": 1.789502645037773e-06,
+      "loss": 0.6673,
+      "step": 7846
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.2279627375179472,
+      "learning_rate": 1.789443348522917e-06,
+      "loss": 0.6647,
+      "step": 7847
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6622756295551799,
+      "learning_rate": 1.789384044640116e-06,
+      "loss": 0.6758,
+      "step": 7848
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.511959044916248,
+      "learning_rate": 1.789324733389924e-06,
+      "loss": 0.6764,
+      "step": 7849
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.550935299872667,
+      "learning_rate": 1.7892654147728943e-06,
+      "loss": 0.6576,
+      "step": 7850
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.468767341108722,
+      "learning_rate": 1.7892060887895808e-06,
+      "loss": 0.6738,
+      "step": 7851
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.1712737390263919,
+      "learning_rate": 1.7891467554405372e-06,
+      "loss": 0.6862,
+      "step": 7852
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 6.718496770290687,
+      "learning_rate": 1.789087414726317e-06,
+      "loss": 0.6888,
+      "step": 7853
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.7164746055104216,
+      "learning_rate": 1.7890280666474742e-06,
+      "loss": 0.653,
+      "step": 7854
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.042885591913077,
+      "learning_rate": 1.7889687112045627e-06,
+      "loss": 0.6621,
+      "step": 7855
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.078562105331707,
+      "learning_rate": 1.7889093483981364e-06,
+      "loss": 0.6784,
+      "step": 7856
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 4.188102271644591,
+      "learning_rate": 1.7888499782287495e-06,
+      "loss": 0.7064,
+      "step": 7857
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0722500551948888,
+      "learning_rate": 1.7887906006969558e-06,
+      "loss": 0.6895,
+      "step": 7858
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5678436076358448,
+      "learning_rate": 1.7887312158033097e-06,
+      "loss": 0.6439,
+      "step": 7859
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.8691149746865143,
+      "learning_rate": 1.7886718235483656e-06,
+      "loss": 0.6523,
+      "step": 7860
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6564549473558243,
+      "learning_rate": 1.7886124239326775e-06,
+      "loss": 0.6927,
+      "step": 7861
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9078109855203387,
+      "learning_rate": 1.7885530169567997e-06,
+      "loss": 0.6836,
+      "step": 7862
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8045019854372317,
+      "learning_rate": 1.7884936026212874e-06,
+      "loss": 0.6927,
+      "step": 7863
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7163109547824256,
+      "learning_rate": 1.788434180926694e-06,
+      "loss": 0.6628,
+      "step": 7864
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0913257430704493,
+      "learning_rate": 1.788374751873575e-06,
+      "loss": 0.6628,
+      "step": 7865
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0661987762305503,
+      "learning_rate": 1.7883153154624845e-06,
+      "loss": 0.6699,
+      "step": 7866
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.528210744882716,
+      "learning_rate": 1.7882558716939777e-06,
+      "loss": 0.6725,
+      "step": 7867
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.355570332792845,
+      "learning_rate": 1.788196420568609e-06,
+      "loss": 0.6888,
+      "step": 7868
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.7690713052946618,
+      "learning_rate": 1.7881369620869332e-06,
+      "loss": 0.6517,
+      "step": 7869
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.527208900338788,
+      "learning_rate": 1.7880774962495058e-06,
+      "loss": 0.6816,
+      "step": 7870
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.0300832617404059,
+      "learning_rate": 1.7880180230568814e-06,
+      "loss": 0.6738,
+      "step": 7871
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.035702393803674,
+      "learning_rate": 1.787958542509615e-06,
+      "loss": 0.6908,
+      "step": 7872
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.304000348507327,
+      "learning_rate": 1.7878990546082618e-06,
+      "loss": 0.7012,
+      "step": 7873
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.084515549847242,
+      "learning_rate": 1.787839559353377e-06,
+      "loss": 0.6504,
+      "step": 7874
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5846048033438138,
+      "learning_rate": 1.7877800567455162e-06,
+      "loss": 0.679,
+      "step": 7875
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.3150627783233995,
+      "learning_rate": 1.787720546785234e-06,
+      "loss": 0.681,
+      "step": 7876
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.5905230084477364,
+      "learning_rate": 1.7876610294730868e-06,
+      "loss": 0.6908,
+      "step": 7877
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.394880933463845,
+      "learning_rate": 1.787601504809629e-06,
+      "loss": 0.6868,
+      "step": 7878
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7939081157118376,
+      "learning_rate": 1.7875419727954171e-06,
+      "loss": 0.6589,
+      "step": 7879
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.59718884197319,
+      "learning_rate": 1.7874824334310063e-06,
+      "loss": 0.6908,
+      "step": 7880
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.7958202352931103,
+      "learning_rate": 1.787422886716952e-06,
+      "loss": 0.6895,
+      "step": 7881
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.4892046959438265,
+      "learning_rate": 1.7873633326538108e-06,
+      "loss": 0.6751,
+      "step": 7882
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8071554345701145,
+      "learning_rate": 1.7873037712421375e-06,
+      "loss": 0.6471,
+      "step": 7883
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.453655701529692,
+      "learning_rate": 1.7872442024824887e-06,
+      "loss": 0.6888,
+      "step": 7884
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8391129683386065,
+      "learning_rate": 1.7871846263754202e-06,
+      "loss": 0.6589,
+      "step": 7885
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1360570450212177,
+      "learning_rate": 1.7871250429214878e-06,
+      "loss": 0.666,
+      "step": 7886
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.652918105758686,
+      "learning_rate": 1.7870654521212475e-06,
+      "loss": 0.6895,
+      "step": 7887
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.950263418861133,
+      "learning_rate": 1.7870058539752563e-06,
+      "loss": 0.6706,
+      "step": 7888
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.1501042424357886,
+      "learning_rate": 1.7869462484840696e-06,
+      "loss": 0.6927,
+      "step": 7889
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.092878997989544,
+      "learning_rate": 1.786886635648244e-06,
+      "loss": 0.6758,
+      "step": 7890
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.020340264145261,
+      "learning_rate": 1.7868270154683358e-06,
+      "loss": 0.638,
+      "step": 7891
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8159307075578195,
+      "learning_rate": 1.7867673879449014e-06,
+      "loss": 0.6465,
+      "step": 7892
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.5277703634514417,
+      "learning_rate": 1.7867077530784976e-06,
+      "loss": 0.696,
+      "step": 7893
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.408823539258897,
+      "learning_rate": 1.7866481108696805e-06,
+      "loss": 0.681,
+      "step": 7894
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9047898830552344,
+      "learning_rate": 1.7865884613190072e-06,
+      "loss": 0.7259,
+      "step": 7895
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.189877047452228,
+      "learning_rate": 1.7865288044270343e-06,
+      "loss": 0.6947,
+      "step": 7896
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9476294688055775,
+      "learning_rate": 1.7864691401943183e-06,
+      "loss": 0.6497,
+      "step": 7897
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.641365426872345,
+      "learning_rate": 1.7864094686214166e-06,
+      "loss": 0.6738,
+      "step": 7898
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7423940343499047,
+      "learning_rate": 1.7863497897088853e-06,
+      "loss": 0.6706,
+      "step": 7899
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.3199560012847718,
+      "learning_rate": 1.7862901034572823e-06,
+      "loss": 0.681,
+      "step": 7900
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.8648846994924737,
+      "learning_rate": 1.786230409867164e-06,
+      "loss": 0.6999,
+      "step": 7901
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7709000094172318,
+      "learning_rate": 1.7861707089390878e-06,
+      "loss": 0.6732,
+      "step": 7902
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.9999503811074812,
+      "learning_rate": 1.7861110006736108e-06,
+      "loss": 0.681,
+      "step": 7903
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8528893335036821,
+      "learning_rate": 1.7860512850712905e-06,
+      "loss": 0.6543,
+      "step": 7904
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.143598037572271,
+      "learning_rate": 1.785991562132684e-06,
+      "loss": 0.6966,
+      "step": 7905
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.5922974029607277,
+      "learning_rate": 1.7859318318583485e-06,
+      "loss": 0.6966,
+      "step": 7906
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.517864618865297,
+      "learning_rate": 1.7858720942488419e-06,
+      "loss": 0.6862,
+      "step": 7907
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.168829847277345,
+      "learning_rate": 1.7858123493047215e-06,
+      "loss": 0.6673,
+      "step": 7908
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.2710237589520341,
+      "learning_rate": 1.785752597026545e-06,
+      "loss": 0.6738,
+      "step": 7909
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.730331934840108,
+      "learning_rate": 1.7856928374148698e-06,
+      "loss": 0.6452,
+      "step": 7910
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5419906940710313,
+      "learning_rate": 1.7856330704702543e-06,
+      "loss": 0.6725,
+      "step": 7911
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.082030789160582,
+      "learning_rate": 1.7855732961932554e-06,
+      "loss": 0.6699,
+      "step": 7912
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.0702445645397023,
+      "learning_rate": 1.7855135145844318e-06,
+      "loss": 0.6478,
+      "step": 7913
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.750838442164652,
+      "learning_rate": 1.785453725644341e-06,
+      "loss": 0.6895,
+      "step": 7914
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.099383694734257,
+      "learning_rate": 1.7853939293735409e-06,
+      "loss": 0.6406,
+      "step": 7915
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.986689545055561,
+      "learning_rate": 1.7853341257725903e-06,
+      "loss": 0.6517,
+      "step": 7916
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.4381911507471634,
+      "learning_rate": 1.7852743148420463e-06,
+      "loss": 0.6901,
+      "step": 7917
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.3752327509555435,
+      "learning_rate": 1.785214496582468e-06,
+      "loss": 0.6816,
+      "step": 7918
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.491070518630572,
+      "learning_rate": 1.7851546709944133e-06,
+      "loss": 0.6693,
+      "step": 7919
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9510820805400848,
+      "learning_rate": 1.7850948380784405e-06,
+      "loss": 0.6927,
+      "step": 7920
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.308565853938603,
+      "learning_rate": 1.7850349978351083e-06,
+      "loss": 0.6901,
+      "step": 7921
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.332693711298609,
+      "learning_rate": 1.784975150264975e-06,
+      "loss": 0.6803,
+      "step": 7922
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8676530021458798,
+      "learning_rate": 1.784915295368599e-06,
+      "loss": 0.6413,
+      "step": 7923
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.3466798584739488,
+      "learning_rate": 1.7848554331465395e-06,
+      "loss": 0.6628,
+      "step": 7924
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9520166403694907,
+      "learning_rate": 1.7847955635993545e-06,
+      "loss": 0.6855,
+      "step": 7925
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.3083587590780483,
+      "learning_rate": 1.7847356867276035e-06,
+      "loss": 0.6719,
+      "step": 7926
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0150824454268024,
+      "learning_rate": 1.7846758025318445e-06,
+      "loss": 0.6634,
+      "step": 7927
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 7.364132045253592,
+      "learning_rate": 1.7846159110126371e-06,
+      "loss": 0.6745,
+      "step": 7928
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.2870558460036676,
+      "learning_rate": 1.7845560121705398e-06,
+      "loss": 0.6445,
+      "step": 7929
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.968728887711417,
+      "learning_rate": 1.784496106006112e-06,
+      "loss": 0.6895,
+      "step": 7930
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.442889441067201,
+      "learning_rate": 1.7844361925199129e-06,
+      "loss": 0.6745,
+      "step": 7931
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.890978634476137,
+      "learning_rate": 1.7843762717125013e-06,
+      "loss": 0.694,
+      "step": 7932
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.840322491246795,
+      "learning_rate": 1.7843163435844363e-06,
+      "loss": 0.709,
+      "step": 7933
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.2702196349291084,
+      "learning_rate": 1.784256408136278e-06,
+      "loss": 0.651,
+      "step": 7934
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.10172839197904,
+      "learning_rate": 1.784196465368585e-06,
+      "loss": 0.6699,
+      "step": 7935
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.9436887931254234,
+      "learning_rate": 1.7841365152819167e-06,
+      "loss": 0.6738,
+      "step": 7936
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.458863935309173,
+      "learning_rate": 1.7840765578768332e-06,
+      "loss": 0.6751,
+      "step": 7937
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9564249001482956,
+      "learning_rate": 1.7840165931538942e-06,
+      "loss": 0.6413,
+      "step": 7938
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.9442963067437598,
+      "learning_rate": 1.7839566211136588e-06,
+      "loss": 0.6628,
+      "step": 7939
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.937356444127516,
+      "learning_rate": 1.7838966417566867e-06,
+      "loss": 0.7168,
+      "step": 7940
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.4815541617559,
+      "learning_rate": 1.7838366550835383e-06,
+      "loss": 0.6855,
+      "step": 7941
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 4.664942771364098,
+      "learning_rate": 1.7837766610947727e-06,
+      "loss": 0.6732,
+      "step": 7942
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 6.433777405836866,
+      "learning_rate": 1.7837166597909504e-06,
+      "loss": 0.6667,
+      "step": 7943
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.113767461735031,
+      "learning_rate": 1.7836566511726312e-06,
+      "loss": 0.6803,
+      "step": 7944
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.410123632088717,
+      "learning_rate": 1.7835966352403752e-06,
+      "loss": 0.6999,
+      "step": 7945
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6894405204513954,
+      "learning_rate": 1.7835366119947425e-06,
+      "loss": 0.6536,
+      "step": 7946
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.182220008108242,
+      "learning_rate": 1.7834765814362931e-06,
+      "loss": 0.679,
+      "step": 7947
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.4204137131175842,
+      "learning_rate": 1.7834165435655877e-06,
+      "loss": 0.6888,
+      "step": 7948
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2880087934299782,
+      "learning_rate": 1.7833564983831864e-06,
+      "loss": 0.6706,
+      "step": 7949
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.222439323796202,
+      "learning_rate": 1.7832964458896496e-06,
+      "loss": 0.6686,
+      "step": 7950
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.4469166927573047,
+      "learning_rate": 1.7832363860855378e-06,
+      "loss": 0.6849,
+      "step": 7951
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.868087156625793,
+      "learning_rate": 1.7831763189714115e-06,
+      "loss": 0.7246,
+      "step": 7952
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0086318327468549,
+      "learning_rate": 1.7831162445478315e-06,
+      "loss": 0.6458,
+      "step": 7953
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.4247742100726828,
+      "learning_rate": 1.7830561628153583e-06,
+      "loss": 0.6829,
+      "step": 7954
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9177633193130327,
+      "learning_rate": 1.7829960737745525e-06,
+      "loss": 0.694,
+      "step": 7955
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.4600022021923327,
+      "learning_rate": 1.7829359774259748e-06,
+      "loss": 0.6777,
+      "step": 7956
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0931654007952838,
+      "learning_rate": 1.782875873770187e-06,
+      "loss": 0.6732,
+      "step": 7957
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8617228561561906,
+      "learning_rate": 1.7828157628077492e-06,
+      "loss": 0.6751,
+      "step": 7958
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7067090315121163,
+      "learning_rate": 1.7827556445392226e-06,
+      "loss": 0.668,
+      "step": 7959
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 5.233583842575185,
+      "learning_rate": 1.7826955189651684e-06,
+      "loss": 0.6901,
+      "step": 7960
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.1522335458775923,
+      "learning_rate": 1.7826353860861478e-06,
+      "loss": 0.6615,
+      "step": 7961
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.3246815518683475,
+      "learning_rate": 1.782575245902722e-06,
+      "loss": 0.6803,
+      "step": 7962
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.821674598862052,
+      "learning_rate": 1.7825150984154519e-06,
+      "loss": 0.6947,
+      "step": 7963
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.351427949773968,
+      "learning_rate": 1.7824549436248994e-06,
+      "loss": 0.6699,
+      "step": 7964
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6110387476100635,
+      "learning_rate": 1.7823947815316255e-06,
+      "loss": 0.6719,
+      "step": 7965
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6450160499577823,
+      "learning_rate": 1.7823346121361922e-06,
+      "loss": 0.666,
+      "step": 7966
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.1320956812878602,
+      "learning_rate": 1.7822744354391607e-06,
+      "loss": 0.6549,
+      "step": 7967
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.747341497806228,
+      "learning_rate": 1.7822142514410928e-06,
+      "loss": 0.6536,
+      "step": 7968
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7152508730975571,
+      "learning_rate": 1.7821540601425498e-06,
+      "loss": 0.6882,
+      "step": 7969
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7956987645243097,
+      "learning_rate": 1.7820938615440939e-06,
+      "loss": 0.6732,
+      "step": 7970
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.478169271202926,
+      "learning_rate": 1.7820336556462873e-06,
+      "loss": 0.6934,
+      "step": 7971
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.0125276187481203,
+      "learning_rate": 1.781973442449691e-06,
+      "loss": 0.7116,
+      "step": 7972
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.5735387055490753,
+      "learning_rate": 1.781913221954867e-06,
+      "loss": 0.7103,
+      "step": 7973
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8854485201221655,
+      "learning_rate": 1.7818529941623783e-06,
+      "loss": 0.6673,
+      "step": 7974
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.5193541434253923,
+      "learning_rate": 1.7817927590727865e-06,
+      "loss": 0.6797,
+      "step": 7975
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.081196683720705,
+      "learning_rate": 1.7817325166866536e-06,
+      "loss": 0.6927,
+      "step": 7976
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.942453064251408,
+      "learning_rate": 1.781672267004542e-06,
+      "loss": 0.6738,
+      "step": 7977
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6570171562823817,
+      "learning_rate": 1.7816120100270137e-06,
+      "loss": 0.6764,
+      "step": 7978
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.5099272251602496,
+      "learning_rate": 1.781551745754632e-06,
+      "loss": 0.6745,
+      "step": 7979
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.578074794207129,
+      "learning_rate": 1.7814914741879583e-06,
+      "loss": 0.6738,
+      "step": 7980
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.4057032251435264,
+      "learning_rate": 1.7814311953275559e-06,
+      "loss": 0.6868,
+      "step": 7981
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.638941873021235,
+      "learning_rate": 1.7813709091739865e-06,
+      "loss": 0.6895,
+      "step": 7982
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.603285314739194,
+      "learning_rate": 1.7813106157278138e-06,
+      "loss": 0.666,
+      "step": 7983
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7153938745575241,
+      "learning_rate": 1.7812503149895998e-06,
+      "loss": 0.6836,
+      "step": 7984
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 7.227496627039111,
+      "learning_rate": 1.7811900069599077e-06,
+      "loss": 0.6797,
+      "step": 7985
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 6.3893677565695315,
+      "learning_rate": 1.7811296916393e-06,
+      "loss": 0.6686,
+      "step": 7986
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8502642582693448,
+      "learning_rate": 1.7810693690283398e-06,
+      "loss": 0.6667,
+      "step": 7987
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.380344279100638,
+      "learning_rate": 1.78100903912759e-06,
+      "loss": 0.6738,
+      "step": 7988
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2538357037385603,
+      "learning_rate": 1.7809487019376141e-06,
+      "loss": 0.6667,
+      "step": 7989
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.349372436846222,
+      "learning_rate": 1.7808883574589747e-06,
+      "loss": 0.6888,
+      "step": 7990
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.18419856557202,
+      "learning_rate": 1.7808280056922354e-06,
+      "loss": 0.6849,
+      "step": 7991
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7470866689820885,
+      "learning_rate": 1.780767646637959e-06,
+      "loss": 0.6432,
+      "step": 7992
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.5877345973873287,
+      "learning_rate": 1.7807072802967094e-06,
+      "loss": 0.6986,
+      "step": 7993
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 8.074806282209495,
+      "learning_rate": 1.7806469066690496e-06,
+      "loss": 0.6992,
+      "step": 7994
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.6698641834329676,
+      "learning_rate": 1.7805865257555429e-06,
+      "loss": 0.6908,
+      "step": 7995
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.2830441886957746,
+      "learning_rate": 1.7805261375567533e-06,
+      "loss": 0.6647,
+      "step": 7996
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.8126479250228622,
+      "learning_rate": 1.7804657420732444e-06,
+      "loss": 0.6497,
+      "step": 7997
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0351347549727155,
+      "learning_rate": 1.7804053393055795e-06,
+      "loss": 0.7083,
+      "step": 7998
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7345615639081828,
+      "learning_rate": 1.7803449292543227e-06,
+      "loss": 0.6849,
+      "step": 7999
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.9525567106600885,
+      "learning_rate": 1.7802845119200374e-06,
+      "loss": 0.6615,
+      "step": 8000
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.161357617841448,
+      "learning_rate": 1.7802240873032881e-06,
+      "loss": 0.6589,
+      "step": 8001
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.2837661028103695,
+      "learning_rate": 1.7801636554046381e-06,
+      "loss": 0.6712,
+      "step": 8002
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.3913921682423656,
+      "learning_rate": 1.780103216224652e-06,
+      "loss": 0.638,
+      "step": 8003
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8902145191422632,
+      "learning_rate": 1.7800427697638934e-06,
+      "loss": 0.6836,
+      "step": 8004
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8584177940055723,
+      "learning_rate": 1.7799823160229268e-06,
+      "loss": 0.6745,
+      "step": 8005
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.034260779743276,
+      "learning_rate": 1.7799218550023159e-06,
+      "loss": 0.6667,
+      "step": 8006
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9446492269179192,
+      "learning_rate": 1.7798613867026257e-06,
+      "loss": 0.7005,
+      "step": 8007
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 4.9809564130333115,
+      "learning_rate": 1.7798009111244202e-06,
+      "loss": 0.6562,
+      "step": 8008
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.640571225648137,
+      "learning_rate": 1.7797404282682638e-06,
+      "loss": 0.6725,
+      "step": 8009
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.3611688690712156,
+      "learning_rate": 1.7796799381347207e-06,
+      "loss": 0.6725,
+      "step": 8010
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.740362735025016,
+      "learning_rate": 1.7796194407243562e-06,
+      "loss": 0.6706,
+      "step": 8011
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.231486523829095,
+      "learning_rate": 1.7795589360377342e-06,
+      "loss": 0.6855,
+      "step": 8012
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.861443682891015,
+      "learning_rate": 1.77949842407542e-06,
+      "loss": 0.7025,
+      "step": 8013
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0317858106698439,
+      "learning_rate": 1.7794379048379779e-06,
+      "loss": 0.6536,
+      "step": 8014
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3639826533633104,
+      "learning_rate": 1.7793773783259732e-06,
+      "loss": 0.737,
+      "step": 8015
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.563400757632636,
+      "learning_rate": 1.77931684453997e-06,
+      "loss": 0.6764,
+      "step": 8016
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.160214259569214,
+      "learning_rate": 1.7792563034805341e-06,
+      "loss": 0.6556,
+      "step": 8017
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.345619208943855,
+      "learning_rate": 1.7791957551482301e-06,
+      "loss": 0.6842,
+      "step": 8018
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.9681055944688124,
+      "learning_rate": 1.7791351995436236e-06,
+      "loss": 0.6823,
+      "step": 8019
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8828944687789668,
+      "learning_rate": 1.779074636667279e-06,
+      "loss": 0.681,
+      "step": 8020
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.5982300331165873,
+      "learning_rate": 1.7790140665197617e-06,
+      "loss": 0.681,
+      "step": 8021
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8973937710767076,
+      "learning_rate": 1.7789534891016374e-06,
+      "loss": 0.6673,
+      "step": 8022
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.405715007041798,
+      "learning_rate": 1.7788929044134714e-06,
+      "loss": 0.6829,
+      "step": 8023
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.017293228407578,
+      "learning_rate": 1.778832312455829e-06,
+      "loss": 0.6888,
+      "step": 8024
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.668766780908194,
+      "learning_rate": 1.778771713229276e-06,
+      "loss": 0.6882,
+      "step": 8025
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.166410294233481,
+      "learning_rate": 1.7787111067343774e-06,
+      "loss": 0.6966,
+      "step": 8026
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.910014908049617,
+      "learning_rate": 1.7786504929716992e-06,
+      "loss": 0.6973,
+      "step": 8027
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.288497264643473,
+      "learning_rate": 1.778589871941807e-06,
+      "loss": 0.709,
+      "step": 8028
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.613456522682263,
+      "learning_rate": 1.778529243645267e-06,
+      "loss": 0.7077,
+      "step": 8029
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3990243345662794,
+      "learning_rate": 1.7784686080826448e-06,
+      "loss": 0.6777,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6208506932638238,
+      "learning_rate": 1.7784079652545056e-06,
+      "loss": 0.668,
+      "step": 8031
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6920250793065694,
+      "learning_rate": 1.7783473151614165e-06,
+      "loss": 0.6797,
+      "step": 8032
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.872956856068526,
+      "learning_rate": 1.778286657803943e-06,
+      "loss": 0.6829,
+      "step": 8033
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4115288510290065,
+      "learning_rate": 1.7782259931826514e-06,
+      "loss": 0.6771,
+      "step": 8034
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6086474496094524,
+      "learning_rate": 1.7781653212981076e-06,
+      "loss": 0.6738,
+      "step": 8035
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 6.150130938832348,
+      "learning_rate": 1.778104642150878e-06,
+      "loss": 0.651,
+      "step": 8036
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.087015834008969,
+      "learning_rate": 1.7780439557415291e-06,
+      "loss": 0.6842,
+      "step": 8037
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.661034417544444,
+      "learning_rate": 1.7779832620706271e-06,
+      "loss": 0.6771,
+      "step": 8038
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.211064548691608,
+      "learning_rate": 1.7779225611387384e-06,
+      "loss": 0.6855,
+      "step": 8039
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 5.088418688993373,
+      "learning_rate": 1.77786185294643e-06,
+      "loss": 0.6842,
+      "step": 8040
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9878617530880718,
+      "learning_rate": 1.7778011374942678e-06,
+      "loss": 0.6901,
+      "step": 8041
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3430529960934066,
+      "learning_rate": 1.777740414782819e-06,
+      "loss": 0.6888,
+      "step": 8042
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4551193520815238,
+      "learning_rate": 1.7776796848126501e-06,
+      "loss": 0.6686,
+      "step": 8043
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5856795747523836,
+      "learning_rate": 1.7776189475843277e-06,
+      "loss": 0.666,
+      "step": 8044
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9560912376916344,
+      "learning_rate": 1.7775582030984191e-06,
+      "loss": 0.6771,
+      "step": 8045
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.091863257253711,
+      "learning_rate": 1.777497451355491e-06,
+      "loss": 0.6888,
+      "step": 8046
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.054007853427597,
+      "learning_rate": 1.7774366923561105e-06,
+      "loss": 0.6816,
+      "step": 8047
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3384654683286454,
+      "learning_rate": 1.7773759261008446e-06,
+      "loss": 0.6934,
+      "step": 8048
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7666109214286962,
+      "learning_rate": 1.77731515259026e-06,
+      "loss": 0.6797,
+      "step": 8049
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0538099440078033,
+      "learning_rate": 1.7772543718249246e-06,
+      "loss": 0.6842,
+      "step": 8050
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0120625580681797,
+      "learning_rate": 1.7771935838054057e-06,
+      "loss": 0.6999,
+      "step": 8051
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.42515998816361,
+      "learning_rate": 1.77713278853227e-06,
+      "loss": 0.6816,
+      "step": 8052
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.319768201391808,
+      "learning_rate": 1.7770719860060854e-06,
+      "loss": 0.6836,
+      "step": 8053
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.263374913898832,
+      "learning_rate": 1.7770111762274194e-06,
+      "loss": 0.6849,
+      "step": 8054
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5559779862371266,
+      "learning_rate": 1.7769503591968391e-06,
+      "loss": 0.6641,
+      "step": 8055
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4199208332517297,
+      "learning_rate": 1.7768895349149125e-06,
+      "loss": 0.6667,
+      "step": 8056
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6925805848341369,
+      "learning_rate": 1.7768287033822072e-06,
+      "loss": 0.6725,
+      "step": 8057
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.659568174932323,
+      "learning_rate": 1.776767864599291e-06,
+      "loss": 0.653,
+      "step": 8058
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.925555844122031,
+      "learning_rate": 1.7767070185667314e-06,
+      "loss": 0.6654,
+      "step": 8059
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.5664422297620746,
+      "learning_rate": 1.7766461652850966e-06,
+      "loss": 0.7038,
+      "step": 8060
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.317980239282158,
+      "learning_rate": 1.7765853047549544e-06,
+      "loss": 0.6732,
+      "step": 8061
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.676664032482997,
+      "learning_rate": 1.7765244369768731e-06,
+      "loss": 0.6497,
+      "step": 8062
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 8.426516133245375,
+      "learning_rate": 1.7764635619514203e-06,
+      "loss": 0.6986,
+      "step": 8063
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4099508765833413,
+      "learning_rate": 1.7764026796791645e-06,
+      "loss": 0.679,
+      "step": 8064
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0063990605849567,
+      "learning_rate": 1.776341790160674e-06,
+      "loss": 0.6608,
+      "step": 8065
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.523918549671847,
+      "learning_rate": 1.7762808933965165e-06,
+      "loss": 0.6777,
+      "step": 8066
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6840534031114682,
+      "learning_rate": 1.7762199893872612e-06,
+      "loss": 0.6947,
+      "step": 8067
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3282259198474333,
+      "learning_rate": 1.776159078133476e-06,
+      "loss": 0.6543,
+      "step": 8068
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3509381452333078,
+      "learning_rate": 1.7760981596357292e-06,
+      "loss": 0.6855,
+      "step": 8069
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4702261360293503,
+      "learning_rate": 1.77603723389459e-06,
+      "loss": 0.6536,
+      "step": 8070
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.7210330133967864,
+      "learning_rate": 1.7759763009106263e-06,
+      "loss": 0.6829,
+      "step": 8071
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.1228160081058776,
+      "learning_rate": 1.7759153606844075e-06,
+      "loss": 0.679,
+      "step": 8072
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.5228855206910525,
+      "learning_rate": 1.7758544132165015e-06,
+      "loss": 0.6895,
+      "step": 8073
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.219122414757881,
+      "learning_rate": 1.7757934585074784e-06,
+      "loss": 0.6829,
+      "step": 8074
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.9453232687609646,
+      "learning_rate": 1.775732496557906e-06,
+      "loss": 0.6999,
+      "step": 8075
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.597482666800879,
+      "learning_rate": 1.7756715273683534e-06,
+      "loss": 0.6543,
+      "step": 8076
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.089912729771731,
+      "learning_rate": 1.77561055093939e-06,
+      "loss": 0.6751,
+      "step": 8077
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 4.4072179290658875,
+      "learning_rate": 1.775549567271585e-06,
+      "loss": 0.679,
+      "step": 8078
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.393404095200482,
+      "learning_rate": 1.775488576365507e-06,
+      "loss": 0.6517,
+      "step": 8079
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6771229606177025,
+      "learning_rate": 1.7754275782217256e-06,
+      "loss": 0.6471,
+      "step": 8080
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.2860233299192523,
+      "learning_rate": 1.77536657284081e-06,
+      "loss": 0.6576,
+      "step": 8081
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.3256924145602005,
+      "learning_rate": 1.7753055602233298e-06,
+      "loss": 0.6738,
+      "step": 8082
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.446997543658499,
+      "learning_rate": 1.775244540369854e-06,
+      "loss": 0.6569,
+      "step": 8083
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.9618897517948515,
+      "learning_rate": 1.7751835132809525e-06,
+      "loss": 0.6927,
+      "step": 8084
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1478375543495918,
+      "learning_rate": 1.7751224789571944e-06,
+      "loss": 0.6921,
+      "step": 8085
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8881801315753346,
+      "learning_rate": 1.7750614373991502e-06,
+      "loss": 0.6497,
+      "step": 8086
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3581606964092017,
+      "learning_rate": 1.7750003886073886e-06,
+      "loss": 0.6732,
+      "step": 8087
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.030436913760241,
+      "learning_rate": 1.7749393325824799e-06,
+      "loss": 0.6816,
+      "step": 8088
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.6628104862269426,
+      "learning_rate": 1.774878269324994e-06,
+      "loss": 0.6829,
+      "step": 8089
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1591816743438672,
+      "learning_rate": 1.7748171988355007e-06,
+      "loss": 0.6908,
+      "step": 8090
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.1259747470057437,
+      "learning_rate": 1.7747561211145698e-06,
+      "loss": 0.6543,
+      "step": 8091
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 6.125890679222732,
+      "learning_rate": 1.7746950361627717e-06,
+      "loss": 0.6842,
+      "step": 8092
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.2230837284195801,
+      "learning_rate": 1.7746339439806764e-06,
+      "loss": 0.6595,
+      "step": 8093
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.3053116297129619,
+      "learning_rate": 1.7745728445688537e-06,
+      "loss": 0.6738,
+      "step": 8094
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.62276005492471,
+      "learning_rate": 1.7745117379278742e-06,
+      "loss": 0.7109,
+      "step": 8095
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.661674291563569,
+      "learning_rate": 1.7744506240583084e-06,
+      "loss": 0.7331,
+      "step": 8096
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.982856647550589,
+      "learning_rate": 1.774389502960726e-06,
+      "loss": 0.6751,
+      "step": 8097
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.5805552531362586,
+      "learning_rate": 1.7743283746356986e-06,
+      "loss": 0.7044,
+      "step": 8098
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9647756602075948,
+      "learning_rate": 1.7742672390837954e-06,
+      "loss": 0.6862,
+      "step": 8099
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9312912556320647,
+      "learning_rate": 1.7742060963055879e-06,
+      "loss": 0.6829,
+      "step": 8100
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.291728778518983,
+      "learning_rate": 1.774144946301646e-06,
+      "loss": 0.6764,
+      "step": 8101
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.5697339194800999,
+      "learning_rate": 1.774083789072541e-06,
+      "loss": 0.6738,
+      "step": 8102
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.5915870815538846,
+      "learning_rate": 1.774022624618844e-06,
+      "loss": 0.6738,
+      "step": 8103
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.816467937397159,
+      "learning_rate": 1.7739614529411248e-06,
+      "loss": 0.6999,
+      "step": 8104
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.0007518427289535,
+      "learning_rate": 1.7739002740399554e-06,
+      "loss": 0.707,
+      "step": 8105
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.044770120101599,
+      "learning_rate": 1.7738390879159061e-06,
+      "loss": 0.6654,
+      "step": 8106
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.4674720852466587,
+      "learning_rate": 1.773777894569548e-06,
+      "loss": 0.6777,
+      "step": 8107
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.038328493601596,
+      "learning_rate": 1.7737166940014525e-06,
+      "loss": 0.6465,
+      "step": 8108
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.787118152520135,
+      "learning_rate": 1.7736554862121908e-06,
+      "loss": 0.6764,
+      "step": 8109
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6846141538367396,
+      "learning_rate": 1.7735942712023339e-06,
+      "loss": 0.6738,
+      "step": 8110
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.7898289905964875,
+      "learning_rate": 1.7735330489724534e-06,
+      "loss": 0.6823,
+      "step": 8111
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.38388505380564,
+      "learning_rate": 1.7734718195231203e-06,
+      "loss": 0.6654,
+      "step": 8112
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8729885540222594,
+      "learning_rate": 1.7734105828549066e-06,
+      "loss": 0.7051,
+      "step": 8113
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2246780821067125,
+      "learning_rate": 1.7733493389683832e-06,
+      "loss": 0.6882,
+      "step": 8114
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.340450906066482,
+      "learning_rate": 1.7732880878641222e-06,
+      "loss": 0.6868,
+      "step": 8115
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2329795023766303,
+      "learning_rate": 1.7732268295426951e-06,
+      "loss": 0.6719,
+      "step": 8116
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7131971619986077,
+      "learning_rate": 1.7731655640046735e-06,
+      "loss": 0.6966,
+      "step": 8117
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.5456415576942315,
+      "learning_rate": 1.7731042912506294e-06,
+      "loss": 0.6764,
+      "step": 8118
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.802590835355331,
+      "learning_rate": 1.7730430112811348e-06,
+      "loss": 0.6999,
+      "step": 8119
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1053457201304784,
+      "learning_rate": 1.7729817240967612e-06,
+      "loss": 0.7122,
+      "step": 8120
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.299477964562134,
+      "learning_rate": 1.7729204296980806e-06,
+      "loss": 0.6686,
+      "step": 8121
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.3040373090961817,
+      "learning_rate": 1.7728591280856655e-06,
+      "loss": 0.6986,
+      "step": 8122
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9413246850467756,
+      "learning_rate": 1.7727978192600878e-06,
+      "loss": 0.6855,
+      "step": 8123
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.601575715707658,
+      "learning_rate": 1.7727365032219198e-06,
+      "loss": 0.6745,
+      "step": 8124
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.2999375867158327,
+      "learning_rate": 1.7726751799717338e-06,
+      "loss": 0.6641,
+      "step": 8125
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.2938969326953424,
+      "learning_rate": 1.7726138495101018e-06,
+      "loss": 0.696,
+      "step": 8126
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.78409393261442,
+      "learning_rate": 1.7725525118375963e-06,
+      "loss": 0.666,
+      "step": 8127
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 5.392146370622397,
+      "learning_rate": 1.77249116695479e-06,
+      "loss": 0.6712,
+      "step": 8128
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 7.765871937694265,
+      "learning_rate": 1.7724298148622553e-06,
+      "loss": 0.694,
+      "step": 8129
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.127001047675583,
+      "learning_rate": 1.772368455560565e-06,
+      "loss": 0.679,
+      "step": 8130
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1189215638281813,
+      "learning_rate": 1.7723070890502915e-06,
+      "loss": 0.6693,
+      "step": 8131
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.0053365032068624,
+      "learning_rate": 1.7722457153320076e-06,
+      "loss": 0.6836,
+      "step": 8132
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.743162437287045,
+      "learning_rate": 1.772184334406286e-06,
+      "loss": 0.7051,
+      "step": 8133
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.202820901249202,
+      "learning_rate": 1.7721229462736998e-06,
+      "loss": 0.6888,
+      "step": 8134
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.2227190539325856,
+      "learning_rate": 1.772061550934822e-06,
+      "loss": 0.6816,
+      "step": 8135
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.9296078476351898,
+      "learning_rate": 1.7720001483902254e-06,
+      "loss": 0.6797,
+      "step": 8136
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.4094857244218904,
+      "learning_rate": 1.7719387386404832e-06,
+      "loss": 0.6738,
+      "step": 8137
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.61277867099441,
+      "learning_rate": 1.7718773216861682e-06,
+      "loss": 0.6712,
+      "step": 8138
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 4.355966743994376,
+      "learning_rate": 1.7718158975278543e-06,
+      "loss": 0.6921,
+      "step": 8139
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7519678321582035,
+      "learning_rate": 1.771754466166114e-06,
+      "loss": 0.6836,
+      "step": 8140
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.141081157598876,
+      "learning_rate": 1.7716930276015214e-06,
+      "loss": 0.6823,
+      "step": 8141
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.1453353169098195,
+      "learning_rate": 1.7716315818346495e-06,
+      "loss": 0.6836,
+      "step": 8142
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8122510155121356,
+      "learning_rate": 1.7715701288660715e-06,
+      "loss": 0.6608,
+      "step": 8143
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.2510360251088728,
+      "learning_rate": 1.7715086686963612e-06,
+      "loss": 0.694,
+      "step": 8144
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7149843358698816,
+      "learning_rate": 1.7714472013260926e-06,
+      "loss": 0.6751,
+      "step": 8145
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.896976054499155,
+      "learning_rate": 1.771385726755839e-06,
+      "loss": 0.6966,
+      "step": 8146
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6931386332700419,
+      "learning_rate": 1.771324244986174e-06,
+      "loss": 0.6673,
+      "step": 8147
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.34364711354458,
+      "learning_rate": 1.7712627560176715e-06,
+      "loss": 0.6615,
+      "step": 8148
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.7357485941270645,
+      "learning_rate": 1.7712012598509059e-06,
+      "loss": 0.6641,
+      "step": 8149
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7956279782142406,
+      "learning_rate": 1.7711397564864505e-06,
+      "loss": 0.6673,
+      "step": 8150
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7082800316012083,
+      "learning_rate": 1.7710782459248796e-06,
+      "loss": 0.651,
+      "step": 8151
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.264009610533429,
+      "learning_rate": 1.7710167281667672e-06,
+      "loss": 0.6803,
+      "step": 8152
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.7197508724951214,
+      "learning_rate": 1.7709552032126875e-06,
+      "loss": 0.6947,
+      "step": 8153
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.4313470798188477,
+      "learning_rate": 1.7708936710632148e-06,
+      "loss": 0.666,
+      "step": 8154
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.52813779146507,
+      "learning_rate": 1.7708321317189233e-06,
+      "loss": 0.6882,
+      "step": 8155
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.8424798962576836,
+      "learning_rate": 1.7707705851803873e-06,
+      "loss": 0.6693,
+      "step": 8156
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.8992045368418347,
+      "learning_rate": 1.770709031448181e-06,
+      "loss": 0.6549,
+      "step": 8157
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.582451843038587,
+      "learning_rate": 1.7706474705228794e-06,
+      "loss": 0.7168,
+      "step": 8158
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 7.243826918951421,
+      "learning_rate": 1.7705859024050568e-06,
+      "loss": 0.7259,
+      "step": 8159
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1529046162296575,
+      "learning_rate": 1.770524327095288e-06,
+      "loss": 0.6628,
+      "step": 8160
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.214789892082113,
+      "learning_rate": 1.7704627445941474e-06,
+      "loss": 0.7044,
+      "step": 8161
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7529585416549178,
+      "learning_rate": 1.7704011549022097e-06,
+      "loss": 0.6862,
+      "step": 8162
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.149854405522985,
+      "learning_rate": 1.77033955802005e-06,
+      "loss": 0.6797,
+      "step": 8163
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.4119175255615324,
+      "learning_rate": 1.770277953948243e-06,
+      "loss": 0.694,
+      "step": 8164
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.2092186990065747,
+      "learning_rate": 1.7702163426873638e-06,
+      "loss": 0.6921,
+      "step": 8165
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3605048872060412,
+      "learning_rate": 1.7701547242379876e-06,
+      "loss": 0.7005,
+      "step": 8166
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7633541672538737,
+      "learning_rate": 1.770093098600689e-06,
+      "loss": 0.6934,
+      "step": 8167
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6695014281244466,
+      "learning_rate": 1.7700314657760434e-06,
+      "loss": 0.668,
+      "step": 8168
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6385630013588564,
+      "learning_rate": 1.769969825764626e-06,
+      "loss": 0.6732,
+      "step": 8169
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6879260657156457,
+      "learning_rate": 1.7699081785670123e-06,
+      "loss": 0.668,
+      "step": 8170
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7459719294779475,
+      "learning_rate": 1.769846524183777e-06,
+      "loss": 0.668,
+      "step": 8171
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.788413024224751,
+      "learning_rate": 1.7697848626154966e-06,
+      "loss": 0.6829,
+      "step": 8172
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6855647258704008,
+      "learning_rate": 1.7697231938627458e-06,
+      "loss": 0.6732,
+      "step": 8173
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.008385905203185,
+      "learning_rate": 1.7696615179261003e-06,
+      "loss": 0.707,
+      "step": 8174
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.15455542580627,
+      "learning_rate": 1.7695998348061357e-06,
+      "loss": 0.6758,
+      "step": 8175
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.980500633382956,
+      "learning_rate": 1.7695381445034281e-06,
+      "loss": 0.6706,
+      "step": 8176
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.520957677985349,
+      "learning_rate": 1.7694764470185525e-06,
+      "loss": 0.6634,
+      "step": 8177
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.379698861131609,
+      "learning_rate": 1.7694147423520856e-06,
+      "loss": 0.6777,
+      "step": 8178
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.2209084734528144,
+      "learning_rate": 1.7693530305046025e-06,
+      "loss": 0.6491,
+      "step": 8179
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 5.6179639975104045,
+      "learning_rate": 1.76929131147668e-06,
+      "loss": 0.6686,
+      "step": 8180
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1146303887758724,
+      "learning_rate": 1.7692295852688931e-06,
+      "loss": 0.6934,
+      "step": 8181
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.783286621290227,
+      "learning_rate": 1.7691678518818188e-06,
+      "loss": 0.681,
+      "step": 8182
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.6912562622230408,
+      "learning_rate": 1.7691061113160328e-06,
+      "loss": 0.6823,
+      "step": 8183
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.3899496637048219,
+      "learning_rate": 1.7690443635721115e-06,
+      "loss": 0.6471,
+      "step": 8184
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 6.736847900391969,
+      "learning_rate": 1.7689826086506311e-06,
+      "loss": 0.6745,
+      "step": 8185
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.5359549229810434,
+      "learning_rate": 1.7689208465521682e-06,
+      "loss": 0.6712,
+      "step": 8186
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1757375208855845,
+      "learning_rate": 1.7688590772772987e-06,
+      "loss": 0.6549,
+      "step": 8187
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7010415049602778,
+      "learning_rate": 1.7687973008265998e-06,
+      "loss": 0.6673,
+      "step": 8188
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0060363621947706,
+      "learning_rate": 1.7687355172006476e-06,
+      "loss": 0.7005,
+      "step": 8189
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3995823685756776,
+      "learning_rate": 1.7686737264000184e-06,
+      "loss": 0.6797,
+      "step": 8190
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3964460786265,
+      "learning_rate": 1.76861192842529e-06,
+      "loss": 0.6901,
+      "step": 8191
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.8270871460565923,
+      "learning_rate": 1.7685501232770381e-06,
+      "loss": 0.7012,
+      "step": 8192
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.605134420658266,
+      "learning_rate": 1.7684883109558399e-06,
+      "loss": 0.6445,
+      "step": 8193
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2137068557397392,
+      "learning_rate": 1.7684264914622724e-06,
+      "loss": 0.6595,
+      "step": 8194
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.2496740608981667,
+      "learning_rate": 1.7683646647969125e-06,
+      "loss": 0.6667,
+      "step": 8195
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.424385354956029,
+      "learning_rate": 1.7683028309603373e-06,
+      "loss": 0.6621,
+      "step": 8196
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9142815550875967,
+      "learning_rate": 1.7682409899531238e-06,
+      "loss": 0.6888,
+      "step": 8197
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.116326621822263,
+      "learning_rate": 1.7681791417758495e-06,
+      "loss": 0.6725,
+      "step": 8198
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.175419675368888,
+      "learning_rate": 1.7681172864290909e-06,
+      "loss": 0.6471,
+      "step": 8199
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.4248473542400282,
+      "learning_rate": 1.768055423913426e-06,
+      "loss": 0.666,
+      "step": 8200
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9246134005067722,
+      "learning_rate": 1.7679935542294318e-06,
+      "loss": 0.6803,
+      "step": 8201
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.867645803507274,
+      "learning_rate": 1.7679316773776857e-06,
+      "loss": 0.6882,
+      "step": 8202
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.2035800219688637,
+      "learning_rate": 1.7678697933587653e-06,
+      "loss": 0.6719,
+      "step": 8203
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.935903801952249,
+      "learning_rate": 1.7678079021732484e-06,
+      "loss": 0.6758,
+      "step": 8204
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3187789304899704,
+      "learning_rate": 1.7677460038217125e-06,
+      "loss": 0.6764,
+      "step": 8205
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.6409512757403943,
+      "learning_rate": 1.7676840983047355e-06,
+      "loss": 0.7083,
+      "step": 8206
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9184512983405025,
+      "learning_rate": 1.7676221856228946e-06,
+      "loss": 0.6686,
+      "step": 8207
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.014943328206297,
+      "learning_rate": 1.7675602657767677e-06,
+      "loss": 0.6777,
+      "step": 8208
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.565836661722948,
+      "learning_rate": 1.7674983387669334e-06,
+      "loss": 0.6589,
+      "step": 8209
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5143088786371517,
+      "learning_rate": 1.7674364045939689e-06,
+      "loss": 0.6654,
+      "step": 8210
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.2323895371490914,
+      "learning_rate": 1.7673744632584526e-06,
+      "loss": 0.7051,
+      "step": 8211
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.548704437811231,
+      "learning_rate": 1.7673125147609626e-06,
+      "loss": 0.6719,
+      "step": 8212
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.0199018355121914,
+      "learning_rate": 1.7672505591020771e-06,
+      "loss": 0.6608,
+      "step": 8213
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2698697524866405,
+      "learning_rate": 1.7671885962823744e-06,
+      "loss": 0.6829,
+      "step": 8214
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7031258780292228,
+      "learning_rate": 1.7671266263024327e-06,
+      "loss": 0.6882,
+      "step": 8215
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.136399521484059,
+      "learning_rate": 1.7670646491628304e-06,
+      "loss": 0.7012,
+      "step": 8216
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0313037312679345,
+      "learning_rate": 1.7670026648641457e-06,
+      "loss": 0.6647,
+      "step": 8217
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.075095039569451,
+      "learning_rate": 1.7669406734069572e-06,
+      "loss": 0.6829,
+      "step": 8218
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.5573693606319854,
+      "learning_rate": 1.7668786747918438e-06,
+      "loss": 0.6784,
+      "step": 8219
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.962938850727921,
+      "learning_rate": 1.7668166690193837e-06,
+      "loss": 0.6862,
+      "step": 8220
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7720515640590893,
+      "learning_rate": 1.766754656090156e-06,
+      "loss": 0.6849,
+      "step": 8221
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6142316470200578,
+      "learning_rate": 1.7666926360047393e-06,
+      "loss": 0.6927,
+      "step": 8222
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2526534836131513,
+      "learning_rate": 1.7666306087637125e-06,
+      "loss": 0.6719,
+      "step": 8223
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.034490408875806,
+      "learning_rate": 1.7665685743676541e-06,
+      "loss": 0.6673,
+      "step": 8224
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.3818874180539242,
+      "learning_rate": 1.7665065328171438e-06,
+      "loss": 0.6602,
+      "step": 8225
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 7.290855308578177,
+      "learning_rate": 1.76644448411276e-06,
+      "loss": 0.6953,
+      "step": 8226
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.841056557950499,
+      "learning_rate": 1.766382428255082e-06,
+      "loss": 0.7233,
+      "step": 8227
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.511663957425788,
+      "learning_rate": 1.766320365244689e-06,
+      "loss": 0.6986,
+      "step": 8228
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.5184178009884994,
+      "learning_rate": 1.7662582950821604e-06,
+      "loss": 0.696,
+      "step": 8229
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6621913900281183,
+      "learning_rate": 1.7661962177680754e-06,
+      "loss": 0.6888,
+      "step": 8230
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.6650757063575634,
+      "learning_rate": 1.766134133303013e-06,
+      "loss": 0.6914,
+      "step": 8231
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.0452233633113086,
+      "learning_rate": 1.7660720416875534e-06,
+      "loss": 0.6738,
+      "step": 8232
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.2661483744980122,
+      "learning_rate": 1.7660099429222757e-06,
+      "loss": 0.6986,
+      "step": 8233
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.420322240065862,
+      "learning_rate": 1.7659478370077591e-06,
+      "loss": 0.6608,
+      "step": 8234
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.8641081392238763,
+      "learning_rate": 1.765885723944584e-06,
+      "loss": 0.6673,
+      "step": 8235
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6541503785422895,
+      "learning_rate": 1.7658236037333294e-06,
+      "loss": 0.6999,
+      "step": 8236
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.4978601258745312,
+      "learning_rate": 1.7657614763745755e-06,
+      "loss": 0.6673,
+      "step": 8237
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.0714027312415375,
+      "learning_rate": 1.7656993418689022e-06,
+      "loss": 0.6738,
+      "step": 8238
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.648879169107128,
+      "learning_rate": 1.7656372002168886e-06,
+      "loss": 0.6914,
+      "step": 8239
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.4318971321190677,
+      "learning_rate": 1.7655750514191158e-06,
+      "loss": 0.6797,
+      "step": 8240
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.195239957366541,
+      "learning_rate": 1.7655128954761635e-06,
+      "loss": 0.6771,
+      "step": 8241
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.3276709569288978,
+      "learning_rate": 1.7654507323886114e-06,
+      "loss": 0.6523,
+      "step": 8242
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7777173560939453,
+      "learning_rate": 1.76538856215704e-06,
+      "loss": 0.6784,
+      "step": 8243
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.5651129849808804,
+      "learning_rate": 1.7653263847820292e-06,
+      "loss": 0.6868,
+      "step": 8244
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0633022339523535,
+      "learning_rate": 1.7652642002641597e-06,
+      "loss": 0.6621,
+      "step": 8245
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.651009633298751,
+      "learning_rate": 1.765202008604012e-06,
+      "loss": 0.681,
+      "step": 8246
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1803580753357932,
+      "learning_rate": 1.7651398098021662e-06,
+      "loss": 0.6706,
+      "step": 8247
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.4081566319215593,
+      "learning_rate": 1.7650776038592027e-06,
+      "loss": 0.6868,
+      "step": 8248
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6338749896252511,
+      "learning_rate": 1.7650153907757024e-06,
+      "loss": 0.6868,
+      "step": 8249
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.6172661787155778,
+      "learning_rate": 1.7649531705522456e-06,
+      "loss": 0.666,
+      "step": 8250
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 5.842230117963634,
+      "learning_rate": 1.7648909431894134e-06,
+      "loss": 0.6823,
+      "step": 8251
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.818393758993432,
+      "learning_rate": 1.7648287086877863e-06,
+      "loss": 0.6836,
+      "step": 8252
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.1326406186725255,
+      "learning_rate": 1.7647664670479455e-06,
+      "loss": 0.6667,
+      "step": 8253
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6568878366838505,
+      "learning_rate": 1.7647042182704712e-06,
+      "loss": 0.6628,
+      "step": 8254
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.8146669391815524,
+      "learning_rate": 1.764641962355945e-06,
+      "loss": 0.6849,
+      "step": 8255
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.7339644212500485,
+      "learning_rate": 1.7645796993049477e-06,
+      "loss": 0.6999,
+      "step": 8256
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.508111828057744,
+      "learning_rate": 1.7645174291180607e-06,
+      "loss": 0.679,
+      "step": 8257
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.7166855283642843,
+      "learning_rate": 1.7644551517958646e-06,
+      "loss": 0.6686,
+      "step": 8258
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.6974609093641714,
+      "learning_rate": 1.764392867338941e-06,
+      "loss": 0.6803,
+      "step": 8259
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.089076446921345,
+      "learning_rate": 1.7643305757478713e-06,
+      "loss": 0.6927,
+      "step": 8260
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8677492948187838,
+      "learning_rate": 1.7642682770232367e-06,
+      "loss": 0.6628,
+      "step": 8261
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.3050916289694765,
+      "learning_rate": 1.7642059711656186e-06,
+      "loss": 0.6576,
+      "step": 8262
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.140313732100836,
+      "learning_rate": 1.7641436581755988e-06,
+      "loss": 0.6706,
+      "step": 8263
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0493675425011166,
+      "learning_rate": 1.7640813380537585e-06,
+      "loss": 0.6966,
+      "step": 8264
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7798459607596451,
+      "learning_rate": 1.7640190108006796e-06,
+      "loss": 0.6758,
+      "step": 8265
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.8160955311859865,
+      "learning_rate": 1.7639566764169436e-06,
+      "loss": 0.666,
+      "step": 8266
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.7588300322555348,
+      "learning_rate": 1.7638943349031323e-06,
+      "loss": 0.6862,
+      "step": 8267
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.3849728026728627,
+      "learning_rate": 1.763831986259828e-06,
+      "loss": 0.6556,
+      "step": 8268
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.136590102870004,
+      "learning_rate": 1.7637696304876118e-06,
+      "loss": 0.6458,
+      "step": 8269
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.0981610571080447,
+      "learning_rate": 1.7637072675870664e-06,
+      "loss": 0.6647,
+      "step": 8270
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.3013501861278263,
+      "learning_rate": 1.7636448975587733e-06,
+      "loss": 0.7122,
+      "step": 8271
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.830554378136484,
+      "learning_rate": 1.7635825204033148e-06,
+      "loss": 0.6855,
+      "step": 8272
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.50938204458683,
+      "learning_rate": 1.7635201361212732e-06,
+      "loss": 0.6732,
+      "step": 8273
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9859260435078687,
+      "learning_rate": 1.763457744713231e-06,
+      "loss": 0.6732,
+      "step": 8274
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6658647293460294,
+      "learning_rate": 1.7633953461797698e-06,
+      "loss": 0.6953,
+      "step": 8275
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 4.93105816920559,
+      "learning_rate": 1.7633329405214722e-06,
+      "loss": 0.668,
+      "step": 8276
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.8291309886595624,
+      "learning_rate": 1.7632705277389209e-06,
+      "loss": 0.6862,
+      "step": 8277
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.3334361972311257,
+      "learning_rate": 1.7632081078326984e-06,
+      "loss": 0.6673,
+      "step": 8278
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 6.672718668178125,
+      "learning_rate": 1.7631456808033866e-06,
+      "loss": 0.696,
+      "step": 8279
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.4459198810863274,
+      "learning_rate": 1.7630832466515692e-06,
+      "loss": 0.6745,
+      "step": 8280
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.011819732756287,
+      "learning_rate": 1.7630208053778284e-06,
+      "loss": 0.6732,
+      "step": 8281
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4273266034255836,
+      "learning_rate": 1.7629583569827467e-06,
+      "loss": 0.6771,
+      "step": 8282
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6821066302691414,
+      "learning_rate": 1.7628959014669074e-06,
+      "loss": 0.6784,
+      "step": 8283
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.2279563110993712,
+      "learning_rate": 1.762833438830893e-06,
+      "loss": 0.6523,
+      "step": 8284
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4964478225566977,
+      "learning_rate": 1.7627709690752867e-06,
+      "loss": 0.6797,
+      "step": 8285
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7613035736281238,
+      "learning_rate": 1.7627084922006716e-06,
+      "loss": 0.6628,
+      "step": 8286
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.308376154090149,
+      "learning_rate": 1.7626460082076307e-06,
+      "loss": 0.6628,
+      "step": 8287
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.7640308642098965,
+      "learning_rate": 1.762583517096747e-06,
+      "loss": 0.6725,
+      "step": 8288
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2261520742141516,
+      "learning_rate": 1.762521018868604e-06,
+      "loss": 0.6829,
+      "step": 8289
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.7437878893456173,
+      "learning_rate": 1.7624585135237849e-06,
+      "loss": 0.6829,
+      "step": 8290
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.01606947286814,
+      "learning_rate": 1.762396001062873e-06,
+      "loss": 0.6556,
+      "step": 8291
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.5427473632290218,
+      "learning_rate": 1.7623334814864517e-06,
+      "loss": 0.6634,
+      "step": 8292
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.1379605067542924,
+      "learning_rate": 1.7622709547951047e-06,
+      "loss": 0.6602,
+      "step": 8293
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9523811727279502,
+      "learning_rate": 1.7622084209894155e-06,
+      "loss": 0.6641,
+      "step": 8294
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.132693475938538,
+      "learning_rate": 1.7621458800699677e-06,
+      "loss": 0.6999,
+      "step": 8295
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.609455016406023,
+      "learning_rate": 1.7620833320373452e-06,
+      "loss": 0.679,
+      "step": 8296
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.210516049181611,
+      "learning_rate": 1.7620207768921311e-06,
+      "loss": 0.6764,
+      "step": 8297
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.881287192366005,
+      "learning_rate": 1.7619582146349101e-06,
+      "loss": 0.6784,
+      "step": 8298
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0740147524569013,
+      "learning_rate": 1.7618956452662654e-06,
+      "loss": 0.6745,
+      "step": 8299
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9332850608717427,
+      "learning_rate": 1.7618330687867815e-06,
+      "loss": 0.679,
+      "step": 8300
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.272861602782909,
+      "learning_rate": 1.761770485197042e-06,
+      "loss": 0.6621,
+      "step": 8301
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.669319098657656,
+      "learning_rate": 1.7617078944976313e-06,
+      "loss": 0.651,
+      "step": 8302
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8736833930219595,
+      "learning_rate": 1.7616452966891335e-06,
+      "loss": 0.6602,
+      "step": 8303
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 5.660221529832421,
+      "learning_rate": 1.7615826917721328e-06,
+      "loss": 0.6836,
+      "step": 8304
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.4802027157881805,
+      "learning_rate": 1.7615200797472136e-06,
+      "loss": 0.7005,
+      "step": 8305
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.2186244419074548,
+      "learning_rate": 1.76145746061496e-06,
+      "loss": 0.6719,
+      "step": 8306
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.2813224927459954,
+      "learning_rate": 1.7613948343759563e-06,
+      "loss": 0.64,
+      "step": 8307
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.063402143950922,
+      "learning_rate": 1.7613322010307877e-06,
+      "loss": 0.6693,
+      "step": 8308
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8668916099258407,
+      "learning_rate": 1.7612695605800383e-06,
+      "loss": 0.6732,
+      "step": 8309
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.2391618177511563,
+      "learning_rate": 1.7612069130242926e-06,
+      "loss": 0.666,
+      "step": 8310
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.5290692136380013,
+      "learning_rate": 1.7611442583641356e-06,
+      "loss": 0.6641,
+      "step": 8311
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.460435191437211,
+      "learning_rate": 1.7610815966001518e-06,
+      "loss": 0.6556,
+      "step": 8312
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.849453255899012,
+      "learning_rate": 1.761018927732926e-06,
+      "loss": 0.6875,
+      "step": 8313
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9174316576932129,
+      "learning_rate": 1.7609562517630437e-06,
+      "loss": 0.6829,
+      "step": 8314
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 5.652115576801831,
+      "learning_rate": 1.760893568691089e-06,
+      "loss": 0.6641,
+      "step": 8315
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 5.287898989315849,
+      "learning_rate": 1.7608308785176476e-06,
+      "loss": 0.6517,
+      "step": 8316
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.028516130335906,
+      "learning_rate": 1.7607681812433039e-06,
+      "loss": 0.6764,
+      "step": 8317
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.6258883119933003,
+      "learning_rate": 1.760705476868644e-06,
+      "loss": 0.6719,
+      "step": 8318
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.7932115266417235,
+      "learning_rate": 1.7606427653942527e-06,
+      "loss": 0.6836,
+      "step": 8319
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9315254493509603,
+      "learning_rate": 1.7605800468207145e-06,
+      "loss": 0.6836,
+      "step": 8320
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.4611501884078304,
+      "learning_rate": 1.7605173211486161e-06,
+      "loss": 0.7025,
+      "step": 8321
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.256899317167705,
+      "learning_rate": 1.760454588378542e-06,
+      "loss": 0.6764,
+      "step": 8322
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.4705917822017156,
+      "learning_rate": 1.760391848511078e-06,
+      "loss": 0.7272,
+      "step": 8323
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9986078069386555,
+      "learning_rate": 1.7603291015468096e-06,
+      "loss": 0.7083,
+      "step": 8324
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6899027446325254,
+      "learning_rate": 1.7602663474863224e-06,
+      "loss": 0.6569,
+      "step": 8325
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 6.670374723060532,
+      "learning_rate": 1.7602035863302024e-06,
+      "loss": 0.7018,
+      "step": 8326
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.127358785799581,
+      "learning_rate": 1.7601408180790349e-06,
+      "loss": 0.6595,
+      "step": 8327
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6549007358110601,
+      "learning_rate": 1.760078042733406e-06,
+      "loss": 0.6914,
+      "step": 8328
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7585197971847122,
+      "learning_rate": 1.7600152602939016e-06,
+      "loss": 0.6706,
+      "step": 8329
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6326049604782407,
+      "learning_rate": 1.7599524707611074e-06,
+      "loss": 0.6738,
+      "step": 8330
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1621457043205343,
+      "learning_rate": 1.7598896741356097e-06,
+      "loss": 0.681,
+      "step": 8331
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.097046593798738,
+      "learning_rate": 1.7598268704179944e-06,
+      "loss": 0.6816,
+      "step": 8332
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9852633995692814,
+      "learning_rate": 1.7597640596088477e-06,
+      "loss": 0.6615,
+      "step": 8333
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.7821357844372407,
+      "learning_rate": 1.759701241708756e-06,
+      "loss": 0.6829,
+      "step": 8334
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.342516555943925,
+      "learning_rate": 1.7596384167183054e-06,
+      "loss": 0.6602,
+      "step": 8335
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.1340567845804848,
+      "learning_rate": 1.759575584638082e-06,
+      "loss": 0.666,
+      "step": 8336
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8813073871417317,
+      "learning_rate": 1.7595127454686727e-06,
+      "loss": 0.6497,
+      "step": 8337
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.129347412582638,
+      "learning_rate": 1.7594498992106638e-06,
+      "loss": 0.6797,
+      "step": 8338
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1310518934610587,
+      "learning_rate": 1.7593870458646419e-06,
+      "loss": 0.6667,
+      "step": 8339
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.433729236853284,
+      "learning_rate": 1.7593241854311933e-06,
+      "loss": 0.694,
+      "step": 8340
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6592604820942095,
+      "learning_rate": 1.759261317910905e-06,
+      "loss": 0.6764,
+      "step": 8341
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.404018492965414,
+      "learning_rate": 1.759198443304364e-06,
+      "loss": 0.6641,
+      "step": 8342
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7519227063821126,
+      "learning_rate": 1.7591355616121564e-06,
+      "loss": 0.6628,
+      "step": 8343
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0037361141013075,
+      "learning_rate": 1.7590726728348693e-06,
+      "loss": 0.6868,
+      "step": 8344
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9447979089266481,
+      "learning_rate": 1.75900977697309e-06,
+      "loss": 0.6634,
+      "step": 8345
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.1969674060010647,
+      "learning_rate": 1.7589468740274058e-06,
+      "loss": 0.6641,
+      "step": 8346
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 4.079204188332253,
+      "learning_rate": 1.7588839639984028e-06,
+      "loss": 0.6771,
+      "step": 8347
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.848377640271139,
+      "learning_rate": 1.758821046886669e-06,
+      "loss": 0.6868,
+      "step": 8348
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7066958657695377,
+      "learning_rate": 1.7587581226927907e-06,
+      "loss": 0.6699,
+      "step": 8349
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7888790568363018,
+      "learning_rate": 1.7586951914173562e-06,
+      "loss": 0.6855,
+      "step": 8350
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.18674036823073,
+      "learning_rate": 1.7586322530609523e-06,
+      "loss": 0.681,
+      "step": 8351
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9589627868263441,
+      "learning_rate": 1.7585693076241666e-06,
+      "loss": 0.6667,
+      "step": 8352
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.3247333178740246,
+      "learning_rate": 1.758506355107586e-06,
+      "loss": 0.6901,
+      "step": 8353
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.016680637587108,
+      "learning_rate": 1.758443395511799e-06,
+      "loss": 0.6823,
+      "step": 8354
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.767005654574231,
+      "learning_rate": 1.7583804288373926e-06,
+      "loss": 0.7155,
+      "step": 8355
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.494591412663007,
+      "learning_rate": 1.7583174550849545e-06,
+      "loss": 0.6608,
+      "step": 8356
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.991625312523906,
+      "learning_rate": 1.7582544742550727e-06,
+      "loss": 0.6823,
+      "step": 8357
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9540029281986886,
+      "learning_rate": 1.7581914863483345e-06,
+      "loss": 0.6582,
+      "step": 8358
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.879703079239246,
+      "learning_rate": 1.7581284913653286e-06,
+      "loss": 0.6758,
+      "step": 8359
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.194850350146564,
+      "learning_rate": 1.758065489306642e-06,
+      "loss": 0.6888,
+      "step": 8360
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.48458269731789,
+      "learning_rate": 1.7580024801728634e-06,
+      "loss": 0.6868,
+      "step": 8361
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.212130830681786,
+      "learning_rate": 1.7579394639645804e-06,
+      "loss": 0.6732,
+      "step": 8362
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.56138677896627,
+      "learning_rate": 1.7578764406823816e-06,
+      "loss": 0.6582,
+      "step": 8363
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.0201426949192203,
+      "learning_rate": 1.7578134103268546e-06,
+      "loss": 0.6745,
+      "step": 8364
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.609016438603995,
+      "learning_rate": 1.7577503728985886e-06,
+      "loss": 0.6895,
+      "step": 8365
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6862526804486803,
+      "learning_rate": 1.757687328398171e-06,
+      "loss": 0.6758,
+      "step": 8366
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.386661140066468,
+      "learning_rate": 1.7576242768261907e-06,
+      "loss": 0.6823,
+      "step": 8367
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9160264309554997,
+      "learning_rate": 1.7575612181832357e-06,
+      "loss": 0.6836,
+      "step": 8368
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.121099031846933,
+      "learning_rate": 1.7574981524698949e-06,
+      "loss": 0.6953,
+      "step": 8369
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 7.670652409003985,
+      "learning_rate": 1.7574350796867571e-06,
+      "loss": 0.7155,
+      "step": 8370
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.2735770808466538,
+      "learning_rate": 1.7573719998344107e-06,
+      "loss": 0.6471,
+      "step": 8371
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.818522136716526,
+      "learning_rate": 1.7573089129134444e-06,
+      "loss": 0.6836,
+      "step": 8372
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6972529539280158,
+      "learning_rate": 1.757245818924447e-06,
+      "loss": 0.668,
+      "step": 8373
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.8258724564074846,
+      "learning_rate": 1.7571827178680074e-06,
+      "loss": 0.6673,
+      "step": 8374
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6538235971643236,
+      "learning_rate": 1.7571196097447145e-06,
+      "loss": 0.6777,
+      "step": 8375
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8044351287810149,
+      "learning_rate": 1.7570564945551575e-06,
+      "loss": 0.6797,
+      "step": 8376
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.171918306651579,
+      "learning_rate": 1.756993372299925e-06,
+      "loss": 0.7012,
+      "step": 8377
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8719582003037981,
+      "learning_rate": 1.7569302429796066e-06,
+      "loss": 0.6465,
+      "step": 8378
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.6208352479236705,
+      "learning_rate": 1.7568671065947914e-06,
+      "loss": 0.6823,
+      "step": 8379
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4511780941652335,
+      "learning_rate": 1.7568039631460684e-06,
+      "loss": 0.6738,
+      "step": 8380
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0374594133032975,
+      "learning_rate": 1.7567408126340273e-06,
+      "loss": 0.6699,
+      "step": 8381
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.3014559488107746,
+      "learning_rate": 1.7566776550592567e-06,
+      "loss": 0.6862,
+      "step": 8382
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6393199796144529,
+      "learning_rate": 1.7566144904223472e-06,
+      "loss": 0.6712,
+      "step": 8383
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.0177075989965076,
+      "learning_rate": 1.7565513187238875e-06,
+      "loss": 0.668,
+      "step": 8384
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.671396750294299,
+      "learning_rate": 1.7564881399644678e-06,
+      "loss": 0.7064,
+      "step": 8385
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7728410231934479,
+      "learning_rate": 1.756424954144677e-06,
+      "loss": 0.6823,
+      "step": 8386
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7434251375106967,
+      "learning_rate": 1.7563617612651051e-06,
+      "loss": 0.6882,
+      "step": 8387
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.431861873114253,
+      "learning_rate": 1.7562985613263423e-06,
+      "loss": 0.6686,
+      "step": 8388
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8348608798646953,
+      "learning_rate": 1.7562353543289782e-06,
+      "loss": 0.6745,
+      "step": 8389
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.3927432572574223,
+      "learning_rate": 1.7561721402736022e-06,
+      "loss": 0.681,
+      "step": 8390
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.6576142932746482,
+      "learning_rate": 1.7561089191608055e-06,
+      "loss": 0.6849,
+      "step": 8391
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.598980616556941,
+      "learning_rate": 1.7560456909911767e-06,
+      "loss": 0.6445,
+      "step": 8392
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1807334197107011,
+      "learning_rate": 1.755982455765307e-06,
+      "loss": 0.6667,
+      "step": 8393
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.3391822098745902,
+      "learning_rate": 1.7559192134837857e-06,
+      "loss": 0.6699,
+      "step": 8394
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.450186613038512,
+      "learning_rate": 1.755855964147204e-06,
+      "loss": 0.6816,
+      "step": 8395
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.189824052796164,
+      "learning_rate": 1.7557927077561517e-06,
+      "loss": 0.681,
+      "step": 8396
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7744826928917756,
+      "learning_rate": 1.755729444311219e-06,
+      "loss": 0.6947,
+      "step": 8397
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.3816991739978386,
+      "learning_rate": 1.7556661738129968e-06,
+      "loss": 0.6693,
+      "step": 8398
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 4.4022262539816515,
+      "learning_rate": 1.7556028962620753e-06,
+      "loss": 0.6634,
+      "step": 8399
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.8543828536630036,
+      "learning_rate": 1.755539611659045e-06,
+      "loss": 0.6777,
+      "step": 8400
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.914908692458497,
+      "learning_rate": 1.755476320004497e-06,
+      "loss": 0.696,
+      "step": 8401
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6007482498641905,
+      "learning_rate": 1.7554130212990213e-06,
+      "loss": 0.6725,
+      "step": 8402
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8778609484805571,
+      "learning_rate": 1.7553497155432094e-06,
+      "loss": 0.6732,
+      "step": 8403
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.526416975642028,
+      "learning_rate": 1.7552864027376515e-06,
+      "loss": 0.6484,
+      "step": 8404
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.7700909587899667,
+      "learning_rate": 1.7552230828829391e-06,
+      "loss": 0.6699,
+      "step": 8405
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 5.002554269648387,
+      "learning_rate": 1.7551597559796628e-06,
+      "loss": 0.6816,
+      "step": 8406
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.331346050318256,
+      "learning_rate": 1.7550964220284134e-06,
+      "loss": 0.6543,
+      "step": 8407
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.5982351407367377,
+      "learning_rate": 1.7550330810297827e-06,
+      "loss": 0.6855,
+      "step": 8408
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.4116218569050676,
+      "learning_rate": 1.7549697329843614e-06,
+      "loss": 0.6608,
+      "step": 8409
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.184525734103381,
+      "learning_rate": 1.7549063778927406e-06,
+      "loss": 0.6875,
+      "step": 8410
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.328063305922707,
+      "learning_rate": 1.754843015755512e-06,
+      "loss": 0.6849,
+      "step": 8411
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.327250988484939,
+      "learning_rate": 1.754779646573267e-06,
+      "loss": 0.679,
+      "step": 8412
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.830545073770895,
+      "learning_rate": 1.7547162703465966e-06,
+      "loss": 0.6667,
+      "step": 8413
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5516215888219336,
+      "learning_rate": 1.7546528870760924e-06,
+      "loss": 0.7096,
+      "step": 8414
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.4850047005465346,
+      "learning_rate": 1.754589496762346e-06,
+      "loss": 0.6986,
+      "step": 8415
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.6891284067867605,
+      "learning_rate": 1.7545260994059494e-06,
+      "loss": 0.6758,
+      "step": 8416
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7914590027524395,
+      "learning_rate": 1.754462695007494e-06,
+      "loss": 0.679,
+      "step": 8417
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.805396171168065,
+      "learning_rate": 1.754399283567571e-06,
+      "loss": 0.6712,
+      "step": 8418
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7520450412408397,
+      "learning_rate": 1.7543358650867733e-06,
+      "loss": 0.6719,
+      "step": 8419
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0838520953632196,
+      "learning_rate": 1.754272439565692e-06,
+      "loss": 0.6882,
+      "step": 8420
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.2959431158327688,
+      "learning_rate": 1.7542090070049196e-06,
+      "loss": 0.6934,
+      "step": 8421
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5706381789350388,
+      "learning_rate": 1.7541455674050476e-06,
+      "loss": 0.6608,
+      "step": 8422
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8076908740572829,
+      "learning_rate": 1.7540821207666686e-06,
+      "loss": 0.6621,
+      "step": 8423
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3118656839738723,
+      "learning_rate": 1.754018667090374e-06,
+      "loss": 0.6842,
+      "step": 8424
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.810183228098484,
+      "learning_rate": 1.7539552063767569e-06,
+      "loss": 0.6921,
+      "step": 8425
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.086831530379851,
+      "learning_rate": 1.753891738626409e-06,
+      "loss": 0.7031,
+      "step": 8426
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7810366730235022,
+      "learning_rate": 1.753828263839923e-06,
+      "loss": 0.6921,
+      "step": 8427
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.4050330248622425,
+      "learning_rate": 1.753764782017891e-06,
+      "loss": 0.6986,
+      "step": 8428
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8879981037843245,
+      "learning_rate": 1.7537012931609057e-06,
+      "loss": 0.6562,
+      "step": 8429
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3186191162567225,
+      "learning_rate": 1.7536377972695596e-06,
+      "loss": 0.6803,
+      "step": 8430
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.7416603316820205,
+      "learning_rate": 1.753574294344445e-06,
+      "loss": 0.6927,
+      "step": 8431
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0906804983254494,
+      "learning_rate": 1.7535107843861549e-06,
+      "loss": 0.6647,
+      "step": 8432
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.011846511821355,
+      "learning_rate": 1.7534472673952824e-06,
+      "loss": 0.6855,
+      "step": 8433
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.157095885151558,
+      "learning_rate": 1.7533837433724199e-06,
+      "loss": 0.6699,
+      "step": 8434
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.552822556937054,
+      "learning_rate": 1.7533202123181598e-06,
+      "loss": 0.6758,
+      "step": 8435
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.574972218792734,
+      "learning_rate": 1.7532566742330958e-06,
+      "loss": 0.6706,
+      "step": 8436
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.5157255471205197,
+      "learning_rate": 1.753193129117821e-06,
+      "loss": 0.6751,
+      "step": 8437
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9557284813526632,
+      "learning_rate": 1.753129576972928e-06,
+      "loss": 0.7031,
+      "step": 8438
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.2562223596306046,
+      "learning_rate": 1.7530660177990098e-06,
+      "loss": 0.6895,
+      "step": 8439
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.737662381901969,
+      "learning_rate": 1.7530024515966598e-06,
+      "loss": 0.6777,
+      "step": 8440
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.3039489062171543,
+      "learning_rate": 1.7529388783664713e-06,
+      "loss": 0.638,
+      "step": 8441
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 6.098959373330654,
+      "learning_rate": 1.7528752981090383e-06,
+      "loss": 0.6589,
+      "step": 8442
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9731203982068775,
+      "learning_rate": 1.752811710824953e-06,
+      "loss": 0.6458,
+      "step": 8443
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.352818939948505,
+      "learning_rate": 1.7527481165148093e-06,
+      "loss": 0.6699,
+      "step": 8444
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.6471607673525788,
+      "learning_rate": 1.752684515179201e-06,
+      "loss": 0.6849,
+      "step": 8445
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.106137288408484,
+      "learning_rate": 1.7526209068187217e-06,
+      "loss": 0.6829,
+      "step": 8446
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 7.292993190055021,
+      "learning_rate": 1.7525572914339648e-06,
+      "loss": 0.7077,
+      "step": 8447
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5182245949063482,
+      "learning_rate": 1.7524936690255242e-06,
+      "loss": 0.6836,
+      "step": 8448
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.7012192165064492,
+      "learning_rate": 1.7524300395939937e-06,
+      "loss": 0.6706,
+      "step": 8449
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.3398245350739004,
+      "learning_rate": 1.752366403139967e-06,
+      "loss": 0.6706,
+      "step": 8450
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.6163281105078706,
+      "learning_rate": 1.7523027596640386e-06,
+      "loss": 0.6849,
+      "step": 8451
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6806198612944929,
+      "learning_rate": 1.7522391091668011e-06,
+      "loss": 0.6758,
+      "step": 8452
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.765977024080551,
+      "learning_rate": 1.75217545164885e-06,
+      "loss": 0.6634,
+      "step": 8453
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7441317326981092,
+      "learning_rate": 1.7521117871107788e-06,
+      "loss": 0.6868,
+      "step": 8454
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6532830212415591,
+      "learning_rate": 1.7520481155531818e-06,
+      "loss": 0.6634,
+      "step": 8455
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.3050592490081656,
+      "learning_rate": 1.7519844369766533e-06,
+      "loss": 0.6738,
+      "step": 8456
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3245890249469847,
+      "learning_rate": 1.7519207513817875e-06,
+      "loss": 0.6816,
+      "step": 8457
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8068534545994837,
+      "learning_rate": 1.751857058769179e-06,
+      "loss": 0.6934,
+      "step": 8458
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.2594090500664876,
+      "learning_rate": 1.7517933591394219e-06,
+      "loss": 0.681,
+      "step": 8459
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.027494340832862,
+      "learning_rate": 1.7517296524931107e-06,
+      "loss": 0.6667,
+      "step": 8460
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.1231782745605874,
+      "learning_rate": 1.7516659388308406e-06,
+      "loss": 0.6738,
+      "step": 8461
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.4134572715618794,
+      "learning_rate": 1.7516022181532055e-06,
+      "loss": 0.6608,
+      "step": 8462
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.0322509247733525,
+      "learning_rate": 1.7515384904608005e-06,
+      "loss": 0.6712,
+      "step": 8463
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.375843737721812,
+      "learning_rate": 1.7514747557542205e-06,
+      "loss": 0.6816,
+      "step": 8464
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9143281235583178,
+      "learning_rate": 1.7514110140340596e-06,
+      "loss": 0.6764,
+      "step": 8465
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3006801213672434,
+      "learning_rate": 1.7513472653009139e-06,
+      "loss": 0.6439,
+      "step": 8466
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 5.4287916874273705,
+      "learning_rate": 1.7512835095553773e-06,
+      "loss": 0.6628,
+      "step": 8467
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.0127081171517234,
+      "learning_rate": 1.7512197467980453e-06,
+      "loss": 0.6823,
+      "step": 8468
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.940401075450121,
+      "learning_rate": 1.7511559770295133e-06,
+      "loss": 0.6484,
+      "step": 8469
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9585260602875434,
+      "learning_rate": 1.751092200250376e-06,
+      "loss": 0.6615,
+      "step": 8470
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9803752905750682,
+      "learning_rate": 1.7510284164612287e-06,
+      "loss": 0.666,
+      "step": 8471
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.1797890838387168,
+      "learning_rate": 1.7509646256626666e-06,
+      "loss": 0.6895,
+      "step": 8472
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.3088519543094574,
+      "learning_rate": 1.7509008278552856e-06,
+      "loss": 0.6732,
+      "step": 8473
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8493948075611242,
+      "learning_rate": 1.7508370230396805e-06,
+      "loss": 0.6823,
+      "step": 8474
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.920594344817036,
+      "learning_rate": 1.7507732112164471e-06,
+      "loss": 0.6992,
+      "step": 8475
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.1530470608442853,
+      "learning_rate": 1.750709392386181e-06,
+      "loss": 0.6712,
+      "step": 8476
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.559354429457932,
+      "learning_rate": 1.7506455665494774e-06,
+      "loss": 0.6823,
+      "step": 8477
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0754780402718862,
+      "learning_rate": 1.7505817337069328e-06,
+      "loss": 0.6354,
+      "step": 8478
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.9250994044382344,
+      "learning_rate": 1.7505178938591426e-06,
+      "loss": 0.6953,
+      "step": 8479
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9806830647790007,
+      "learning_rate": 1.7504540470067022e-06,
+      "loss": 0.6706,
+      "step": 8480
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.7192833508689822,
+      "learning_rate": 1.7503901931502079e-06,
+      "loss": 0.6842,
+      "step": 8481
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.481953064756009,
+      "learning_rate": 1.7503263322902555e-06,
+      "loss": 0.6751,
+      "step": 8482
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9273338406407773,
+      "learning_rate": 1.750262464427441e-06,
+      "loss": 0.6478,
+      "step": 8483
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.284239175085265,
+      "learning_rate": 1.7501985895623607e-06,
+      "loss": 0.6471,
+      "step": 8484
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9056733640241625,
+      "learning_rate": 1.7501347076956108e-06,
+      "loss": 0.6497,
+      "step": 8485
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.09495563120318,
+      "learning_rate": 1.7500708188277873e-06,
+      "loss": 0.6947,
+      "step": 8486
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9567989487105927,
+      "learning_rate": 1.7500069229594862e-06,
+      "loss": 0.6647,
+      "step": 8487
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8777924239302743,
+      "learning_rate": 1.7499430200913046e-06,
+      "loss": 0.6855,
+      "step": 8488
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9828746078637152,
+      "learning_rate": 1.749879110223838e-06,
+      "loss": 0.6829,
+      "step": 8489
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9142780010096849,
+      "learning_rate": 1.7498151933576835e-06,
+      "loss": 0.6823,
+      "step": 8490
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5064803930640844,
+      "learning_rate": 1.7497512694934375e-06,
+      "loss": 0.668,
+      "step": 8491
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.437777595836175,
+      "learning_rate": 1.7496873386316968e-06,
+      "loss": 0.7142,
+      "step": 8492
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8442274039133193,
+      "learning_rate": 1.7496234007730577e-06,
+      "loss": 0.6745,
+      "step": 8493
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.0838284109248855,
+      "learning_rate": 1.749559455918117e-06,
+      "loss": 0.6654,
+      "step": 8494
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7630137615554881,
+      "learning_rate": 1.7494955040674718e-06,
+      "loss": 0.681,
+      "step": 8495
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1896222241889551,
+      "learning_rate": 1.7494315452217188e-06,
+      "loss": 0.6523,
+      "step": 8496
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.963320570602255,
+      "learning_rate": 1.7493675793814549e-06,
+      "loss": 0.7142,
+      "step": 8497
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2481704954078534,
+      "learning_rate": 1.749303606547277e-06,
+      "loss": 0.6602,
+      "step": 8498
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.976789133883316,
+      "learning_rate": 1.7492396267197824e-06,
+      "loss": 0.6842,
+      "step": 8499
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.7762452648919975,
+      "learning_rate": 1.749175639899568e-06,
+      "loss": 0.6816,
+      "step": 8500
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2088855440228539,
+      "learning_rate": 1.749111646087231e-06,
+      "loss": 0.6452,
+      "step": 8501
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.1065632862748913,
+      "learning_rate": 1.749047645283369e-06,
+      "loss": 0.6628,
+      "step": 8502
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.104437801511193,
+      "learning_rate": 1.748983637488579e-06,
+      "loss": 0.6751,
+      "step": 8503
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8746949728965709,
+      "learning_rate": 1.7489196227034585e-06,
+      "loss": 0.6712,
+      "step": 8504
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.795150525095519,
+      "learning_rate": 1.748855600928605e-06,
+      "loss": 0.6829,
+      "step": 8505
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.4478281827080473,
+      "learning_rate": 1.7487915721646156e-06,
+      "loss": 0.7103,
+      "step": 8506
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 7.732142088541005,
+      "learning_rate": 1.7487275364120883e-06,
+      "loss": 0.6999,
+      "step": 8507
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.536941319082531,
+      "learning_rate": 1.748663493671621e-06,
+      "loss": 0.6706,
+      "step": 8508
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.801299316366276,
+      "learning_rate": 1.748599443943811e-06,
+      "loss": 0.6771,
+      "step": 8509
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.3614798437817095,
+      "learning_rate": 1.748535387229256e-06,
+      "loss": 0.6901,
+      "step": 8510
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4729496307429395,
+      "learning_rate": 1.7484713235285542e-06,
+      "loss": 0.6862,
+      "step": 8511
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.4753724907635064,
+      "learning_rate": 1.7484072528423032e-06,
+      "loss": 0.7109,
+      "step": 8512
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.316171453622901,
+      "learning_rate": 1.7483431751711014e-06,
+      "loss": 0.6803,
+      "step": 8513
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7096170867602362,
+      "learning_rate": 1.7482790905155464e-06,
+      "loss": 0.6966,
+      "step": 8514
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.526390170704584,
+      "learning_rate": 1.7482149988762366e-06,
+      "loss": 0.6745,
+      "step": 8515
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.543338276950688,
+      "learning_rate": 1.74815090025377e-06,
+      "loss": 0.653,
+      "step": 8516
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.846694830341309,
+      "learning_rate": 1.7480867946487449e-06,
+      "loss": 0.6829,
+      "step": 8517
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.7884385230418576,
+      "learning_rate": 1.7480226820617592e-06,
+      "loss": 0.6849,
+      "step": 8518
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2420483678246617,
+      "learning_rate": 1.7479585624934118e-06,
+      "loss": 0.6693,
+      "step": 8519
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6881122698977804,
+      "learning_rate": 1.7478944359443014e-06,
+      "loss": 0.6855,
+      "step": 8520
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8132351310639084,
+      "learning_rate": 1.7478303024150259e-06,
+      "loss": 0.707,
+      "step": 8521
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.1978628563605405,
+      "learning_rate": 1.7477661619061838e-06,
+      "loss": 0.6868,
+      "step": 8522
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.1267955714651876,
+      "learning_rate": 1.7477020144183743e-06,
+      "loss": 0.6784,
+      "step": 8523
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8116351982818888,
+      "learning_rate": 1.7476378599521957e-06,
+      "loss": 0.6647,
+      "step": 8524
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.6960964516078314,
+      "learning_rate": 1.7475736985082468e-06,
+      "loss": 0.6647,
+      "step": 8525
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.4487680890802155,
+      "learning_rate": 1.7475095300871263e-06,
+      "loss": 0.6615,
+      "step": 8526
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.8111870107296688,
+      "learning_rate": 1.7474453546894333e-06,
+      "loss": 0.6758,
+      "step": 8527
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1792483444198771,
+      "learning_rate": 1.7473811723157668e-06,
+      "loss": 0.679,
+      "step": 8528
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.589244876786775,
+      "learning_rate": 1.7473169829667259e-06,
+      "loss": 0.679,
+      "step": 8529
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.0317367902242551,
+      "learning_rate": 1.7472527866429095e-06,
+      "loss": 0.6712,
+      "step": 8530
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.806616933824005,
+      "learning_rate": 1.7471885833449162e-06,
+      "loss": 0.6862,
+      "step": 8531
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.079459072355763,
+      "learning_rate": 1.7471243730733462e-06,
+      "loss": 0.668,
+      "step": 8532
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 6.378378146181488,
+      "learning_rate": 1.7470601558287986e-06,
+      "loss": 0.6875,
+      "step": 8533
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.108626539142797,
+      "learning_rate": 1.7469959316118722e-06,
+      "loss": 0.6849,
+      "step": 8534
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.7244307756990667,
+      "learning_rate": 1.7469317004231666e-06,
+      "loss": 0.6849,
+      "step": 8535
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7105755623878527,
+      "learning_rate": 1.7468674622632814e-06,
+      "loss": 0.6576,
+      "step": 8536
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.1031419772696784,
+      "learning_rate": 1.7468032171328163e-06,
+      "loss": 0.6855,
+      "step": 8537
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.6976314273814137,
+      "learning_rate": 1.7467389650323708e-06,
+      "loss": 0.6921,
+      "step": 8538
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.103390802417593,
+      "learning_rate": 1.746674705962544e-06,
+      "loss": 0.6868,
+      "step": 8539
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.9158051022246703,
+      "learning_rate": 1.7466104399239368e-06,
+      "loss": 0.6634,
+      "step": 8540
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.429036637466942,
+      "learning_rate": 1.7465461669171479e-06,
+      "loss": 0.6777,
+      "step": 8541
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 3.2845180189060077,
+      "learning_rate": 1.7464818869427778e-06,
+      "loss": 0.6725,
+      "step": 8542
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.7314876439224441,
+      "learning_rate": 1.7464176000014262e-06,
+      "loss": 0.6621,
+      "step": 8543
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.592626015204027,
+      "learning_rate": 1.746353306093693e-06,
+      "loss": 0.7077,
+      "step": 8544
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.380233018967456,
+      "learning_rate": 1.7462890052201783e-06,
+      "loss": 0.6732,
+      "step": 8545
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.440335367801142,
+      "learning_rate": 1.746224697381483e-06,
+      "loss": 0.6738,
+      "step": 8546
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 5.446996739734463,
+      "learning_rate": 1.746160382578206e-06,
+      "loss": 0.6517,
+      "step": 8547
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5230542555389208,
+      "learning_rate": 1.7460960608109483e-06,
+      "loss": 0.6478,
+      "step": 8548
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9883243999463356,
+      "learning_rate": 1.74603173208031e-06,
+      "loss": 0.6771,
+      "step": 8549
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8473990682574692,
+      "learning_rate": 1.7459673963868916e-06,
+      "loss": 0.6855,
+      "step": 8550
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.0587778563531656,
+      "learning_rate": 1.7459030537312936e-06,
+      "loss": 0.696,
+      "step": 8551
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7882983785286577,
+      "learning_rate": 1.7458387041141162e-06,
+      "loss": 0.6641,
+      "step": 8552
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.013509860283508,
+      "learning_rate": 1.7457743475359606e-06,
+      "loss": 0.6862,
+      "step": 8553
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.955328087559887,
+      "learning_rate": 1.7457099839974268e-06,
+      "loss": 0.6719,
+      "step": 8554
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.3427087310290355,
+      "learning_rate": 1.745645613499116e-06,
+      "loss": 0.7057,
+      "step": 8555
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.619607951049931,
+      "learning_rate": 1.7455812360416285e-06,
+      "loss": 0.6927,
+      "step": 8556
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7867250471669366,
+      "learning_rate": 1.7455168516255655e-06,
+      "loss": 0.6439,
+      "step": 8557
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2132963992102743,
+      "learning_rate": 1.7454524602515277e-06,
+      "loss": 0.6536,
+      "step": 8558
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.841489482460608,
+      "learning_rate": 1.7453880619201163e-06,
+      "loss": 0.6686,
+      "step": 8559
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.417478351794759,
+      "learning_rate": 1.7453236566319318e-06,
+      "loss": 0.6706,
+      "step": 8560
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.453820469197068,
+      "learning_rate": 1.7452592443875762e-06,
+      "loss": 0.694,
+      "step": 8561
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9712870083912312,
+      "learning_rate": 1.74519482518765e-06,
+      "loss": 0.6882,
+      "step": 8562
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.360277202157476,
+      "learning_rate": 1.7451303990327546e-06,
+      "loss": 0.6823,
+      "step": 8563
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.214270964471408,
+      "learning_rate": 1.7450659659234911e-06,
+      "loss": 0.6764,
+      "step": 8564
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.479412739817417,
+      "learning_rate": 1.7450015258604615e-06,
+      "loss": 0.6712,
+      "step": 8565
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.545678446674888,
+      "learning_rate": 1.7449370788442664e-06,
+      "loss": 0.6673,
+      "step": 8566
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8112376211292398,
+      "learning_rate": 1.7448726248755077e-06,
+      "loss": 0.6536,
+      "step": 8567
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.793086598890094,
+      "learning_rate": 1.7448081639547868e-06,
+      "loss": 0.6621,
+      "step": 8568
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.7385051992679705,
+      "learning_rate": 1.7447436960827059e-06,
+      "loss": 0.6549,
+      "step": 8569
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.584117114376808,
+      "learning_rate": 1.744679221259866e-06,
+      "loss": 0.6888,
+      "step": 8570
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.4853189058853213,
+      "learning_rate": 1.7446147394868688e-06,
+      "loss": 0.6673,
+      "step": 8571
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.568714974176752,
+      "learning_rate": 1.7445502507643167e-06,
+      "loss": 0.6725,
+      "step": 8572
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.2802926761397955,
+      "learning_rate": 1.7444857550928114e-06,
+      "loss": 0.6855,
+      "step": 8573
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8066935255361387,
+      "learning_rate": 1.7444212524729544e-06,
+      "loss": 0.6771,
+      "step": 8574
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 4.448783248434127,
+      "learning_rate": 1.7443567429053482e-06,
+      "loss": 0.6647,
+      "step": 8575
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.7239523387010345,
+      "learning_rate": 1.7442922263905944e-06,
+      "loss": 0.6901,
+      "step": 8576
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.9451374629370396,
+      "learning_rate": 1.7442277029292956e-06,
+      "loss": 0.6543,
+      "step": 8577
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.573702380359975,
+      "learning_rate": 1.7441631725220537e-06,
+      "loss": 0.6803,
+      "step": 8578
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1455324298318579,
+      "learning_rate": 1.7440986351694716e-06,
+      "loss": 0.6927,
+      "step": 8579
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.029363590719137,
+      "learning_rate": 1.7440340908721508e-06,
+      "loss": 0.6895,
+      "step": 8580
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8245273979157051,
+      "learning_rate": 1.7439695396306936e-06,
+      "loss": 0.6536,
+      "step": 8581
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6905307557643108,
+      "learning_rate": 1.7439049814457036e-06,
+      "loss": 0.6973,
+      "step": 8582
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.8510217774342013,
+      "learning_rate": 1.7438404163177824e-06,
+      "loss": 0.6868,
+      "step": 8583
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.338330397663637,
+      "learning_rate": 1.7437758442475326e-06,
+      "loss": 0.6693,
+      "step": 8584
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.3751631707923275,
+      "learning_rate": 1.7437112652355571e-06,
+      "loss": 0.7298,
+      "step": 8585
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6305762958686594,
+      "learning_rate": 1.7436466792824587e-06,
+      "loss": 0.6764,
+      "step": 8586
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.928691895041909,
+      "learning_rate": 1.7435820863888404e-06,
+      "loss": 0.679,
+      "step": 8587
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.992309083185412,
+      "learning_rate": 1.7435174865553042e-06,
+      "loss": 0.679,
+      "step": 8588
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7163559668236464,
+      "learning_rate": 1.7434528797824539e-06,
+      "loss": 0.6686,
+      "step": 8589
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9373885826063735,
+      "learning_rate": 1.743388266070892e-06,
+      "loss": 0.6693,
+      "step": 8590
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.5760188621246303,
+      "learning_rate": 1.7433236454212214e-06,
+      "loss": 0.6628,
+      "step": 8591
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.205869437416245,
+      "learning_rate": 1.7432590178340458e-06,
+      "loss": 0.6654,
+      "step": 8592
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 9.57333716043537,
+      "learning_rate": 1.743194383309968e-06,
+      "loss": 0.7142,
+      "step": 8593
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.5165106773053543,
+      "learning_rate": 1.743129741849591e-06,
+      "loss": 0.6706,
+      "step": 8594
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.690107670131931,
+      "learning_rate": 1.7430650934535192e-06,
+      "loss": 0.6816,
+      "step": 8595
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.086692497805524,
+      "learning_rate": 1.7430004381223545e-06,
+      "loss": 0.6758,
+      "step": 8596
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.1108308936250895,
+      "learning_rate": 1.7429357758567011e-06,
+      "loss": 0.6842,
+      "step": 8597
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.966299553508105,
+      "learning_rate": 1.7428711066571623e-06,
+      "loss": 0.6732,
+      "step": 8598
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6110687143046547,
+      "learning_rate": 1.742806430524342e-06,
+      "loss": 0.6803,
+      "step": 8599
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.287697503617918,
+      "learning_rate": 1.7427417474588437e-06,
+      "loss": 0.6868,
+      "step": 8600
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.012531085532044,
+      "learning_rate": 1.7426770574612708e-06,
+      "loss": 0.6699,
+      "step": 8601
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.9346150418282706,
+      "learning_rate": 1.7426123605322272e-06,
+      "loss": 0.6361,
+      "step": 8602
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6303003476220026,
+      "learning_rate": 1.7425476566723169e-06,
+      "loss": 0.6719,
+      "step": 8603
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0417090105381916,
+      "learning_rate": 1.7424829458821434e-06,
+      "loss": 0.6543,
+      "step": 8604
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.9034032314276104,
+      "learning_rate": 1.742418228162311e-06,
+      "loss": 0.6934,
+      "step": 8605
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.3248426229828385,
+      "learning_rate": 1.7423535035134238e-06,
+      "loss": 0.6699,
+      "step": 8606
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.062324806491953,
+      "learning_rate": 1.7422887719360854e-06,
+      "loss": 0.6562,
+      "step": 8607
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 6.74948101398741,
+      "learning_rate": 1.7422240334309005e-06,
+      "loss": 0.6595,
+      "step": 8608
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.1906481618090172,
+      "learning_rate": 1.742159287998473e-06,
+      "loss": 0.6647,
+      "step": 8609
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.816044648132965,
+      "learning_rate": 1.7420945356394072e-06,
+      "loss": 0.6751,
+      "step": 8610
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.1575065802310935,
+      "learning_rate": 1.7420297763543076e-06,
+      "loss": 0.7142,
+      "step": 8611
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.764068950059657,
+      "learning_rate": 1.7419650101437785e-06,
+      "loss": 0.6725,
+      "step": 8612
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.3113165421020896,
+      "learning_rate": 1.741900237008424e-06,
+      "loss": 0.6764,
+      "step": 8613
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 5.101864599003163,
+      "learning_rate": 1.7418354569488493e-06,
+      "loss": 0.6882,
+      "step": 8614
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.666652163127594,
+      "learning_rate": 1.7417706699656586e-06,
+      "loss": 0.6888,
+      "step": 8615
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.069993944491255,
+      "learning_rate": 1.7417058760594565e-06,
+      "loss": 0.6758,
+      "step": 8616
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.9229938310977372,
+      "learning_rate": 1.741641075230848e-06,
+      "loss": 0.694,
+      "step": 8617
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7653551451982106,
+      "learning_rate": 1.741576267480438e-06,
+      "loss": 0.651,
+      "step": 8618
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.998201188288485,
+      "learning_rate": 1.7415114528088307e-06,
+      "loss": 0.7031,
+      "step": 8619
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.705855057853482,
+      "learning_rate": 1.7414466312166317e-06,
+      "loss": 0.6927,
+      "step": 8620
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.003262680919501,
+      "learning_rate": 1.7413818027044456e-06,
+      "loss": 0.6745,
+      "step": 8621
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.0170995028702727,
+      "learning_rate": 1.7413169672728774e-06,
+      "loss": 0.6882,
+      "step": 8622
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2978838986957328,
+      "learning_rate": 1.7412521249225325e-06,
+      "loss": 0.6888,
+      "step": 8623
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8108160283538626,
+      "learning_rate": 1.741187275654016e-06,
+      "loss": 0.681,
+      "step": 8624
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.311653748428833,
+      "learning_rate": 1.7411224194679332e-06,
+      "loss": 0.6842,
+      "step": 8625
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9223820298381749,
+      "learning_rate": 1.7410575563648892e-06,
+      "loss": 0.6751,
+      "step": 8626
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.427443275437862,
+      "learning_rate": 1.7409926863454896e-06,
+      "loss": 0.6979,
+      "step": 8627
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.710968981056539,
+      "learning_rate": 1.7409278094103397e-06,
+      "loss": 0.6686,
+      "step": 8628
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.226504440699719,
+      "learning_rate": 1.7408629255600449e-06,
+      "loss": 0.681,
+      "step": 8629
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.5816441811236768,
+      "learning_rate": 1.7407980347952108e-06,
+      "loss": 0.6797,
+      "step": 8630
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.957668221765823,
+      "learning_rate": 1.7407331371164435e-06,
+      "loss": 0.6868,
+      "step": 8631
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1463433221233437,
+      "learning_rate": 1.7406682325243482e-06,
+      "loss": 0.707,
+      "step": 8632
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3913961877465364,
+      "learning_rate": 1.7406033210195309e-06,
+      "loss": 0.6686,
+      "step": 8633
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.386073872657376,
+      "learning_rate": 1.7405384026025969e-06,
+      "loss": 0.6706,
+      "step": 8634
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.7645145232151362,
+      "learning_rate": 1.7404734772741527e-06,
+      "loss": 0.6855,
+      "step": 8635
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8205902170339454,
+      "learning_rate": 1.7404085450348042e-06,
+      "loss": 0.6641,
+      "step": 8636
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.0050577797160725,
+      "learning_rate": 1.7403436058851572e-06,
+      "loss": 0.679,
+      "step": 8637
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.6981148302400655,
+      "learning_rate": 1.7402786598258178e-06,
+      "loss": 0.6979,
+      "step": 8638
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.613579036528833,
+      "learning_rate": 1.7402137068573924e-06,
+      "loss": 0.6706,
+      "step": 8639
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.59763358761898,
+      "learning_rate": 1.7401487469804867e-06,
+      "loss": 0.6882,
+      "step": 8640
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.8730100243426784,
+      "learning_rate": 1.7400837801957073e-06,
+      "loss": 0.6621,
+      "step": 8641
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.0577151055701948,
+      "learning_rate": 1.7400188065036607e-06,
+      "loss": 0.6641,
+      "step": 8642
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9431599979935923,
+      "learning_rate": 1.739953825904953e-06,
+      "loss": 0.6849,
+      "step": 8643
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.9852734243642334,
+      "learning_rate": 1.7398888384001911e-06,
+      "loss": 0.6751,
+      "step": 8644
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.010915227218595,
+      "learning_rate": 1.739823843989981e-06,
+      "loss": 0.6751,
+      "step": 8645
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.84799394774548,
+      "learning_rate": 1.7397588426749294e-06,
+      "loss": 0.6699,
+      "step": 8646
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9834780734940043,
+      "learning_rate": 1.7396938344556435e-06,
+      "loss": 0.6712,
+      "step": 8647
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5863694362322177,
+      "learning_rate": 1.7396288193327292e-06,
+      "loss": 0.6628,
+      "step": 8648
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.688169907241975,
+      "learning_rate": 1.739563797306794e-06,
+      "loss": 0.6836,
+      "step": 8649
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.806665307385985,
+      "learning_rate": 1.7394987683784446e-06,
+      "loss": 0.6882,
+      "step": 8650
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8972501823720941,
+      "learning_rate": 1.7394337325482874e-06,
+      "loss": 0.6725,
+      "step": 8651
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.5953402561630483,
+      "learning_rate": 1.7393686898169299e-06,
+      "loss": 0.6868,
+      "step": 8652
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8327396892990179,
+      "learning_rate": 1.7393036401849792e-06,
+      "loss": 0.6536,
+      "step": 8653
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2306636862181517,
+      "learning_rate": 1.7392385836530422e-06,
+      "loss": 0.6673,
+      "step": 8654
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.252107476667633,
+      "learning_rate": 1.7391735202217262e-06,
+      "loss": 0.6947,
+      "step": 8655
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.471484149013074,
+      "learning_rate": 1.739108449891638e-06,
+      "loss": 0.6706,
+      "step": 8656
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.7682815078226106,
+      "learning_rate": 1.7390433726633856e-06,
+      "loss": 0.6699,
+      "step": 8657
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.532289143632362,
+      "learning_rate": 1.7389782885375758e-06,
+      "loss": 0.7038,
+      "step": 8658
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.7280841199444859,
+      "learning_rate": 1.7389131975148166e-06,
+      "loss": 0.707,
+      "step": 8659
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.1837131753588235,
+      "learning_rate": 1.7388480995957152e-06,
+      "loss": 0.6921,
+      "step": 8660
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.812125550496958,
+      "learning_rate": 1.738782994780879e-06,
+      "loss": 0.696,
+      "step": 8661
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.025326709442193,
+      "learning_rate": 1.7387178830709154e-06,
+      "loss": 0.6842,
+      "step": 8662
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.1277862933142069,
+      "learning_rate": 1.7386527644664328e-06,
+      "loss": 0.6777,
+      "step": 8663
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8314189126826788,
+      "learning_rate": 1.7385876389680388e-06,
+      "loss": 0.6706,
+      "step": 8664
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6649068976514594,
+      "learning_rate": 1.738522506576341e-06,
+      "loss": 0.6608,
+      "step": 8665
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.3073782178497555,
+      "learning_rate": 1.7384573672919468e-06,
+      "loss": 0.6686,
+      "step": 8666
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.8469043215876674,
+      "learning_rate": 1.7383922211154654e-06,
+      "loss": 0.6908,
+      "step": 8667
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.752106714556416,
+      "learning_rate": 1.7383270680475038e-06,
+      "loss": 0.6595,
+      "step": 8668
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.498684657089412,
+      "learning_rate": 1.7382619080886704e-06,
+      "loss": 0.6686,
+      "step": 8669
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.359397435968665,
+      "learning_rate": 1.7381967412395732e-06,
+      "loss": 0.6348,
+      "step": 8670
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.08454247820394,
+      "learning_rate": 1.7381315675008205e-06,
+      "loss": 0.6582,
+      "step": 8671
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.484554262160724,
+      "learning_rate": 1.738066386873021e-06,
+      "loss": 0.6328,
+      "step": 8672
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.1632319371045567,
+      "learning_rate": 1.7380011993567823e-06,
+      "loss": 0.6771,
+      "step": 8673
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.022571126956604,
+      "learning_rate": 1.7379360049527133e-06,
+      "loss": 0.6875,
+      "step": 8674
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.7836973973995485,
+      "learning_rate": 1.7378708036614222e-06,
+      "loss": 0.6829,
+      "step": 8675
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.928155181410454,
+      "learning_rate": 1.7378055954835177e-06,
+      "loss": 0.6387,
+      "step": 8676
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.627353419571463,
+      "learning_rate": 1.7377403804196083e-06,
+      "loss": 0.6777,
+      "step": 8677
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 4.444197965302287,
+      "learning_rate": 1.7376751584703028e-06,
+      "loss": 0.6992,
+      "step": 8678
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.9624734928780465,
+      "learning_rate": 1.7376099296362096e-06,
+      "loss": 0.748,
+      "step": 8679
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.285291308348364,
+      "learning_rate": 1.7375446939179378e-06,
+      "loss": 0.679,
+      "step": 8680
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.2793882373826264,
+      "learning_rate": 1.737479451316096e-06,
+      "loss": 0.6712,
+      "step": 8681
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9039189443790199,
+      "learning_rate": 1.7374142018312934e-06,
+      "loss": 0.6914,
+      "step": 8682
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7969641499298772,
+      "learning_rate": 1.7373489454641388e-06,
+      "loss": 0.6803,
+      "step": 8683
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.327779195188425,
+      "learning_rate": 1.7372836822152412e-06,
+      "loss": 0.6484,
+      "step": 8684
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0298270317414167,
+      "learning_rate": 1.73721841208521e-06,
+      "loss": 0.6784,
+      "step": 8685
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.2142224313450019,
+      "learning_rate": 1.7371531350746543e-06,
+      "loss": 0.6452,
+      "step": 8686
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9102664863547918,
+      "learning_rate": 1.7370878511841828e-06,
+      "loss": 0.6751,
+      "step": 8687
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7710773208201414,
+      "learning_rate": 1.7370225604144055e-06,
+      "loss": 0.6621,
+      "step": 8688
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.0713174801580645,
+      "learning_rate": 1.7369572627659313e-06,
+      "loss": 0.6914,
+      "step": 8689
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.9477469575993099,
+      "learning_rate": 1.7368919582393697e-06,
+      "loss": 0.6647,
+      "step": 8690
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.3923085444066967,
+      "learning_rate": 1.7368266468353303e-06,
+      "loss": 0.681,
+      "step": 8691
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1160131243573588,
+      "learning_rate": 1.7367613285544228e-06,
+      "loss": 0.6745,
+      "step": 8692
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.299833819630635,
+      "learning_rate": 1.7366960033972568e-06,
+      "loss": 0.6615,
+      "step": 8693
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.194810764555172,
+      "learning_rate": 1.7366306713644416e-06,
+      "loss": 0.6732,
+      "step": 8694
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.8071008126246544,
+      "learning_rate": 1.7365653324565872e-06,
+      "loss": 0.6419,
+      "step": 8695
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9362980430046428,
+      "learning_rate": 1.7364999866743038e-06,
+      "loss": 0.6868,
+      "step": 8696
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6912331673251444,
+      "learning_rate": 1.7364346340182003e-06,
+      "loss": 0.6699,
+      "step": 8697
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8860868063441683,
+      "learning_rate": 1.7363692744888877e-06,
+      "loss": 0.7012,
+      "step": 8698
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1570679798159986,
+      "learning_rate": 1.7363039080869752e-06,
+      "loss": 0.6667,
+      "step": 8699
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0029080309189116,
+      "learning_rate": 1.7362385348130735e-06,
+      "loss": 0.7012,
+      "step": 8700
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.948924524147168,
+      "learning_rate": 1.7361731546677923e-06,
+      "loss": 0.6706,
+      "step": 8701
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7674398817469692,
+      "learning_rate": 1.736107767651742e-06,
+      "loss": 0.6862,
+      "step": 8702
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5489592754562365,
+      "learning_rate": 1.7360423737655328e-06,
+      "loss": 0.6849,
+      "step": 8703
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.9386395700676684,
+      "learning_rate": 1.735976973009775e-06,
+      "loss": 0.6947,
+      "step": 8704
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.0207471574654423,
+      "learning_rate": 1.7359115653850789e-06,
+      "loss": 0.6706,
+      "step": 8705
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.74410453800444,
+      "learning_rate": 1.7358461508920552e-06,
+      "loss": 0.6992,
+      "step": 8706
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.9776932656124546,
+      "learning_rate": 1.7357807295313142e-06,
+      "loss": 0.6908,
+      "step": 8707
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.2469403953987217,
+      "learning_rate": 1.7357153013034668e-06,
+      "loss": 0.6829,
+      "step": 8708
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6431222642616439,
+      "learning_rate": 1.7356498662091235e-06,
+      "loss": 0.6725,
+      "step": 8709
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8700562847706494,
+      "learning_rate": 1.7355844242488945e-06,
+      "loss": 0.6745,
+      "step": 8710
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.4760935818938612,
+      "learning_rate": 1.7355189754233912e-06,
+      "loss": 0.681,
+      "step": 8711
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0946056643397983,
+      "learning_rate": 1.7354535197332245e-06,
+      "loss": 0.6842,
+      "step": 8712
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 5.966866482135429,
+      "learning_rate": 1.7353880571790049e-06,
+      "loss": 0.6914,
+      "step": 8713
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.1104615051732918,
+      "learning_rate": 1.7353225877613435e-06,
+      "loss": 0.6706,
+      "step": 8714
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.534656421975636,
+      "learning_rate": 1.7352571114808514e-06,
+      "loss": 0.6556,
+      "step": 8715
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.4051887440715234,
+      "learning_rate": 1.7351916283381396e-06,
+      "loss": 0.6875,
+      "step": 8716
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.4476133685196044,
+      "learning_rate": 1.7351261383338191e-06,
+      "loss": 0.6576,
+      "step": 8717
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.941265053633604,
+      "learning_rate": 1.7350606414685015e-06,
+      "loss": 0.6914,
+      "step": 8718
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.1895700186682023,
+      "learning_rate": 1.7349951377427983e-06,
+      "loss": 0.6895,
+      "step": 8719
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.075940149057537,
+      "learning_rate": 1.7349296271573202e-06,
+      "loss": 0.6569,
+      "step": 8720
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.6263566082799543,
+      "learning_rate": 1.7348641097126787e-06,
+      "loss": 0.6719,
+      "step": 8721
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.881612400436302,
+      "learning_rate": 1.7347985854094857e-06,
+      "loss": 0.6947,
+      "step": 8722
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.6598521973784672,
+      "learning_rate": 1.7347330542483527e-06,
+      "loss": 0.6758,
+      "step": 8723
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.763836633188647,
+      "learning_rate": 1.7346675162298907e-06,
+      "loss": 0.6868,
+      "step": 8724
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.078072228822213,
+      "learning_rate": 1.7346019713547121e-06,
+      "loss": 0.6836,
+      "step": 8725
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.2415446583076757,
+      "learning_rate": 1.7345364196234283e-06,
+      "loss": 0.666,
+      "step": 8726
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.149863561297166,
+      "learning_rate": 1.7344708610366513e-06,
+      "loss": 0.6569,
+      "step": 8727
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6856837107617393,
+      "learning_rate": 1.7344052955949925e-06,
+      "loss": 0.681,
+      "step": 8728
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.869814296368122,
+      "learning_rate": 1.7343397232990639e-06,
+      "loss": 0.6602,
+      "step": 8729
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.3809752666395663,
+      "learning_rate": 1.734274144149478e-06,
+      "loss": 0.6966,
+      "step": 8730
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.053519122936646,
+      "learning_rate": 1.734208558146847e-06,
+      "loss": 0.6693,
+      "step": 8731
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.031355299793135,
+      "learning_rate": 1.734142965291782e-06,
+      "loss": 0.6875,
+      "step": 8732
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.53050077064896,
+      "learning_rate": 1.734077365584896e-06,
+      "loss": 0.6875,
+      "step": 8733
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.348940896979694,
+      "learning_rate": 1.734011759026801e-06,
+      "loss": 0.6738,
+      "step": 8734
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.082292300900019,
+      "learning_rate": 1.7339461456181094e-06,
+      "loss": 0.6725,
+      "step": 8735
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6367561327672218,
+      "learning_rate": 1.7338805253594336e-06,
+      "loss": 0.6615,
+      "step": 8736
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7630645943007089,
+      "learning_rate": 1.733814898251386e-06,
+      "loss": 0.6576,
+      "step": 8737
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.015022626700978,
+      "learning_rate": 1.7337492642945786e-06,
+      "loss": 0.6868,
+      "step": 8738
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.764733522115458,
+      "learning_rate": 1.7336836234896249e-06,
+      "loss": 0.666,
+      "step": 8739
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.945018509097129,
+      "learning_rate": 1.7336179758371365e-06,
+      "loss": 0.6875,
+      "step": 8740
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.103060267797686,
+      "learning_rate": 1.7335523213377272e-06,
+      "loss": 0.6947,
+      "step": 8741
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.7893749942705137,
+      "learning_rate": 1.733486659992009e-06,
+      "loss": 0.6738,
+      "step": 8742
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8073967038313508,
+      "learning_rate": 1.733420991800595e-06,
+      "loss": 0.6868,
+      "step": 8743
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5582940260119535,
+      "learning_rate": 1.733355316764098e-06,
+      "loss": 0.6927,
+      "step": 8744
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.1357760714649452,
+      "learning_rate": 1.733289634883131e-06,
+      "loss": 0.681,
+      "step": 8745
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6953873607791772,
+      "learning_rate": 1.7332239461583069e-06,
+      "loss": 0.668,
+      "step": 8746
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 3.213975427753867,
+      "learning_rate": 1.7331582505902392e-06,
+      "loss": 0.7188,
+      "step": 8747
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.544165665782981,
+      "learning_rate": 1.7330925481795401e-06,
+      "loss": 0.7038,
+      "step": 8748
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.0097536983474775,
+      "learning_rate": 1.7330268389268238e-06,
+      "loss": 0.6751,
+      "step": 8749
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.132360379088005,
+      "learning_rate": 1.7329611228327032e-06,
+      "loss": 0.6706,
+      "step": 8750
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7264294314559426,
+      "learning_rate": 1.7328953998977915e-06,
+      "loss": 0.6693,
+      "step": 8751
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.262001565668996,
+      "learning_rate": 1.7328296701227023e-06,
+      "loss": 0.6842,
+      "step": 8752
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6458084443989123,
+      "learning_rate": 1.732763933508049e-06,
+      "loss": 0.6712,
+      "step": 8753
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6676803155124256,
+      "learning_rate": 1.732698190054445e-06,
+      "loss": 0.6589,
+      "step": 8754
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.60158415297403,
+      "learning_rate": 1.7326324397625043e-06,
+      "loss": 0.6764,
+      "step": 8755
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7065413562882679,
+      "learning_rate": 1.7325666826328397e-06,
+      "loss": 0.6862,
+      "step": 8756
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.5092030985363736,
+      "learning_rate": 1.732500918666066e-06,
+      "loss": 0.7077,
+      "step": 8757
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8455426533221444,
+      "learning_rate": 1.7324351478627963e-06,
+      "loss": 0.6882,
+      "step": 8758
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.1377082761407857,
+      "learning_rate": 1.7323693702236444e-06,
+      "loss": 0.6654,
+      "step": 8759
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.501693476572139,
+      "learning_rate": 1.7323035857492245e-06,
+      "loss": 0.6888,
+      "step": 8760
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.096526735235738,
+      "learning_rate": 1.7322377944401505e-06,
+      "loss": 0.6836,
+      "step": 8761
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.1829377107060752,
+      "learning_rate": 1.7321719962970364e-06,
+      "loss": 0.6725,
+      "step": 8762
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.17619494895312,
+      "learning_rate": 1.7321061913204961e-06,
+      "loss": 0.7096,
+      "step": 8763
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.792708162060531,
+      "learning_rate": 1.7320403795111445e-06,
+      "loss": 0.6927,
+      "step": 8764
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.406465476165838,
+      "learning_rate": 1.7319745608695949e-06,
+      "loss": 0.6719,
+      "step": 8765
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.309530222601276,
+      "learning_rate": 1.7319087353964622e-06,
+      "loss": 0.6706,
+      "step": 8766
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6506438277895225,
+      "learning_rate": 1.7318429030923602e-06,
+      "loss": 0.6621,
+      "step": 8767
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.642996833054726,
+      "learning_rate": 1.7317770639579038e-06,
+      "loss": 0.668,
+      "step": 8768
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.4150116397612136,
+      "learning_rate": 1.7317112179937077e-06,
+      "loss": 0.6895,
+      "step": 8769
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.793736012766643,
+      "learning_rate": 1.7316453652003858e-06,
+      "loss": 0.6966,
+      "step": 8770
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.2481956908295462,
+      "learning_rate": 1.7315795055785529e-06,
+      "loss": 0.6914,
+      "step": 8771
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.2894044121976247,
+      "learning_rate": 1.731513639128824e-06,
+      "loss": 0.6628,
+      "step": 8772
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6046972370431685,
+      "learning_rate": 1.7314477658518136e-06,
+      "loss": 0.6686,
+      "step": 8773
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9419026595249262,
+      "learning_rate": 1.7313818857481364e-06,
+      "loss": 0.6543,
+      "step": 8774
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.2929727748964135,
+      "learning_rate": 1.7313159988184073e-06,
+      "loss": 0.6764,
+      "step": 8775
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.1695867997158182,
+      "learning_rate": 1.7312501050632414e-06,
+      "loss": 0.6699,
+      "step": 8776
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.516213890860853,
+      "learning_rate": 1.7311842044832536e-06,
+      "loss": 0.6947,
+      "step": 8777
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.938374668105928,
+      "learning_rate": 1.731118297079059e-06,
+      "loss": 0.696,
+      "step": 8778
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6090860249218704,
+      "learning_rate": 1.7310523828512729e-06,
+      "loss": 0.6745,
+      "step": 8779
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6695876715514129,
+      "learning_rate": 1.7309864618005098e-06,
+      "loss": 0.7227,
+      "step": 8780
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3421695492315915,
+      "learning_rate": 1.7309205339273855e-06,
+      "loss": 0.6582,
+      "step": 8781
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6231930666731943,
+      "learning_rate": 1.7308545992325154e-06,
+      "loss": 0.6641,
+      "step": 8782
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0798289190483816,
+      "learning_rate": 1.7307886577165148e-06,
+      "loss": 0.6387,
+      "step": 8783
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8075747139747669,
+      "learning_rate": 1.7307227093799988e-06,
+      "loss": 0.6829,
+      "step": 8784
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9408456653316726,
+      "learning_rate": 1.7306567542235833e-06,
+      "loss": 0.6426,
+      "step": 8785
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.746036704031865,
+      "learning_rate": 1.7305907922478836e-06,
+      "loss": 0.6914,
+      "step": 8786
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.7895115780652353,
+      "learning_rate": 1.7305248234535156e-06,
+      "loss": 0.6849,
+      "step": 8787
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 5.665032227658894,
+      "learning_rate": 1.7304588478410945e-06,
+      "loss": 0.6816,
+      "step": 8788
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7330684481812412,
+      "learning_rate": 1.7303928654112364e-06,
+      "loss": 0.668,
+      "step": 8789
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0858575120099037,
+      "learning_rate": 1.7303268761645572e-06,
+      "loss": 0.6784,
+      "step": 8790
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.3811784318036375,
+      "learning_rate": 1.7302608801016727e-06,
+      "loss": 0.6777,
+      "step": 8791
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7366448046178056,
+      "learning_rate": 1.7301948772231988e-06,
+      "loss": 0.6803,
+      "step": 8792
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.662478302697057,
+      "learning_rate": 1.7301288675297515e-06,
+      "loss": 0.6673,
+      "step": 8793
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.6038862071963496,
+      "learning_rate": 1.730062851021947e-06,
+      "loss": 0.6615,
+      "step": 8794
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.2796677297832253,
+      "learning_rate": 1.7299968277004014e-06,
+      "loss": 0.6862,
+      "step": 8795
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8293653164791005,
+      "learning_rate": 1.7299307975657307e-06,
+      "loss": 0.6445,
+      "step": 8796
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.270043409634695,
+      "learning_rate": 1.7298647606185513e-06,
+      "loss": 0.6602,
+      "step": 8797
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.7367892257334551,
+      "learning_rate": 1.7297987168594796e-06,
+      "loss": 0.6478,
+      "step": 8798
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.325972397404438,
+      "learning_rate": 1.729732666289132e-06,
+      "loss": 0.6686,
+      "step": 8799
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.7246127718698703,
+      "learning_rate": 1.7296666089081247e-06,
+      "loss": 0.6602,
+      "step": 8800
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6692165898234563,
+      "learning_rate": 1.7296005447170746e-06,
+      "loss": 0.6849,
+      "step": 8801
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.570445799429104,
+      "learning_rate": 1.729534473716598e-06,
+      "loss": 0.6751,
+      "step": 8802
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2538519743644607,
+      "learning_rate": 1.7294683959073118e-06,
+      "loss": 0.6823,
+      "step": 8803
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.971970884688457,
+      "learning_rate": 1.7294023112898324e-06,
+      "loss": 0.6797,
+      "step": 8804
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.9892619208855646,
+      "learning_rate": 1.7293362198647767e-06,
+      "loss": 0.6927,
+      "step": 8805
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.2512484747594534,
+      "learning_rate": 1.7292701216327616e-06,
+      "loss": 0.6836,
+      "step": 8806
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.616089950621106,
+      "learning_rate": 1.729204016594404e-06,
+      "loss": 0.7135,
+      "step": 8807
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.3096177416523338,
+      "learning_rate": 1.729137904750321e-06,
+      "loss": 0.6582,
+      "step": 8808
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.2663430258016568,
+      "learning_rate": 1.7290717861011292e-06,
+      "loss": 0.668,
+      "step": 8809
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 10.532387295330262,
+      "learning_rate": 1.7290056606474458e-06,
+      "loss": 0.709,
+      "step": 8810
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8276412091954252,
+      "learning_rate": 1.7289395283898885e-06,
+      "loss": 0.6413,
+      "step": 8811
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 7.047795965201187,
+      "learning_rate": 1.7288733893290738e-06,
+      "loss": 0.6966,
+      "step": 8812
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.2277070043818499,
+      "learning_rate": 1.7288072434656193e-06,
+      "loss": 0.6693,
+      "step": 8813
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0181676698722704,
+      "learning_rate": 1.7287410908001427e-06,
+      "loss": 0.6719,
+      "step": 8814
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.1475741581814176,
+      "learning_rate": 1.7286749313332607e-06,
+      "loss": 0.696,
+      "step": 8815
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.761581955998432,
+      "learning_rate": 1.7286087650655913e-06,
+      "loss": 0.7044,
+      "step": 8816
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 3.793193886477874,
+      "learning_rate": 1.7285425919977517e-06,
+      "loss": 0.6842,
+      "step": 8817
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.472404567877617,
+      "learning_rate": 1.7284764121303599e-06,
+      "loss": 0.6816,
+      "step": 8818
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.322948390153908,
+      "learning_rate": 1.7284102254640332e-06,
+      "loss": 0.6803,
+      "step": 8819
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.2262786865714033,
+      "learning_rate": 1.7283440319993892e-06,
+      "loss": 0.6842,
+      "step": 8820
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4032273959672943,
+      "learning_rate": 1.7282778317370461e-06,
+      "loss": 0.6699,
+      "step": 8821
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6172637314183689,
+      "learning_rate": 1.728211624677622e-06,
+      "loss": 0.6829,
+      "step": 8822
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.7704463771251495,
+      "learning_rate": 1.728145410821734e-06,
+      "loss": 0.6882,
+      "step": 8823
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.752821834819236,
+      "learning_rate": 1.7280791901700003e-06,
+      "loss": 0.6667,
+      "step": 8824
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.5211493736056285,
+      "learning_rate": 1.7280129627230396e-06,
+      "loss": 0.6823,
+      "step": 8825
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.074552822050978,
+      "learning_rate": 1.7279467284814693e-06,
+      "loss": 0.6608,
+      "step": 8826
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.509814231368676,
+      "learning_rate": 1.7278804874459078e-06,
+      "loss": 0.6523,
+      "step": 8827
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.354120908684249,
+      "learning_rate": 1.7278142396169732e-06,
+      "loss": 0.6719,
+      "step": 8828
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.2789932232434846,
+      "learning_rate": 1.7277479849952842e-06,
+      "loss": 0.6608,
+      "step": 8829
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7642396549000932,
+      "learning_rate": 1.727681723581459e-06,
+      "loss": 0.6543,
+      "step": 8830
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4931908921745864,
+      "learning_rate": 1.7276154553761155e-06,
+      "loss": 0.6628,
+      "step": 8831
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.2458546730350175,
+      "learning_rate": 1.727549180379873e-06,
+      "loss": 0.6686,
+      "step": 8832
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3218320176732796,
+      "learning_rate": 1.7274828985933493e-06,
+      "loss": 0.6517,
+      "step": 8833
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6610253919258459,
+      "learning_rate": 1.7274166100171637e-06,
+      "loss": 0.6908,
+      "step": 8834
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.9143594586844466,
+      "learning_rate": 1.7273503146519343e-06,
+      "loss": 0.6621,
+      "step": 8835
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.204021390281893,
+      "learning_rate": 1.7272840124982802e-06,
+      "loss": 0.6719,
+      "step": 8836
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.35358689688705,
+      "learning_rate": 1.7272177035568203e-06,
+      "loss": 0.6842,
+      "step": 8837
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.257972536602473,
+      "learning_rate": 1.727151387828173e-06,
+      "loss": 0.6908,
+      "step": 8838
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9515137471920472,
+      "learning_rate": 1.7270850653129574e-06,
+      "loss": 0.6758,
+      "step": 8839
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3344057620165082,
+      "learning_rate": 1.7270187360117927e-06,
+      "loss": 0.6556,
+      "step": 8840
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.0141991366948298,
+      "learning_rate": 1.726952399925298e-06,
+      "loss": 0.6595,
+      "step": 8841
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.7674301586904955,
+      "learning_rate": 1.726886057054092e-06,
+      "loss": 0.6797,
+      "step": 8842
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9421841285468386,
+      "learning_rate": 1.7268197073987945e-06,
+      "loss": 0.6823,
+      "step": 8843
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.718578400755337,
+      "learning_rate": 1.726753350960024e-06,
+      "loss": 0.6895,
+      "step": 8844
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7499969339068997,
+      "learning_rate": 1.7266869877384002e-06,
+      "loss": 0.6491,
+      "step": 8845
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.6011021176138267,
+      "learning_rate": 1.726620617734543e-06,
+      "loss": 0.6536,
+      "step": 8846
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8786071417302911,
+      "learning_rate": 1.726554240949071e-06,
+      "loss": 0.6523,
+      "step": 8847
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9079358381779234,
+      "learning_rate": 1.7264878573826037e-06,
+      "loss": 0.6732,
+      "step": 8848
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 6.905143704816516,
+      "learning_rate": 1.7264214670357613e-06,
+      "loss": 0.6992,
+      "step": 8849
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.7374506965879248,
+      "learning_rate": 1.726355069909163e-06,
+      "loss": 0.681,
+      "step": 8850
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5982772796360636,
+      "learning_rate": 1.726288666003429e-06,
+      "loss": 0.6914,
+      "step": 8851
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.49288101615905,
+      "learning_rate": 1.7262222553191784e-06,
+      "loss": 0.6888,
+      "step": 8852
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.724715173183867,
+      "learning_rate": 1.7261558378570312e-06,
+      "loss": 0.7168,
+      "step": 8853
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.82568094355982,
+      "learning_rate": 1.7260894136176072e-06,
+      "loss": 0.6504,
+      "step": 8854
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.691939716701735,
+      "learning_rate": 1.7260229826015266e-06,
+      "loss": 0.6712,
+      "step": 8855
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.2936571956201168,
+      "learning_rate": 1.7259565448094094e-06,
+      "loss": 0.6517,
+      "step": 8856
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.786189050787584,
+      "learning_rate": 1.7258901002418753e-06,
+      "loss": 0.6777,
+      "step": 8857
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.322120231698932,
+      "learning_rate": 1.725823648899545e-06,
+      "loss": 0.6901,
+      "step": 8858
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.2151469451858095,
+      "learning_rate": 1.7257571907830383e-06,
+      "loss": 0.6868,
+      "step": 8859
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8478687153285074,
+      "learning_rate": 1.7256907258929756e-06,
+      "loss": 0.666,
+      "step": 8860
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7828230892233147,
+      "learning_rate": 1.7256242542299772e-06,
+      "loss": 0.6536,
+      "step": 8861
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.9979085334748747,
+      "learning_rate": 1.7255577757946633e-06,
+      "loss": 0.6862,
+      "step": 8862
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.1892228103731854,
+      "learning_rate": 1.7254912905876547e-06,
+      "loss": 0.6862,
+      "step": 8863
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.188542450658862,
+      "learning_rate": 1.7254247986095715e-06,
+      "loss": 0.6615,
+      "step": 8864
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.794873462052883,
+      "learning_rate": 1.7253582998610345e-06,
+      "loss": 0.6862,
+      "step": 8865
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.7327366212053796,
+      "learning_rate": 1.7252917943426648e-06,
+      "loss": 0.6771,
+      "step": 8866
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.0358409935197996,
+      "learning_rate": 1.7252252820550822e-06,
+      "loss": 0.6608,
+      "step": 8867
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7723390622333534,
+      "learning_rate": 1.725158762998908e-06,
+      "loss": 0.6699,
+      "step": 8868
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.1653314200966682,
+      "learning_rate": 1.725092237174763e-06,
+      "loss": 0.6738,
+      "step": 8869
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.6681059897154251,
+      "learning_rate": 1.7250257045832678e-06,
+      "loss": 0.696,
+      "step": 8870
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.031042182273189,
+      "learning_rate": 1.724959165225044e-06,
+      "loss": 0.6842,
+      "step": 8871
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 7.331151586476754,
+      "learning_rate": 1.7248926191007118e-06,
+      "loss": 0.6745,
+      "step": 8872
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.942590142772518,
+      "learning_rate": 1.7248260662108929e-06,
+      "loss": 0.6706,
+      "step": 8873
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.92331399125714,
+      "learning_rate": 1.7247595065562083e-06,
+      "loss": 0.6973,
+      "step": 8874
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8686324440851039,
+      "learning_rate": 1.7246929401372788e-06,
+      "loss": 0.6745,
+      "step": 8875
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.063176679380842,
+      "learning_rate": 1.7246263669547263e-06,
+      "loss": 0.6855,
+      "step": 8876
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.5072139027826363,
+      "learning_rate": 1.7245597870091718e-06,
+      "loss": 0.6667,
+      "step": 8877
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.755698548754282,
+      "learning_rate": 1.7244932003012365e-06,
+      "loss": 0.6836,
+      "step": 8878
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 3.168152652898917,
+      "learning_rate": 1.7244266068315423e-06,
+      "loss": 0.6999,
+      "step": 8879
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5914601078408233,
+      "learning_rate": 1.7243600066007104e-06,
+      "loss": 0.6732,
+      "step": 8880
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.7269442003154436,
+      "learning_rate": 1.7242933996093627e-06,
+      "loss": 0.6829,
+      "step": 8881
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 5.956251529599911,
+      "learning_rate": 1.7242267858581206e-06,
+      "loss": 0.681,
+      "step": 8882
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 4.542881693866246,
+      "learning_rate": 1.7241601653476058e-06,
+      "loss": 0.6615,
+      "step": 8883
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.947873306369048,
+      "learning_rate": 1.72409353807844e-06,
+      "loss": 0.6576,
+      "step": 8884
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.057290921221241,
+      "learning_rate": 1.7240269040512452e-06,
+      "loss": 0.6992,
+      "step": 8885
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7908925774558436,
+      "learning_rate": 1.7239602632666435e-06,
+      "loss": 0.6667,
+      "step": 8886
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.466658879034757,
+      "learning_rate": 1.7238936157252564e-06,
+      "loss": 0.7012,
+      "step": 8887
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7707319025035271,
+      "learning_rate": 1.723826961427706e-06,
+      "loss": 0.6634,
+      "step": 8888
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.58518602616954,
+      "learning_rate": 1.723760300374615e-06,
+      "loss": 0.7096,
+      "step": 8889
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.9002180163407503,
+      "learning_rate": 1.723693632566605e-06,
+      "loss": 0.6803,
+      "step": 8890
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3758428543721204,
+      "learning_rate": 1.7236269580042983e-06,
+      "loss": 0.6673,
+      "step": 8891
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.1519109447752305,
+      "learning_rate": 1.723560276688317e-06,
+      "loss": 0.6777,
+      "step": 8892
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8467363705955273,
+      "learning_rate": 1.723493588619284e-06,
+      "loss": 0.638,
+      "step": 8893
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7214537435057161,
+      "learning_rate": 1.723426893797821e-06,
+      "loss": 0.668,
+      "step": 8894
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.681073888970992,
+      "learning_rate": 1.723360192224551e-06,
+      "loss": 0.6699,
+      "step": 8895
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8049452680503447,
+      "learning_rate": 1.7232934839000962e-06,
+      "loss": 0.6934,
+      "step": 8896
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9155597768029925,
+      "learning_rate": 1.72322676882508e-06,
+      "loss": 0.6888,
+      "step": 8897
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.7875551277240542,
+      "learning_rate": 1.7231600470001238e-06,
+      "loss": 0.6816,
+      "step": 8898
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.815487254358227,
+      "learning_rate": 1.7230933184258509e-06,
+      "loss": 0.6621,
+      "step": 8899
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.848603312303086,
+      "learning_rate": 1.723026583102884e-06,
+      "loss": 0.6725,
+      "step": 8900
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.080170171640607,
+      "learning_rate": 1.7229598410318467e-06,
+      "loss": 0.6712,
+      "step": 8901
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.5944293916933554,
+      "learning_rate": 1.7228930922133608e-06,
+      "loss": 0.696,
+      "step": 8902
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.375237080500811,
+      "learning_rate": 1.7228263366480497e-06,
+      "loss": 0.6758,
+      "step": 8903
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3231627018656,
+      "learning_rate": 1.7227595743365367e-06,
+      "loss": 0.6751,
+      "step": 8904
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.928961404732532,
+      "learning_rate": 1.7226928052794445e-06,
+      "loss": 0.6569,
+      "step": 8905
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8805288856145922,
+      "learning_rate": 1.7226260294773964e-06,
+      "loss": 0.6523,
+      "step": 8906
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8145969214093784,
+      "learning_rate": 1.722559246931016e-06,
+      "loss": 0.6615,
+      "step": 8907
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9340631496124185,
+      "learning_rate": 1.7224924576409258e-06,
+      "loss": 0.6706,
+      "step": 8908
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.4959148367664916,
+      "learning_rate": 1.72242566160775e-06,
+      "loss": 0.6745,
+      "step": 8909
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.9246124123760664,
+      "learning_rate": 1.7223588588321115e-06,
+      "loss": 0.6777,
+      "step": 8910
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.2117500490849693,
+      "learning_rate": 1.7222920493146336e-06,
+      "loss": 0.6882,
+      "step": 8911
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.627028522380986,
+      "learning_rate": 1.7222252330559403e-06,
+      "loss": 0.7279,
+      "step": 8912
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0943449256372277,
+      "learning_rate": 1.722158410056655e-06,
+      "loss": 0.6875,
+      "step": 8913
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1980721884834362,
+      "learning_rate": 1.7220915803174014e-06,
+      "loss": 0.6758,
+      "step": 8914
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.0099090905423724,
+      "learning_rate": 1.7220247438388037e-06,
+      "loss": 0.694,
+      "step": 8915
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.281954987246569,
+      "learning_rate": 1.7219579006214845e-06,
+      "loss": 0.6836,
+      "step": 8916
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.526386016605229,
+      "learning_rate": 1.721891050666069e-06,
+      "loss": 0.6602,
+      "step": 8917
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3694532711039822,
+      "learning_rate": 1.7218241939731798e-06,
+      "loss": 0.6868,
+      "step": 8918
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.631463680950229,
+      "learning_rate": 1.7217573305434421e-06,
+      "loss": 0.6693,
+      "step": 8919
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.088072656772336,
+      "learning_rate": 1.721690460377479e-06,
+      "loss": 0.681,
+      "step": 8920
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.0426219532016585,
+      "learning_rate": 1.7216235834759155e-06,
+      "loss": 0.666,
+      "step": 8921
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8122273206066355,
+      "learning_rate": 1.7215566998393748e-06,
+      "loss": 0.6803,
+      "step": 8922
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.222676558823178,
+      "learning_rate": 1.721489809468482e-06,
+      "loss": 0.6654,
+      "step": 8923
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.3833223822896643,
+      "learning_rate": 1.7214229123638608e-06,
+      "loss": 0.7031,
+      "step": 8924
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.9727163282368998,
+      "learning_rate": 1.7213560085261358e-06,
+      "loss": 0.7005,
+      "step": 8925
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.8045025472350986,
+      "learning_rate": 1.7212890979559314e-06,
+      "loss": 0.679,
+      "step": 8926
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1839241453551494,
+      "learning_rate": 1.7212221806538719e-06,
+      "loss": 0.6549,
+      "step": 8927
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.022170277905354,
+      "learning_rate": 1.7211552566205821e-06,
+      "loss": 0.6934,
+      "step": 8928
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.945872877354285,
+      "learning_rate": 1.7210883258566864e-06,
+      "loss": 0.6842,
+      "step": 8929
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.0462772155134767,
+      "learning_rate": 1.7210213883628098e-06,
+      "loss": 0.6895,
+      "step": 8930
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.7245322848630193,
+      "learning_rate": 1.7209544441395769e-06,
+      "loss": 0.6693,
+      "step": 8931
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.0098884328115327,
+      "learning_rate": 1.7208874931876123e-06,
+      "loss": 0.6777,
+      "step": 8932
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 5.655052754527628,
+      "learning_rate": 1.7208205355075406e-06,
+      "loss": 0.6999,
+      "step": 8933
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.558723715190061,
+      "learning_rate": 1.7207535710999877e-06,
+      "loss": 0.6654,
+      "step": 8934
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8786926064027464,
+      "learning_rate": 1.7206865999655776e-06,
+      "loss": 0.6784,
+      "step": 8935
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 3.540243090942768,
+      "learning_rate": 1.7206196221049359e-06,
+      "loss": 0.6602,
+      "step": 8936
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9562317582188503,
+      "learning_rate": 1.7205526375186875e-06,
+      "loss": 0.7116,
+      "step": 8937
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.752739640187885,
+      "learning_rate": 1.7204856462074575e-06,
+      "loss": 0.6764,
+      "step": 8938
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.062872676567447,
+      "learning_rate": 1.7204186481718712e-06,
+      "loss": 0.6829,
+      "step": 8939
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.292550801676416,
+      "learning_rate": 1.7203516434125543e-06,
+      "loss": 0.6582,
+      "step": 8940
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.0596487887995374,
+      "learning_rate": 1.7202846319301317e-06,
+      "loss": 0.6738,
+      "step": 8941
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.852387304909222,
+      "learning_rate": 1.7202176137252287e-06,
+      "loss": 0.6914,
+      "step": 8942
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.1730734840099117,
+      "learning_rate": 1.720150588798471e-06,
+      "loss": 0.6745,
+      "step": 8943
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.235266270874155,
+      "learning_rate": 1.7200835571504843e-06,
+      "loss": 0.6888,
+      "step": 8944
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1521314577594863,
+      "learning_rate": 1.7200165187818942e-06,
+      "loss": 0.6654,
+      "step": 8945
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.0906972084446815,
+      "learning_rate": 1.719949473693326e-06,
+      "loss": 0.6452,
+      "step": 8946
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7090749711913243,
+      "learning_rate": 1.7198824218854058e-06,
+      "loss": 0.6862,
+      "step": 8947
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.850040034957921,
+      "learning_rate": 1.7198153633587595e-06,
+      "loss": 0.6484,
+      "step": 8948
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.083662407807541,
+      "learning_rate": 1.7197482981140126e-06,
+      "loss": 0.6608,
+      "step": 8949
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.892679997056147,
+      "learning_rate": 1.7196812261517912e-06,
+      "loss": 0.6712,
+      "step": 8950
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.257342204904894,
+      "learning_rate": 1.719614147472721e-06,
+      "loss": 0.6829,
+      "step": 8951
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.219260950588435,
+      "learning_rate": 1.7195470620774288e-06,
+      "loss": 0.6771,
+      "step": 8952
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.9024747058986833,
+      "learning_rate": 1.7194799699665399e-06,
+      "loss": 0.6348,
+      "step": 8953
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9710137471050568,
+      "learning_rate": 1.719412871140681e-06,
+      "loss": 0.7259,
+      "step": 8954
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.338944812860526,
+      "learning_rate": 1.7193457656004778e-06,
+      "loss": 0.6719,
+      "step": 8955
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.6506175841493045,
+      "learning_rate": 1.7192786533465573e-06,
+      "loss": 0.681,
+      "step": 8956
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.253399055398987,
+      "learning_rate": 1.7192115343795454e-06,
+      "loss": 0.7083,
+      "step": 8957
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0840830239781793,
+      "learning_rate": 1.7191444087000688e-06,
+      "loss": 0.651,
+      "step": 8958
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.6009415904154167,
+      "learning_rate": 1.7190772763087538e-06,
+      "loss": 0.6784,
+      "step": 8959
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.0742297735226773,
+      "learning_rate": 1.719010137206227e-06,
+      "loss": 0.6745,
+      "step": 8960
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.2027862734183685,
+      "learning_rate": 1.7189429913931147e-06,
+      "loss": 0.679,
+      "step": 8961
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.084060899449546,
+      "learning_rate": 1.718875838870044e-06,
+      "loss": 0.694,
+      "step": 8962
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.54030144316122,
+      "learning_rate": 1.7188086796376417e-06,
+      "loss": 0.6934,
+      "step": 8963
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2235415731819725,
+      "learning_rate": 1.7187415136965341e-06,
+      "loss": 0.6634,
+      "step": 8964
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.884145322856984,
+      "learning_rate": 1.718674341047349e-06,
+      "loss": 0.722,
+      "step": 8965
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.410164089873417,
+      "learning_rate": 1.7186071616907124e-06,
+      "loss": 0.6452,
+      "step": 8966
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.506247751634845,
+      "learning_rate": 1.7185399756272514e-06,
+      "loss": 0.6764,
+      "step": 8967
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.352671192525237,
+      "learning_rate": 1.7184727828575934e-06,
+      "loss": 0.6875,
+      "step": 8968
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.768524571885998,
+      "learning_rate": 1.7184055833823655e-06,
+      "loss": 0.6816,
+      "step": 8969
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.1496836739007787,
+      "learning_rate": 1.7183383772021947e-06,
+      "loss": 0.6934,
+      "step": 8970
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6178470161154719,
+      "learning_rate": 1.7182711643177082e-06,
+      "loss": 0.6797,
+      "step": 8971
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 7.741940453811778,
+      "learning_rate": 1.7182039447295334e-06,
+      "loss": 0.6992,
+      "step": 8972
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3112629370397975,
+      "learning_rate": 1.7181367184382975e-06,
+      "loss": 0.6647,
+      "step": 8973
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.553267187452984,
+      "learning_rate": 1.7180694854446284e-06,
+      "loss": 0.6784,
+      "step": 8974
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.661821915354852,
+      "learning_rate": 1.7180022457491535e-06,
+      "loss": 0.6686,
+      "step": 8975
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0747408209109655,
+      "learning_rate": 1.7179349993524999e-06,
+      "loss": 0.6562,
+      "step": 8976
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8771873195393747,
+      "learning_rate": 1.7178677462552953e-06,
+      "loss": 0.6966,
+      "step": 8977
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.5189881553442321,
+      "learning_rate": 1.7178004864581675e-06,
+      "loss": 0.6621,
+      "step": 8978
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.078066732786764,
+      "learning_rate": 1.7177332199617445e-06,
+      "loss": 0.7181,
+      "step": 8979
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8254588586703813,
+      "learning_rate": 1.7176659467666536e-06,
+      "loss": 0.6868,
+      "step": 8980
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2976231202819928,
+      "learning_rate": 1.7175986668735234e-06,
+      "loss": 0.6556,
+      "step": 8981
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.640736012450371,
+      "learning_rate": 1.7175313802829812e-06,
+      "loss": 0.6693,
+      "step": 8982
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3723374220918134,
+      "learning_rate": 1.717464086995655e-06,
+      "loss": 0.6706,
+      "step": 8983
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.215788691671679,
+      "learning_rate": 1.7173967870121733e-06,
+      "loss": 0.6647,
+      "step": 8984
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.867268714902981,
+      "learning_rate": 1.7173294803331637e-06,
+      "loss": 0.6615,
+      "step": 8985
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2745063296705421,
+      "learning_rate": 1.7172621669592547e-06,
+      "loss": 0.6562,
+      "step": 8986
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.403608603923426,
+      "learning_rate": 1.7171948468910746e-06,
+      "loss": 0.6465,
+      "step": 8987
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3433467692233598,
+      "learning_rate": 1.7171275201292512e-06,
+      "loss": 0.666,
+      "step": 8988
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5717171019738085,
+      "learning_rate": 1.7170601866744136e-06,
+      "loss": 0.6914,
+      "step": 8989
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.9930685814672735,
+      "learning_rate": 1.7169928465271898e-06,
+      "loss": 0.7096,
+      "step": 8990
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8856419301827728,
+      "learning_rate": 1.7169254996882082e-06,
+      "loss": 0.6947,
+      "step": 8991
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.544012088366,
+      "learning_rate": 1.7168581461580977e-06,
+      "loss": 0.6732,
+      "step": 8992
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.084472836823921,
+      "learning_rate": 1.7167907859374867e-06,
+      "loss": 0.679,
+      "step": 8993
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9000898613294417,
+      "learning_rate": 1.7167234190270038e-06,
+      "loss": 0.6901,
+      "step": 8994
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.178034544564859,
+      "learning_rate": 1.7166560454272778e-06,
+      "loss": 0.6914,
+      "step": 8995
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.375699448575205,
+      "learning_rate": 1.7165886651389378e-06,
+      "loss": 0.6634,
+      "step": 8996
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.0653827565566047,
+      "learning_rate": 1.7165212781626122e-06,
+      "loss": 0.6764,
+      "step": 8997
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 6.3482200380249125,
+      "learning_rate": 1.7164538844989301e-06,
+      "loss": 0.6771,
+      "step": 8998
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.083264404060306,
+      "learning_rate": 1.7163864841485208e-06,
+      "loss": 0.6667,
+      "step": 8999
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.074538824067115,
+      "learning_rate": 1.7163190771120132e-06,
+      "loss": 0.6693,
+      "step": 9000
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.7788532985539869,
+      "learning_rate": 1.7162516633900357e-06,
+      "loss": 0.6855,
+      "step": 9001
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 4.725358093385939,
+      "learning_rate": 1.7161842429832185e-06,
+      "loss": 0.6882,
+      "step": 9002
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.384930412897024,
+      "learning_rate": 1.7161168158921905e-06,
+      "loss": 0.7005,
+      "step": 9003
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.3553788760032313,
+      "learning_rate": 1.7160493821175806e-06,
+      "loss": 0.6784,
+      "step": 9004
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.1353755295399244,
+      "learning_rate": 1.7159819416600184e-06,
+      "loss": 0.6549,
+      "step": 9005
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.1368043743697673,
+      "learning_rate": 1.7159144945201337e-06,
+      "loss": 0.6576,
+      "step": 9006
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.9340387595810604,
+      "learning_rate": 1.7158470406985558e-06,
+      "loss": 0.6628,
+      "step": 9007
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7174178562105439,
+      "learning_rate": 1.7157795801959143e-06,
+      "loss": 0.6745,
+      "step": 9008
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.3276216843714925,
+      "learning_rate": 1.7157121130128384e-06,
+      "loss": 0.6868,
+      "step": 9009
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.1672819462742563,
+      "learning_rate": 1.7156446391499578e-06,
+      "loss": 0.6745,
+      "step": 9010
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.5629731749653266,
+      "learning_rate": 1.715577158607903e-06,
+      "loss": 0.6699,
+      "step": 9011
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.505547713333929,
+      "learning_rate": 1.715509671387303e-06,
+      "loss": 0.668,
+      "step": 9012
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.490603145544369,
+      "learning_rate": 1.715442177488788e-06,
+      "loss": 0.6562,
+      "step": 9013
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0346071929544385,
+      "learning_rate": 1.715374676912988e-06,
+      "loss": 0.6699,
+      "step": 9014
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.0249043576247887,
+      "learning_rate": 1.7153071696605328e-06,
+      "loss": 0.6615,
+      "step": 9015
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.8217915899058746,
+      "learning_rate": 1.7152396557320525e-06,
+      "loss": 0.6699,
+      "step": 9016
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.892651502458677,
+      "learning_rate": 1.7151721351281774e-06,
+      "loss": 0.668,
+      "step": 9017
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.9720393540597205,
+      "learning_rate": 1.7151046078495374e-06,
+      "loss": 0.6751,
+      "step": 9018
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0197153100111715,
+      "learning_rate": 1.7150370738967627e-06,
+      "loss": 0.6953,
+      "step": 9019
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.119644306477457,
+      "learning_rate": 1.714969533270484e-06,
+      "loss": 0.6927,
+      "step": 9020
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.9104934637935873,
+      "learning_rate": 1.7149019859713314e-06,
+      "loss": 0.6615,
+      "step": 9021
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.915249049429312,
+      "learning_rate": 1.714834431999935e-06,
+      "loss": 0.6758,
+      "step": 9022
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.4187473346374686,
+      "learning_rate": 1.714766871356926e-06,
+      "loss": 0.6934,
+      "step": 9023
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7675734208695795,
+      "learning_rate": 1.7146993040429345e-06,
+      "loss": 0.6784,
+      "step": 9024
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.427500884512883,
+      "learning_rate": 1.7146317300585911e-06,
+      "loss": 0.6797,
+      "step": 9025
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7647487942347952,
+      "learning_rate": 1.7145641494045266e-06,
+      "loss": 0.6647,
+      "step": 9026
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.109123530498625,
+      "learning_rate": 1.7144965620813718e-06,
+      "loss": 0.6432,
+      "step": 9027
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9875505046055676,
+      "learning_rate": 1.7144289680897574e-06,
+      "loss": 0.6764,
+      "step": 9028
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.874156930572412,
+      "learning_rate": 1.7143613674303141e-06,
+      "loss": 0.6549,
+      "step": 9029
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.97330863411936,
+      "learning_rate": 1.714293760103673e-06,
+      "loss": 0.666,
+      "step": 9030
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.211470045528498,
+      "learning_rate": 1.7142261461104652e-06,
+      "loss": 0.6615,
+      "step": 9031
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.8280705405206157,
+      "learning_rate": 1.7141585254513216e-06,
+      "loss": 0.6738,
+      "step": 9032
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8074704181700543,
+      "learning_rate": 1.7140908981268733e-06,
+      "loss": 0.6556,
+      "step": 9033
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.848355039218546,
+      "learning_rate": 1.7140232641377512e-06,
+      "loss": 0.6589,
+      "step": 9034
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6001004303403215,
+      "learning_rate": 1.7139556234845874e-06,
+      "loss": 0.6523,
+      "step": 9035
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.50740861022956,
+      "learning_rate": 1.7138879761680122e-06,
+      "loss": 0.6895,
+      "step": 9036
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.6293222014507354,
+      "learning_rate": 1.7138203221886576e-06,
+      "loss": 0.7012,
+      "step": 9037
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3326909990901106,
+      "learning_rate": 1.7137526615471546e-06,
+      "loss": 0.6595,
+      "step": 9038
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.041469772606308,
+      "learning_rate": 1.7136849942441352e-06,
+      "loss": 0.6758,
+      "step": 9039
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9683833265499198,
+      "learning_rate": 1.7136173202802304e-06,
+      "loss": 0.6823,
+      "step": 9040
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.938900594767453,
+      "learning_rate": 1.713549639656072e-06,
+      "loss": 0.6732,
+      "step": 9041
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.1086422282866333,
+      "learning_rate": 1.7134819523722919e-06,
+      "loss": 0.6523,
+      "step": 9042
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8357929820014929,
+      "learning_rate": 1.7134142584295215e-06,
+      "loss": 0.6686,
+      "step": 9043
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8671366356368536,
+      "learning_rate": 1.7133465578283927e-06,
+      "loss": 0.666,
+      "step": 9044
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 5.284288204450902,
+      "learning_rate": 1.7132788505695374e-06,
+      "loss": 0.6882,
+      "step": 9045
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8948781987477697,
+      "learning_rate": 1.7132111366535877e-06,
+      "loss": 0.6634,
+      "step": 9046
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.541112436110081,
+      "learning_rate": 1.713143416081175e-06,
+      "loss": 0.6771,
+      "step": 9047
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.2783511799684615,
+      "learning_rate": 1.7130756888529324e-06,
+      "loss": 0.6562,
+      "step": 9048
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.1659786065043574,
+      "learning_rate": 1.7130079549694906e-06,
+      "loss": 0.6478,
+      "step": 9049
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3106995162728103,
+      "learning_rate": 1.712940214431483e-06,
+      "loss": 0.6641,
+      "step": 9050
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9946008180456891,
+      "learning_rate": 1.7128724672395412e-06,
+      "loss": 0.6823,
+      "step": 9051
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.687744686691879,
+      "learning_rate": 1.7128047133942976e-06,
+      "loss": 0.6777,
+      "step": 9052
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3503451745715793,
+      "learning_rate": 1.7127369528963844e-06,
+      "loss": 0.6855,
+      "step": 9053
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.1352038693910043,
+      "learning_rate": 1.712669185746434e-06,
+      "loss": 0.7064,
+      "step": 9054
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.3773344460794612,
+      "learning_rate": 1.7126014119450796e-06,
+      "loss": 0.6888,
+      "step": 9055
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.511860599937161,
+      "learning_rate": 1.712533631492953e-06,
+      "loss": 0.6914,
+      "step": 9056
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.894526713962332,
+      "learning_rate": 1.712465844390687e-06,
+      "loss": 0.6823,
+      "step": 9057
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9740906651503238,
+      "learning_rate": 1.7123980506389144e-06,
+      "loss": 0.6621,
+      "step": 9058
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.749367763918816,
+      "learning_rate": 1.7123302502382676e-06,
+      "loss": 0.7025,
+      "step": 9059
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.3987918344718833,
+      "learning_rate": 1.7122624431893798e-06,
+      "loss": 0.6927,
+      "step": 9060
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.1280028527367096,
+      "learning_rate": 1.7121946294928834e-06,
+      "loss": 0.6712,
+      "step": 9061
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.350331840125984,
+      "learning_rate": 1.7121268091494114e-06,
+      "loss": 0.6895,
+      "step": 9062
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.4263239966174013,
+      "learning_rate": 1.7120589821595973e-06,
+      "loss": 0.6732,
+      "step": 9063
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6434971649063116,
+      "learning_rate": 1.7119911485240736e-06,
+      "loss": 0.6738,
+      "step": 9064
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.4170458487773365,
+      "learning_rate": 1.7119233082434734e-06,
+      "loss": 0.6686,
+      "step": 9065
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.9560868793371984,
+      "learning_rate": 1.7118554613184302e-06,
+      "loss": 0.6693,
+      "step": 9066
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7210832389355212,
+      "learning_rate": 1.711787607749577e-06,
+      "loss": 0.6745,
+      "step": 9067
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6597669433952074,
+      "learning_rate": 1.7117197475375468e-06,
+      "loss": 0.6777,
+      "step": 9068
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.8732359824918814,
+      "learning_rate": 1.7116518806829739e-06,
+      "loss": 0.6875,
+      "step": 9069
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.1662406421361484,
+      "learning_rate": 1.7115840071864904e-06,
+      "loss": 0.6771,
+      "step": 9070
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7226183457878842,
+      "learning_rate": 1.7115161270487306e-06,
+      "loss": 0.6836,
+      "step": 9071
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.2130716590572,
+      "learning_rate": 1.7114482402703281e-06,
+      "loss": 0.679,
+      "step": 9072
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8223146986778914,
+      "learning_rate": 1.7113803468519162e-06,
+      "loss": 0.6536,
+      "step": 9073
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.407373747077449,
+      "learning_rate": 1.7113124467941286e-06,
+      "loss": 0.6406,
+      "step": 9074
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.024380876847112,
+      "learning_rate": 1.7112445400975988e-06,
+      "loss": 0.6673,
+      "step": 9075
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.755740003041189,
+      "learning_rate": 1.711176626762961e-06,
+      "loss": 0.6562,
+      "step": 9076
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8439304433221655,
+      "learning_rate": 1.7111087067908487e-06,
+      "loss": 0.6849,
+      "step": 9077
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7354205028354484,
+      "learning_rate": 1.7110407801818962e-06,
+      "loss": 0.6641,
+      "step": 9078
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.7390369337098868,
+      "learning_rate": 1.710972846936737e-06,
+      "loss": 0.6934,
+      "step": 9079
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.9272767780723676,
+      "learning_rate": 1.7109049070560055e-06,
+      "loss": 0.6816,
+      "step": 9080
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 9.028404461263111,
+      "learning_rate": 1.7108369605403357e-06,
+      "loss": 0.6992,
+      "step": 9081
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7974605530165487,
+      "learning_rate": 1.7107690073903616e-06,
+      "loss": 0.6576,
+      "step": 9082
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 4.013495534104482,
+      "learning_rate": 1.7107010476067175e-06,
+      "loss": 0.7109,
+      "step": 9083
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.321939947300424,
+      "learning_rate": 1.7106330811900374e-06,
+      "loss": 0.6602,
+      "step": 9084
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.0039742453854488,
+      "learning_rate": 1.710565108140956e-06,
+      "loss": 0.6458,
+      "step": 9085
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.325559642864358,
+      "learning_rate": 1.710497128460108e-06,
+      "loss": 0.6862,
+      "step": 9086
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.878290321759715,
+      "learning_rate": 1.7104291421481272e-06,
+      "loss": 0.7038,
+      "step": 9087
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.1012525503769233,
+      "learning_rate": 1.7103611492056485e-06,
+      "loss": 0.6862,
+      "step": 9088
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.5048180602999537,
+      "learning_rate": 1.7102931496333062e-06,
+      "loss": 0.6862,
+      "step": 9089
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.351036201653561,
+      "learning_rate": 1.7102251434317352e-06,
+      "loss": 0.7109,
+      "step": 9090
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.818849148778638,
+      "learning_rate": 1.71015713060157e-06,
+      "loss": 0.6758,
+      "step": 9091
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8614200668414003,
+      "learning_rate": 1.7100891111434456e-06,
+      "loss": 0.6973,
+      "step": 9092
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.492782335610241,
+      "learning_rate": 1.7100210850579968e-06,
+      "loss": 0.6777,
+      "step": 9093
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.010033144454473,
+      "learning_rate": 1.7099530523458583e-06,
+      "loss": 0.679,
+      "step": 9094
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.5303941367958394,
+      "learning_rate": 1.7098850130076653e-06,
+      "loss": 0.6745,
+      "step": 9095
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.489790973865067,
+      "learning_rate": 1.7098169670440524e-06,
+      "loss": 0.6641,
+      "step": 9096
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.590792985774741,
+      "learning_rate": 1.7097489144556553e-06,
+      "loss": 0.6595,
+      "step": 9097
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8175106743161007,
+      "learning_rate": 1.7096808552431089e-06,
+      "loss": 0.6979,
+      "step": 9098
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8251667946520137,
+      "learning_rate": 1.709612789407048e-06,
+      "loss": 0.6966,
+      "step": 9099
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.717400213158441,
+      "learning_rate": 1.7095447169481082e-06,
+      "loss": 0.653,
+      "step": 9100
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6602295525474112,
+      "learning_rate": 1.709476637866925e-06,
+      "loss": 0.6745,
+      "step": 9101
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6393800728247847,
+      "learning_rate": 1.7094085521641338e-06,
+      "loss": 0.6758,
+      "step": 9102
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.9733177461150109,
+      "learning_rate": 1.7093404598403693e-06,
+      "loss": 0.6608,
+      "step": 9103
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.223699148599248,
+      "learning_rate": 1.7092723608962678e-06,
+      "loss": 0.6771,
+      "step": 9104
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.8676541560964033,
+      "learning_rate": 1.7092042553324646e-06,
+      "loss": 0.6719,
+      "step": 9105
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.5129203623465743,
+      "learning_rate": 1.7091361431495955e-06,
+      "loss": 0.7044,
+      "step": 9106
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.228429862898769,
+      "learning_rate": 1.709068024348296e-06,
+      "loss": 0.6628,
+      "step": 9107
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.54767917846934,
+      "learning_rate": 1.708999898929202e-06,
+      "loss": 0.6745,
+      "step": 9108
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8309529874787356,
+      "learning_rate": 1.7089317668929492e-06,
+      "loss": 0.6842,
+      "step": 9109
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 5.3499200550488455,
+      "learning_rate": 1.7088636282401735e-06,
+      "loss": 0.668,
+      "step": 9110
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7949153142228425,
+      "learning_rate": 1.7087954829715108e-06,
+      "loss": 0.6738,
+      "step": 9111
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.9300368443364615,
+      "learning_rate": 1.7087273310875972e-06,
+      "loss": 0.6875,
+      "step": 9112
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8036696797588058,
+      "learning_rate": 1.7086591725890686e-06,
+      "loss": 0.638,
+      "step": 9113
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.5640424537198738,
+      "learning_rate": 1.7085910074765616e-06,
+      "loss": 0.6901,
+      "step": 9114
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8000382282718134,
+      "learning_rate": 1.708522835750712e-06,
+      "loss": 0.6836,
+      "step": 9115
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.7430277704671875,
+      "learning_rate": 1.7084546574121558e-06,
+      "loss": 0.6738,
+      "step": 9116
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.946832770091644,
+      "learning_rate": 1.7083864724615297e-06,
+      "loss": 0.696,
+      "step": 9117
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.725773843208377,
+      "learning_rate": 1.7083182808994705e-06,
+      "loss": 0.6842,
+      "step": 9118
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.704531900905469,
+      "learning_rate": 1.7082500827266137e-06,
+      "loss": 0.6953,
+      "step": 9119
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.9647894785419417,
+      "learning_rate": 1.7081818779435964e-06,
+      "loss": 0.7044,
+      "step": 9120
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.50612295057099,
+      "learning_rate": 1.7081136665510547e-06,
+      "loss": 0.6771,
+      "step": 9121
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8464149537748066,
+      "learning_rate": 1.708045448549626e-06,
+      "loss": 0.6621,
+      "step": 9122
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8605808096500263,
+      "learning_rate": 1.7079772239399462e-06,
+      "loss": 0.6953,
+      "step": 9123
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.017125772260472,
+      "learning_rate": 1.7079089927226525e-06,
+      "loss": 0.6803,
+      "step": 9124
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 4.039000481028101,
+      "learning_rate": 1.7078407548983816e-06,
+      "loss": 0.6895,
+      "step": 9125
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9734383884144266,
+      "learning_rate": 1.7077725104677701e-06,
+      "loss": 0.6992,
+      "step": 9126
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.5348185884774788,
+      "learning_rate": 1.7077042594314554e-06,
+      "loss": 0.7031,
+      "step": 9127
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.435702693767827,
+      "learning_rate": 1.7076360017900742e-06,
+      "loss": 0.6621,
+      "step": 9128
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.3869823993023864,
+      "learning_rate": 1.7075677375442635e-06,
+      "loss": 0.6921,
+      "step": 9129
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.567544583706478,
+      "learning_rate": 1.7074994666946605e-06,
+      "loss": 0.6803,
+      "step": 9130
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.4869485978466845,
+      "learning_rate": 1.7074311892419026e-06,
+      "loss": 0.6686,
+      "step": 9131
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.3029160208151074,
+      "learning_rate": 1.7073629051866267e-06,
+      "loss": 0.6751,
+      "step": 9132
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.3148795680212846,
+      "learning_rate": 1.7072946145294704e-06,
+      "loss": 0.6771,
+      "step": 9133
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.2751413563973062,
+      "learning_rate": 1.7072263172710709e-06,
+      "loss": 0.6628,
+      "step": 9134
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.191143974686877,
+      "learning_rate": 1.7071580134120656e-06,
+      "loss": 0.7018,
+      "step": 9135
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 6.49999325251229,
+      "learning_rate": 1.707089702953092e-06,
+      "loss": 0.694,
+      "step": 9136
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.065365630302391,
+      "learning_rate": 1.7070213858947877e-06,
+      "loss": 0.6738,
+      "step": 9137
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.874218493775202,
+      "learning_rate": 1.7069530622377902e-06,
+      "loss": 0.681,
+      "step": 9138
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.531184477262752,
+      "learning_rate": 1.706884731982737e-06,
+      "loss": 0.6895,
+      "step": 9139
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7390628639342364,
+      "learning_rate": 1.7068163951302664e-06,
+      "loss": 0.6667,
+      "step": 9140
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.508200683477671,
+      "learning_rate": 1.706748051681016e-06,
+      "loss": 0.6966,
+      "step": 9141
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8896050996807681,
+      "learning_rate": 1.7066797016356233e-06,
+      "loss": 0.6784,
+      "step": 9142
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.5103271688703384,
+      "learning_rate": 1.7066113449947267e-06,
+      "loss": 0.6921,
+      "step": 9143
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.8714912983514442,
+      "learning_rate": 1.7065429817589638e-06,
+      "loss": 0.6712,
+      "step": 9144
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6069904328109078,
+      "learning_rate": 1.7064746119289728e-06,
+      "loss": 0.6719,
+      "step": 9145
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9651477494093823,
+      "learning_rate": 1.7064062355053917e-06,
+      "loss": 0.6549,
+      "step": 9146
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.4323148038383406,
+      "learning_rate": 1.7063378524888585e-06,
+      "loss": 0.6908,
+      "step": 9147
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.9316522834444787,
+      "learning_rate": 1.706269462880012e-06,
+      "loss": 0.6908,
+      "step": 9148
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0844973967009657,
+      "learning_rate": 1.70620106667949e-06,
+      "loss": 0.6738,
+      "step": 9149
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.4672287281384517,
+      "learning_rate": 1.706132663887931e-06,
+      "loss": 0.6621,
+      "step": 9150
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.6019516846017468,
+      "learning_rate": 1.7060642545059736e-06,
+      "loss": 0.6758,
+      "step": 9151
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.028753949485863,
+      "learning_rate": 1.7059958385342559e-06,
+      "loss": 0.6549,
+      "step": 9152
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6074441456602453,
+      "learning_rate": 1.7059274159734165e-06,
+      "loss": 0.6849,
+      "step": 9153
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.0613249609574735,
+      "learning_rate": 1.7058589868240942e-06,
+      "loss": 0.6888,
+      "step": 9154
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.5841330069596347,
+      "learning_rate": 1.7057905510869276e-06,
+      "loss": 0.7057,
+      "step": 9155
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.1745438346160024,
+      "learning_rate": 1.7057221087625556e-06,
+      "loss": 0.6719,
+      "step": 9156
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0862078183324,
+      "learning_rate": 1.7056536598516162e-06,
+      "loss": 0.679,
+      "step": 9157
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0018854958153858,
+      "learning_rate": 1.705585204354749e-06,
+      "loss": 0.681,
+      "step": 9158
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7166568353622885,
+      "learning_rate": 1.705516742272593e-06,
+      "loss": 0.6777,
+      "step": 9159
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.965679843057323,
+      "learning_rate": 1.7054482736057864e-06,
+      "loss": 0.6641,
+      "step": 9160
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.313719889776869,
+      "learning_rate": 1.7053797983549688e-06,
+      "loss": 0.6986,
+      "step": 9161
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.266355297620178,
+      "learning_rate": 1.7053113165207793e-06,
+      "loss": 0.6621,
+      "step": 9162
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.261614024405649,
+      "learning_rate": 1.705242828103857e-06,
+      "loss": 0.6829,
+      "step": 9163
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.451123486434517,
+      "learning_rate": 1.7051743331048406e-06,
+      "loss": 0.6992,
+      "step": 9164
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7129984101409237,
+      "learning_rate": 1.7051058315243702e-06,
+      "loss": 0.6836,
+      "step": 9165
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.3421253669107116,
+      "learning_rate": 1.7050373233630846e-06,
+      "loss": 0.6829,
+      "step": 9166
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7734909927153423,
+      "learning_rate": 1.7049688086216232e-06,
+      "loss": 0.6602,
+      "step": 9167
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.8995125575614367,
+      "learning_rate": 1.7049002873006258e-06,
+      "loss": 0.6602,
+      "step": 9168
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8490212454214706,
+      "learning_rate": 1.7048317594007316e-06,
+      "loss": 0.6849,
+      "step": 9169
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.095084113345477,
+      "learning_rate": 1.7047632249225798e-06,
+      "loss": 0.6738,
+      "step": 9170
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6568380624402392,
+      "learning_rate": 1.704694683866811e-06,
+      "loss": 0.6738,
+      "step": 9171
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.748982304955807,
+      "learning_rate": 1.704626136234064e-06,
+      "loss": 0.6855,
+      "step": 9172
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.0512377617443445,
+      "learning_rate": 1.7045575820249797e-06,
+      "loss": 0.709,
+      "step": 9173
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7080458453296153,
+      "learning_rate": 1.7044890212401966e-06,
+      "loss": 0.6634,
+      "step": 9174
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.7710001355294278,
+      "learning_rate": 1.7044204538803553e-06,
+      "loss": 0.6686,
+      "step": 9175
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.461655425232297,
+      "learning_rate": 1.7043518799460955e-06,
+      "loss": 0.6576,
+      "step": 9176
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.183856585613049,
+      "learning_rate": 1.7042832994380575e-06,
+      "loss": 0.6654,
+      "step": 9177
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5882430614906857,
+      "learning_rate": 1.704214712356881e-06,
+      "loss": 0.6862,
+      "step": 9178
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.791024639901277,
+      "learning_rate": 1.7041461187032065e-06,
+      "loss": 0.6725,
+      "step": 9179
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.7885994989919434,
+      "learning_rate": 1.704077518477674e-06,
+      "loss": 0.6615,
+      "step": 9180
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.1591723940737797,
+      "learning_rate": 1.7040089116809236e-06,
+      "loss": 0.6979,
+      "step": 9181
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 7.945481342484333,
+      "learning_rate": 1.7039402983135957e-06,
+      "loss": 0.6712,
+      "step": 9182
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.6319068602947735,
+      "learning_rate": 1.7038716783763309e-06,
+      "loss": 0.6829,
+      "step": 9183
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.21590405605672,
+      "learning_rate": 1.7038030518697695e-06,
+      "loss": 0.6641,
+      "step": 9184
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.6655017144368864,
+      "learning_rate": 1.7037344187945516e-06,
+      "loss": 0.6849,
+      "step": 9185
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.685988350735724,
+      "learning_rate": 1.7036657791513186e-06,
+      "loss": 0.6706,
+      "step": 9186
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9414390392878395,
+      "learning_rate": 1.7035971329407106e-06,
+      "loss": 0.6706,
+      "step": 9187
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.492355920422382,
+      "learning_rate": 1.703528480163368e-06,
+      "loss": 0.6829,
+      "step": 9188
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.810618030144529,
+      "learning_rate": 1.7034598208199324e-06,
+      "loss": 0.6784,
+      "step": 9189
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.6100306352710447,
+      "learning_rate": 1.7033911549110438e-06,
+      "loss": 0.6725,
+      "step": 9190
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.0278234902747703,
+      "learning_rate": 1.7033224824373431e-06,
+      "loss": 0.6758,
+      "step": 9191
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.555473390876823,
+      "learning_rate": 1.7032538033994718e-06,
+      "loss": 0.6543,
+      "step": 9192
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7213471878150616,
+      "learning_rate": 1.7031851177980706e-06,
+      "loss": 0.6686,
+      "step": 9193
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8962557986173637,
+      "learning_rate": 1.7031164256337801e-06,
+      "loss": 0.6947,
+      "step": 9194
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.762864164715707,
+      "learning_rate": 1.7030477269072423e-06,
+      "loss": 0.6914,
+      "step": 9195
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.6149481963694345,
+      "learning_rate": 1.7029790216190975e-06,
+      "loss": 0.6725,
+      "step": 9196
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.604933797800781,
+      "learning_rate": 1.7029103097699877e-06,
+      "loss": 0.6615,
+      "step": 9197
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3555465400871327,
+      "learning_rate": 1.7028415913605536e-06,
+      "loss": 0.6888,
+      "step": 9198
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9710991646282793,
+      "learning_rate": 1.7027728663914366e-06,
+      "loss": 0.6764,
+      "step": 9199
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7729877208009758,
+      "learning_rate": 1.7027041348632786e-06,
+      "loss": 0.651,
+      "step": 9200
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.9186274370402197,
+      "learning_rate": 1.7026353967767206e-06,
+      "loss": 0.679,
+      "step": 9201
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.753907342009414,
+      "learning_rate": 1.7025666521324048e-06,
+      "loss": 0.6862,
+      "step": 9202
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.3031610119966466,
+      "learning_rate": 1.7024979009309716e-06,
+      "loss": 0.6953,
+      "step": 9203
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.986126659338499,
+      "learning_rate": 1.7024291431730636e-06,
+      "loss": 0.6576,
+      "step": 9204
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5504480955600592,
+      "learning_rate": 1.7023603788593222e-06,
+      "loss": 0.6478,
+      "step": 9205
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9136992614309805,
+      "learning_rate": 1.7022916079903898e-06,
+      "loss": 0.6738,
+      "step": 9206
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.6162410039276436,
+      "learning_rate": 1.702222830566907e-06,
+      "loss": 0.6966,
+      "step": 9207
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.100546818149164,
+      "learning_rate": 1.7021540465895169e-06,
+      "loss": 0.6602,
+      "step": 9208
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8854581471213354,
+      "learning_rate": 1.7020852560588608e-06,
+      "loss": 0.6582,
+      "step": 9209
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 5.07030333328606,
+      "learning_rate": 1.702016458975581e-06,
+      "loss": 0.6921,
+      "step": 9210
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.843672335659326,
+      "learning_rate": 1.7019476553403193e-06,
+      "loss": 0.6816,
+      "step": 9211
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7588860282312812,
+      "learning_rate": 1.7018788451537183e-06,
+      "loss": 0.6797,
+      "step": 9212
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.747944481720809,
+      "learning_rate": 1.70181002841642e-06,
+      "loss": 0.6797,
+      "step": 9213
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.86937832862698,
+      "learning_rate": 1.7017412051290662e-06,
+      "loss": 0.6484,
+      "step": 9214
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.257944195346281,
+      "learning_rate": 1.7016723752922998e-06,
+      "loss": 0.6699,
+      "step": 9215
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.8346938858305233,
+      "learning_rate": 1.7016035389067637e-06,
+      "loss": 0.6777,
+      "step": 9216
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.5600545979239464,
+      "learning_rate": 1.7015346959730989e-06,
+      "loss": 0.6725,
+      "step": 9217
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.215169648204341,
+      "learning_rate": 1.701465846491949e-06,
+      "loss": 0.6862,
+      "step": 9218
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.9276252050051682,
+      "learning_rate": 1.7013969904639564e-06,
+      "loss": 0.651,
+      "step": 9219
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9777845171284809,
+      "learning_rate": 1.7013281278897638e-06,
+      "loss": 0.6784,
+      "step": 9220
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9540514205283603,
+      "learning_rate": 1.7012592587700137e-06,
+      "loss": 0.6673,
+      "step": 9221
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.93581571617637,
+      "learning_rate": 1.7011903831053492e-06,
+      "loss": 0.6842,
+      "step": 9222
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.754659321422443,
+      "learning_rate": 1.7011215008964124e-06,
+      "loss": 0.6634,
+      "step": 9223
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0643051295250954,
+      "learning_rate": 1.7010526121438465e-06,
+      "loss": 0.6908,
+      "step": 9224
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.2003032003878245,
+      "learning_rate": 1.7009837168482947e-06,
+      "loss": 0.653,
+      "step": 9225
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.519125353024893,
+      "learning_rate": 1.7009148150104002e-06,
+      "loss": 0.6784,
+      "step": 9226
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.846464295347159,
+      "learning_rate": 1.7008459066308057e-06,
+      "loss": 0.6842,
+      "step": 9227
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4014656416218518,
+      "learning_rate": 1.700776991710154e-06,
+      "loss": 0.6673,
+      "step": 9228
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.806527408636845,
+      "learning_rate": 1.7007080702490888e-06,
+      "loss": 0.6699,
+      "step": 9229
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7363400644522,
+      "learning_rate": 1.7006391422482533e-06,
+      "loss": 0.694,
+      "step": 9230
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.632746499560052,
+      "learning_rate": 1.7005702077082907e-06,
+      "loss": 0.6771,
+      "step": 9231
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9638449625533294,
+      "learning_rate": 1.7005012666298444e-06,
+      "loss": 0.6361,
+      "step": 9232
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9142968489034125,
+      "learning_rate": 1.700432319013558e-06,
+      "loss": 0.6758,
+      "step": 9233
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.071139256007235,
+      "learning_rate": 1.7003633648600746e-06,
+      "loss": 0.6908,
+      "step": 9234
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.0301639055777452,
+      "learning_rate": 1.7002944041700381e-06,
+      "loss": 0.6426,
+      "step": 9235
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.2929957754705725,
+      "learning_rate": 1.700225436944092e-06,
+      "loss": 0.6641,
+      "step": 9236
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2580719448363882,
+      "learning_rate": 1.7001564631828797e-06,
+      "loss": 0.6458,
+      "step": 9237
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4174130799584796,
+      "learning_rate": 1.7000874828870452e-06,
+      "loss": 0.6895,
+      "step": 9238
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5543424245884347,
+      "learning_rate": 1.700018496057233e-06,
+      "loss": 0.6445,
+      "step": 9239
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.001391578860759,
+      "learning_rate": 1.6999495026940855e-06,
+      "loss": 0.6842,
+      "step": 9240
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9184626453559437,
+      "learning_rate": 1.6998805027982477e-06,
+      "loss": 0.653,
+      "step": 9241
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.3882352023017643,
+      "learning_rate": 1.6998114963703635e-06,
+      "loss": 0.6829,
+      "step": 9242
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.3240905247038905,
+      "learning_rate": 1.6997424834110766e-06,
+      "loss": 0.7077,
+      "step": 9243
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4071501734525,
+      "learning_rate": 1.6996734639210308e-06,
+      "loss": 0.6589,
+      "step": 9244
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.9893327972856283,
+      "learning_rate": 1.6996044379008709e-06,
+      "loss": 0.6309,
+      "step": 9245
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7437779847913333,
+      "learning_rate": 1.6995354053512413e-06,
+      "loss": 0.666,
+      "step": 9246
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.7830669891356767,
+      "learning_rate": 1.6994663662727854e-06,
+      "loss": 0.6693,
+      "step": 9247
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1010201619915208,
+      "learning_rate": 1.699397320666148e-06,
+      "loss": 0.6595,
+      "step": 9248
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.9205206977296005,
+      "learning_rate": 1.699328268531974e-06,
+      "loss": 0.7142,
+      "step": 9249
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4787205168402404,
+      "learning_rate": 1.6992592098709073e-06,
+      "loss": 0.6823,
+      "step": 9250
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4743904948749715,
+      "learning_rate": 1.6991901446835924e-06,
+      "loss": 0.6647,
+      "step": 9251
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 9.083251947336375,
+      "learning_rate": 1.6991210729706743e-06,
+      "loss": 0.7409,
+      "step": 9252
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4641272462664014,
+      "learning_rate": 1.699051994732797e-06,
+      "loss": 0.6549,
+      "step": 9253
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4412885383149336,
+      "learning_rate": 1.6989829099706058e-06,
+      "loss": 0.6504,
+      "step": 9254
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1034240417070411,
+      "learning_rate": 1.6989138186847453e-06,
+      "loss": 0.6882,
+      "step": 9255
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.8608770423778367,
+      "learning_rate": 1.6988447208758606e-06,
+      "loss": 0.6803,
+      "step": 9256
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.4692086561065985,
+      "learning_rate": 1.6987756165445956e-06,
+      "loss": 0.679,
+      "step": 9257
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4619958212595745,
+      "learning_rate": 1.6987065056915966e-06,
+      "loss": 0.6699,
+      "step": 9258
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1558976759213753,
+      "learning_rate": 1.698637388317508e-06,
+      "loss": 0.6999,
+      "step": 9259
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.1765687926866297,
+      "learning_rate": 1.6985682644229746e-06,
+      "loss": 0.6595,
+      "step": 9260
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 7.614954625654576,
+      "learning_rate": 1.6984991340086418e-06,
+      "loss": 0.696,
+      "step": 9261
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.8402896958533694,
+      "learning_rate": 1.6984299970751553e-06,
+      "loss": 0.6615,
+      "step": 9262
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1821108291014037,
+      "learning_rate": 1.6983608536231593e-06,
+      "loss": 0.7012,
+      "step": 9263
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4323500174207981,
+      "learning_rate": 1.6982917036532998e-06,
+      "loss": 0.6471,
+      "step": 9264
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.3236960389341141,
+      "learning_rate": 1.6982225471662222e-06,
+      "loss": 0.6855,
+      "step": 9265
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2490745638099767,
+      "learning_rate": 1.6981533841625717e-06,
+      "loss": 0.7012,
+      "step": 9266
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6293106911619734,
+      "learning_rate": 1.6980842146429942e-06,
+      "loss": 0.6777,
+      "step": 9267
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.1816183059661998,
+      "learning_rate": 1.6980150386081347e-06,
+      "loss": 0.6478,
+      "step": 9268
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2338383032353804,
+      "learning_rate": 1.697945856058639e-06,
+      "loss": 0.6823,
+      "step": 9269
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.360867952892332,
+      "learning_rate": 1.697876666995153e-06,
+      "loss": 0.7064,
+      "step": 9270
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.5198799400694996,
+      "learning_rate": 1.6978074714183226e-06,
+      "loss": 0.6582,
+      "step": 9271
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.2388099687397123,
+      "learning_rate": 1.6977382693287933e-06,
+      "loss": 0.6719,
+      "step": 9272
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.722195216090002,
+      "learning_rate": 1.697669060727211e-06,
+      "loss": 0.6868,
+      "step": 9273
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.3493618016380704,
+      "learning_rate": 1.6975998456142215e-06,
+      "loss": 0.6628,
+      "step": 9274
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.020946446508889,
+      "learning_rate": 1.6975306239904713e-06,
+      "loss": 0.6914,
+      "step": 9275
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 8.194982463635421,
+      "learning_rate": 1.6974613958566058e-06,
+      "loss": 0.6842,
+      "step": 9276
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.6387306422896812,
+      "learning_rate": 1.6973921612132713e-06,
+      "loss": 0.6647,
+      "step": 9277
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.0802581081939264,
+      "learning_rate": 1.697322920061114e-06,
+      "loss": 0.6758,
+      "step": 9278
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.7604496611494076,
+      "learning_rate": 1.6972536724007807e-06,
+      "loss": 0.6738,
+      "step": 9279
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2110662509351289,
+      "learning_rate": 1.697184418232917e-06,
+      "loss": 0.6901,
+      "step": 9280
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 4.167766531472868,
+      "learning_rate": 1.6971151575581694e-06,
+      "loss": 0.6777,
+      "step": 9281
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9565195007208538,
+      "learning_rate": 1.6970458903771844e-06,
+      "loss": 0.6771,
+      "step": 9282
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.4605601189022863,
+      "learning_rate": 1.6969766166906085e-06,
+      "loss": 0.6576,
+      "step": 9283
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.666487977870133,
+      "learning_rate": 1.696907336499088e-06,
+      "loss": 0.6816,
+      "step": 9284
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.320914187840947,
+      "learning_rate": 1.69683804980327e-06,
+      "loss": 0.6615,
+      "step": 9285
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5169158859572844,
+      "learning_rate": 1.6967687566038008e-06,
+      "loss": 0.6582,
+      "step": 9286
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.595790475012993,
+      "learning_rate": 1.696699456901327e-06,
+      "loss": 0.6686,
+      "step": 9287
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5560296537805414,
+      "learning_rate": 1.696630150696496e-06,
+      "loss": 0.6693,
+      "step": 9288
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.742913319059911,
+      "learning_rate": 1.6965608379899536e-06,
+      "loss": 0.6979,
+      "step": 9289
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.380188957222264,
+      "learning_rate": 1.6964915187823477e-06,
+      "loss": 0.6927,
+      "step": 9290
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.0113947751765453,
+      "learning_rate": 1.6964221930743245e-06,
+      "loss": 0.6901,
+      "step": 9291
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.41076979382996,
+      "learning_rate": 1.696352860866532e-06,
+      "loss": 0.6641,
+      "step": 9292
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.809427436592986,
+      "learning_rate": 1.696283522159616e-06,
+      "loss": 0.6719,
+      "step": 9293
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.08766669923799,
+      "learning_rate": 1.6962141769542247e-06,
+      "loss": 0.6536,
+      "step": 9294
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8587126851771527,
+      "learning_rate": 1.6961448252510047e-06,
+      "loss": 0.6712,
+      "step": 9295
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.4867019979892033,
+      "learning_rate": 1.6960754670506037e-06,
+      "loss": 0.6777,
+      "step": 9296
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8130152066729462,
+      "learning_rate": 1.6960061023536687e-06,
+      "loss": 0.7005,
+      "step": 9297
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.548251305563582,
+      "learning_rate": 1.695936731160847e-06,
+      "loss": 0.668,
+      "step": 9298
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.9710310387617938,
+      "learning_rate": 1.6958673534727865e-06,
+      "loss": 0.6882,
+      "step": 9299
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.296568962646546,
+      "learning_rate": 1.6957979692901345e-06,
+      "loss": 0.6712,
+      "step": 9300
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 9.356611060151176,
+      "learning_rate": 1.6957285786135384e-06,
+      "loss": 0.7018,
+      "step": 9301
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5458378032170146,
+      "learning_rate": 1.6956591814436458e-06,
+      "loss": 0.6582,
+      "step": 9302
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.936353252863101,
+      "learning_rate": 1.6955897777811048e-06,
+      "loss": 0.6868,
+      "step": 9303
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.965068365278645,
+      "learning_rate": 1.6955203676265625e-06,
+      "loss": 0.6673,
+      "step": 9304
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.306067799302311,
+      "learning_rate": 1.6954509509806675e-06,
+      "loss": 0.6908,
+      "step": 9305
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.3387753185445157,
+      "learning_rate": 1.6953815278440674e-06,
+      "loss": 0.6745,
+      "step": 9306
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7358375387814916,
+      "learning_rate": 1.6953120982174096e-06,
+      "loss": 0.6901,
+      "step": 9307
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.8954066360837714,
+      "learning_rate": 1.6952426621013429e-06,
+      "loss": 0.6797,
+      "step": 9308
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.069287756809119,
+      "learning_rate": 1.6951732194965147e-06,
+      "loss": 0.6732,
+      "step": 9309
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.751977673360162,
+      "learning_rate": 1.6951037704035733e-06,
+      "loss": 0.6693,
+      "step": 9310
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7975903164056686,
+      "learning_rate": 1.695034314823167e-06,
+      "loss": 0.6732,
+      "step": 9311
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.6775739291675027,
+      "learning_rate": 1.694964852755944e-06,
+      "loss": 0.6784,
+      "step": 9312
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.990000246189686,
+      "learning_rate": 1.6948953842025525e-06,
+      "loss": 0.6556,
+      "step": 9313
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9453625766833835,
+      "learning_rate": 1.694825909163641e-06,
+      "loss": 0.6908,
+      "step": 9314
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.06918278363075,
+      "learning_rate": 1.6947564276398579e-06,
+      "loss": 0.6901,
+      "step": 9315
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.2908503805802036,
+      "learning_rate": 1.6946869396318512e-06,
+      "loss": 0.6842,
+      "step": 9316
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.893141167889451,
+      "learning_rate": 1.6946174451402705e-06,
+      "loss": 0.6816,
+      "step": 9317
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.0611304873604026,
+      "learning_rate": 1.6945479441657634e-06,
+      "loss": 0.6986,
+      "step": 9318
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7537534323367286,
+      "learning_rate": 1.6944784367089788e-06,
+      "loss": 0.6732,
+      "step": 9319
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.2480598692004086,
+      "learning_rate": 1.6944089227705658e-06,
+      "loss": 0.6764,
+      "step": 9320
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8459384925908271,
+      "learning_rate": 1.694339402351173e-06,
+      "loss": 0.666,
+      "step": 9321
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.087434198676277,
+      "learning_rate": 1.6942698754514488e-06,
+      "loss": 0.6784,
+      "step": 9322
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.454286272509224,
+      "learning_rate": 1.6942003420720425e-06,
+      "loss": 0.679,
+      "step": 9323
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.000519503421371,
+      "learning_rate": 1.6941308022136032e-06,
+      "loss": 0.681,
+      "step": 9324
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.3366397708062216,
+      "learning_rate": 1.6940612558767796e-06,
+      "loss": 0.6725,
+      "step": 9325
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.1641869959608444,
+      "learning_rate": 1.6939917030622212e-06,
+      "loss": 0.6849,
+      "step": 9326
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.332850682359284,
+      "learning_rate": 1.6939221437705765e-06,
+      "loss": 0.6888,
+      "step": 9327
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.16652152219531,
+      "learning_rate": 1.6938525780024953e-06,
+      "loss": 0.6725,
+      "step": 9328
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.421387783317438,
+      "learning_rate": 1.6937830057586267e-06,
+      "loss": 0.7051,
+      "step": 9329
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.520427668749435,
+      "learning_rate": 1.6937134270396198e-06,
+      "loss": 0.6699,
+      "step": 9330
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6553375080138726,
+      "learning_rate": 1.6936438418461242e-06,
+      "loss": 0.6784,
+      "step": 9331
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.427728941110382,
+      "learning_rate": 1.6935742501787891e-06,
+      "loss": 0.6647,
+      "step": 9332
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7017076628636436,
+      "learning_rate": 1.6935046520382642e-06,
+      "loss": 0.6888,
+      "step": 9333
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.7553809924917252,
+      "learning_rate": 1.6934350474251992e-06,
+      "loss": 0.696,
+      "step": 9334
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.507619793209572,
+      "learning_rate": 1.6933654363402433e-06,
+      "loss": 0.6719,
+      "step": 9335
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.7693307359645383,
+      "learning_rate": 1.693295818784047e-06,
+      "loss": 0.7188,
+      "step": 9336
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.901234000321596,
+      "learning_rate": 1.6932261947572592e-06,
+      "loss": 0.7038,
+      "step": 9337
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.083228949780308,
+      "learning_rate": 1.6931565642605301e-06,
+      "loss": 0.668,
+      "step": 9338
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.957144282707367,
+      "learning_rate": 1.6930869272945094e-06,
+      "loss": 0.6979,
+      "step": 9339
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8313734226464049,
+      "learning_rate": 1.693017283859847e-06,
+      "loss": 0.6738,
+      "step": 9340
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.1737405455026442,
+      "learning_rate": 1.6929476339571934e-06,
+      "loss": 0.6797,
+      "step": 9341
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8515904359096019,
+      "learning_rate": 1.692877977587198e-06,
+      "loss": 0.6784,
+      "step": 9342
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8844046893646197,
+      "learning_rate": 1.692808314750511e-06,
+      "loss": 0.6667,
+      "step": 9343
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5609913310669468,
+      "learning_rate": 1.692738645447783e-06,
+      "loss": 0.6803,
+      "step": 9344
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.8378815906557593,
+      "learning_rate": 1.6926689696796636e-06,
+      "loss": 0.666,
+      "step": 9345
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.3169628596681522,
+      "learning_rate": 1.6925992874468035e-06,
+      "loss": 0.6777,
+      "step": 9346
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.920950314146007,
+      "learning_rate": 1.6925295987498534e-06,
+      "loss": 0.6784,
+      "step": 9347
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5845963492813664,
+      "learning_rate": 1.6924599035894632e-06,
+      "loss": 0.6816,
+      "step": 9348
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.35927708818465,
+      "learning_rate": 1.6923902019662831e-06,
+      "loss": 0.6745,
+      "step": 9349
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 4.446936550602115,
+      "learning_rate": 1.6923204938809644e-06,
+      "loss": 0.6595,
+      "step": 9350
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.567365107014403,
+      "learning_rate": 1.6922507793341572e-06,
+      "loss": 0.6738,
+      "step": 9351
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 5.250154305495749,
+      "learning_rate": 1.6921810583265123e-06,
+      "loss": 0.6745,
+      "step": 9352
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.7169281886020977,
+      "learning_rate": 1.6921113308586804e-06,
+      "loss": 0.668,
+      "step": 9353
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.1883472457990956,
+      "learning_rate": 1.6920415969313123e-06,
+      "loss": 0.6908,
+      "step": 9354
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.9470986701941384,
+      "learning_rate": 1.6919718565450585e-06,
+      "loss": 0.6875,
+      "step": 9355
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.0638073370622294,
+      "learning_rate": 1.6919021097005701e-06,
+      "loss": 0.6758,
+      "step": 9356
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.105468540000647,
+      "learning_rate": 1.6918323563984984e-06,
+      "loss": 0.6582,
+      "step": 9357
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 7.0073942942611955,
+      "learning_rate": 1.691762596639494e-06,
+      "loss": 0.696,
+      "step": 9358
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8150929049624134,
+      "learning_rate": 1.6916928304242084e-06,
+      "loss": 0.6667,
+      "step": 9359
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.180566894279906,
+      "learning_rate": 1.6916230577532922e-06,
+      "loss": 0.6829,
+      "step": 9360
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.5772733079251298,
+      "learning_rate": 1.6915532786273968e-06,
+      "loss": 0.6686,
+      "step": 9361
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.850781702565314,
+      "learning_rate": 1.6914834930471736e-06,
+      "loss": 0.6719,
+      "step": 9362
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7407921056957307,
+      "learning_rate": 1.691413701013274e-06,
+      "loss": 0.6914,
+      "step": 9363
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8528674590491975,
+      "learning_rate": 1.6913439025263488e-06,
+      "loss": 0.6608,
+      "step": 9364
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.006634154149408,
+      "learning_rate": 1.69127409758705e-06,
+      "loss": 0.6947,
+      "step": 9365
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.8251215074600293,
+      "learning_rate": 1.6912042861960285e-06,
+      "loss": 0.6895,
+      "step": 9366
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.286712713562804,
+      "learning_rate": 1.6911344683539368e-06,
+      "loss": 0.6823,
+      "step": 9367
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.0504496079296235,
+      "learning_rate": 1.6910646440614258e-06,
+      "loss": 0.6855,
+      "step": 9368
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9176122146603708,
+      "learning_rate": 1.6909948133191473e-06,
+      "loss": 0.6738,
+      "step": 9369
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6792945963371197,
+      "learning_rate": 1.6909249761277533e-06,
+      "loss": 0.679,
+      "step": 9370
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.9747039970563067,
+      "learning_rate": 1.6908551324878953e-06,
+      "loss": 0.6823,
+      "step": 9371
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.5939745533216465,
+      "learning_rate": 1.6907852824002253e-06,
+      "loss": 0.6549,
+      "step": 9372
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.637600148775928,
+      "learning_rate": 1.6907154258653948e-06,
+      "loss": 0.6751,
+      "step": 9373
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.3133373305441025,
+      "learning_rate": 1.6906455628840568e-06,
+      "loss": 0.6419,
+      "step": 9374
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.899026958475936,
+      "learning_rate": 1.6905756934568622e-06,
+      "loss": 0.6784,
+      "step": 9375
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6043946509922165,
+      "learning_rate": 1.6905058175844637e-06,
+      "loss": 0.6536,
+      "step": 9376
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.696957891930431,
+      "learning_rate": 1.6904359352675135e-06,
+      "loss": 0.6608,
+      "step": 9377
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8894273941730979,
+      "learning_rate": 1.6903660465066634e-06,
+      "loss": 0.6693,
+      "step": 9378
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7946605556350421,
+      "learning_rate": 1.690296151302566e-06,
+      "loss": 0.6953,
+      "step": 9379
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.2766195226355226,
+      "learning_rate": 1.6902262496558737e-06,
+      "loss": 0.6751,
+      "step": 9380
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7969676734748228,
+      "learning_rate": 1.6901563415672388e-06,
+      "loss": 0.668,
+      "step": 9381
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.646379725268616,
+      "learning_rate": 1.6900864270373134e-06,
+      "loss": 0.6673,
+      "step": 9382
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.783697781402308,
+      "learning_rate": 1.6900165060667507e-06,
+      "loss": 0.7077,
+      "step": 9383
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9004929651885427,
+      "learning_rate": 1.6899465786562029e-06,
+      "loss": 0.6478,
+      "step": 9384
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0191617058793816,
+      "learning_rate": 1.6898766448063223e-06,
+      "loss": 0.6693,
+      "step": 9385
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.837662663821036,
+      "learning_rate": 1.689806704517762e-06,
+      "loss": 0.6426,
+      "step": 9386
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.1366253537216506,
+      "learning_rate": 1.6897367577911755e-06,
+      "loss": 0.6634,
+      "step": 9387
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.113216514182604,
+      "learning_rate": 1.689666804627214e-06,
+      "loss": 0.6777,
+      "step": 9388
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.778054926002196,
+      "learning_rate": 1.6895968450265316e-06,
+      "loss": 0.6992,
+      "step": 9389
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.5357119429453845,
+      "learning_rate": 1.6895268789897808e-06,
+      "loss": 0.6445,
+      "step": 9390
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.115141194756565,
+      "learning_rate": 1.6894569065176147e-06,
+      "loss": 0.6732,
+      "step": 9391
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.279064424789582,
+      "learning_rate": 1.6893869276106862e-06,
+      "loss": 0.7161,
+      "step": 9392
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4014780891383043,
+      "learning_rate": 1.6893169422696486e-06,
+      "loss": 0.6582,
+      "step": 9393
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.3764270229088553,
+      "learning_rate": 1.6892469504951553e-06,
+      "loss": 0.7285,
+      "step": 9394
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.7998921900386944,
+      "learning_rate": 1.6891769522878587e-06,
+      "loss": 0.6615,
+      "step": 9395
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.9316786878579295,
+      "learning_rate": 1.6891069476484128e-06,
+      "loss": 0.6797,
+      "step": 9396
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.3995602102541627,
+      "learning_rate": 1.689036936577471e-06,
+      "loss": 0.6686,
+      "step": 9397
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.064932272305162,
+      "learning_rate": 1.6889669190756866e-06,
+      "loss": 0.64,
+      "step": 9398
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7061225565275777,
+      "learning_rate": 1.6888968951437128e-06,
+      "loss": 0.6654,
+      "step": 9399
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0928339093536188,
+      "learning_rate": 1.6888268647822032e-06,
+      "loss": 0.6901,
+      "step": 9400
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.492561742717643,
+      "learning_rate": 1.688756827991812e-06,
+      "loss": 0.6647,
+      "step": 9401
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.264074640237215,
+      "learning_rate": 1.6886867847731922e-06,
+      "loss": 0.7031,
+      "step": 9402
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8550715634422468,
+      "learning_rate": 1.6886167351269978e-06,
+      "loss": 0.6654,
+      "step": 9403
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.70893022801471,
+      "learning_rate": 1.6885466790538826e-06,
+      "loss": 0.6908,
+      "step": 9404
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0155061281974755,
+      "learning_rate": 1.6884766165545e-06,
+      "loss": 0.6608,
+      "step": 9405
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.474309964405541,
+      "learning_rate": 1.6884065476295044e-06,
+      "loss": 0.6725,
+      "step": 9406
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.161578842596586,
+      "learning_rate": 1.6883364722795498e-06,
+      "loss": 0.6582,
+      "step": 9407
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.5119190749676243,
+      "learning_rate": 1.68826639050529e-06,
+      "loss": 0.6686,
+      "step": 9408
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7934815191151756,
+      "learning_rate": 1.688196302307379e-06,
+      "loss": 0.6706,
+      "step": 9409
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.740643978788661,
+      "learning_rate": 1.6881262076864711e-06,
+      "loss": 0.6719,
+      "step": 9410
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.831169632217328,
+      "learning_rate": 1.6880561066432203e-06,
+      "loss": 0.6543,
+      "step": 9411
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.7716168953286284,
+      "learning_rate": 1.6879859991782813e-06,
+      "loss": 0.6686,
+      "step": 9412
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.1891441527450586,
+      "learning_rate": 1.6879158852923076e-06,
+      "loss": 0.6725,
+      "step": 9413
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 5.07082621642335,
+      "learning_rate": 1.6878457649859548e-06,
+      "loss": 0.7122,
+      "step": 9414
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0180750962768559,
+      "learning_rate": 1.6877756382598762e-06,
+      "loss": 0.6224,
+      "step": 9415
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 6.656445890763785,
+      "learning_rate": 1.6877055051147267e-06,
+      "loss": 0.6719,
+      "step": 9416
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.683727727664329,
+      "learning_rate": 1.687635365551161e-06,
+      "loss": 0.6582,
+      "step": 9417
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8546732376565696,
+      "learning_rate": 1.687565219569834e-06,
+      "loss": 0.668,
+      "step": 9418
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9346118892582027,
+      "learning_rate": 1.6874950671713994e-06,
+      "loss": 0.6693,
+      "step": 9419
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.836182612071113,
+      "learning_rate": 1.687424908356513e-06,
+      "loss": 0.6855,
+      "step": 9420
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0953967571725485,
+      "learning_rate": 1.6873547431258288e-06,
+      "loss": 0.6576,
+      "step": 9421
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.6879593264151707,
+      "learning_rate": 1.687284571480002e-06,
+      "loss": 0.6836,
+      "step": 9422
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.185626434297404,
+      "learning_rate": 1.6872143934196878e-06,
+      "loss": 0.6706,
+      "step": 9423
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0679871414618531,
+      "learning_rate": 1.6871442089455407e-06,
+      "loss": 0.6549,
+      "step": 9424
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.517089891405334,
+      "learning_rate": 1.687074018058216e-06,
+      "loss": 0.6803,
+      "step": 9425
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0183511683858577,
+      "learning_rate": 1.687003820758369e-06,
+      "loss": 0.6523,
+      "step": 9426
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9465096852469961,
+      "learning_rate": 1.6869336170466541e-06,
+      "loss": 0.6712,
+      "step": 9427
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.918348642576582,
+      "learning_rate": 1.6868634069237273e-06,
+      "loss": 0.7044,
+      "step": 9428
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.1941449585144936,
+      "learning_rate": 1.6867931903902433e-06,
+      "loss": 0.6953,
+      "step": 9429
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1340850892730312,
+      "learning_rate": 1.6867229674468582e-06,
+      "loss": 0.6927,
+      "step": 9430
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.4546227107958103,
+      "learning_rate": 1.6866527380942267e-06,
+      "loss": 0.6543,
+      "step": 9431
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.4511453730651218,
+      "learning_rate": 1.686582502333004e-06,
+      "loss": 0.6719,
+      "step": 9432
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.161099628772299,
+      "learning_rate": 1.6865122601638463e-06,
+      "loss": 0.6953,
+      "step": 9433
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.5199968843923886,
+      "learning_rate": 1.6864420115874094e-06,
+      "loss": 0.6478,
+      "step": 9434
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.269144450952579,
+      "learning_rate": 1.6863717566043484e-06,
+      "loss": 0.6719,
+      "step": 9435
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0550287767882534,
+      "learning_rate": 1.6863014952153188e-06,
+      "loss": 0.6999,
+      "step": 9436
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.784207435200631,
+      "learning_rate": 1.686231227420977e-06,
+      "loss": 0.668,
+      "step": 9437
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.820037458513027,
+      "learning_rate": 1.686160953221978e-06,
+      "loss": 0.6992,
+      "step": 9438
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9485436552014328,
+      "learning_rate": 1.6860906726189783e-06,
+      "loss": 0.6784,
+      "step": 9439
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.3003058839431505,
+      "learning_rate": 1.686020385612634e-06,
+      "loss": 0.7109,
+      "step": 9440
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9216149090716865,
+      "learning_rate": 1.6859500922036003e-06,
+      "loss": 0.6673,
+      "step": 9441
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0455685168900786,
+      "learning_rate": 1.6858797923925341e-06,
+      "loss": 0.6803,
+      "step": 9442
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.1439178418596687,
+      "learning_rate": 1.6858094861800912e-06,
+      "loss": 0.6595,
+      "step": 9443
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.24636718381097,
+      "learning_rate": 1.6857391735669273e-06,
+      "loss": 0.6855,
+      "step": 9444
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0471175078297352,
+      "learning_rate": 1.6856688545536992e-06,
+      "loss": 0.6803,
+      "step": 9445
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 5.081322515106662,
+      "learning_rate": 1.685598529141063e-06,
+      "loss": 0.6712,
+      "step": 9446
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.372979690987149,
+      "learning_rate": 1.6855281973296753e-06,
+      "loss": 0.6699,
+      "step": 9447
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1197689475692203,
+      "learning_rate": 1.6854578591201922e-06,
+      "loss": 0.6693,
+      "step": 9448
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0738605304826307,
+      "learning_rate": 1.6853875145132706e-06,
+      "loss": 0.6895,
+      "step": 9449
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.6163463130067712,
+      "learning_rate": 1.6853171635095662e-06,
+      "loss": 0.681,
+      "step": 9450
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.5948615146172496,
+      "learning_rate": 1.6852468061097364e-06,
+      "loss": 0.6647,
+      "step": 9451
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.2132380597939316,
+      "learning_rate": 1.6851764423144375e-06,
+      "loss": 0.6862,
+      "step": 9452
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.874119342516642,
+      "learning_rate": 1.6851060721243262e-06,
+      "loss": 0.6569,
+      "step": 9453
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.234991554377252,
+      "learning_rate": 1.6850356955400593e-06,
+      "loss": 0.668,
+      "step": 9454
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1654921024287894,
+      "learning_rate": 1.6849653125622938e-06,
+      "loss": 0.679,
+      "step": 9455
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.5876076225891507,
+      "learning_rate": 1.6848949231916864e-06,
+      "loss": 0.6908,
+      "step": 9456
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7171485961424795,
+      "learning_rate": 1.6848245274288942e-06,
+      "loss": 0.6777,
+      "step": 9457
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7386109512663415,
+      "learning_rate": 1.684754125274574e-06,
+      "loss": 0.6615,
+      "step": 9458
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.8154856911262427,
+      "learning_rate": 1.684683716729383e-06,
+      "loss": 0.6641,
+      "step": 9459
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8159785496003683,
+      "learning_rate": 1.6846133017939788e-06,
+      "loss": 0.6686,
+      "step": 9460
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.3839417750423542,
+      "learning_rate": 1.6845428804690176e-06,
+      "loss": 0.6621,
+      "step": 9461
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0558139623113854,
+      "learning_rate": 1.6844724527551571e-06,
+      "loss": 0.6862,
+      "step": 9462
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.842692750213089,
+      "learning_rate": 1.684402018653055e-06,
+      "loss": 0.6699,
+      "step": 9463
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.573705511646552,
+      "learning_rate": 1.684331578163368e-06,
+      "loss": 0.6901,
+      "step": 9464
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7585211988995161,
+      "learning_rate": 1.6842611312867542e-06,
+      "loss": 0.6602,
+      "step": 9465
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.7219635226735255,
+      "learning_rate": 1.6841906780238704e-06,
+      "loss": 0.666,
+      "step": 9466
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.5152354726942137,
+      "learning_rate": 1.6841202183753747e-06,
+      "loss": 0.6686,
+      "step": 9467
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.7825971387068322,
+      "learning_rate": 1.6840497523419243e-06,
+      "loss": 0.6771,
+      "step": 9468
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 4.026658063191153,
+      "learning_rate": 1.6839792799241771e-06,
+      "loss": 0.6797,
+      "step": 9469
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8743055383823879,
+      "learning_rate": 1.6839088011227913e-06,
+      "loss": 0.6497,
+      "step": 9470
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.285448521932554,
+      "learning_rate": 1.6838383159384236e-06,
+      "loss": 0.6634,
+      "step": 9471
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.916469271916746,
+      "learning_rate": 1.6837678243717325e-06,
+      "loss": 0.6875,
+      "step": 9472
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1258012591044695,
+      "learning_rate": 1.6836973264233762e-06,
+      "loss": 0.6888,
+      "step": 9473
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.55754232376877,
+      "learning_rate": 1.6836268220940117e-06,
+      "loss": 0.6374,
+      "step": 9474
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0140192122446967,
+      "learning_rate": 1.683556311384298e-06,
+      "loss": 0.6699,
+      "step": 9475
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.1256606316162507,
+      "learning_rate": 1.6834857942948928e-06,
+      "loss": 0.668,
+      "step": 9476
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9802641930659526,
+      "learning_rate": 1.6834152708264542e-06,
+      "loss": 0.7025,
+      "step": 9477
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.4445690619439913,
+      "learning_rate": 1.6833447409796403e-06,
+      "loss": 0.7174,
+      "step": 9478
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 6.025900856740322,
+      "learning_rate": 1.6832742047551096e-06,
+      "loss": 0.6712,
+      "step": 9479
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 3.563850114418661,
+      "learning_rate": 1.6832036621535207e-06,
+      "loss": 0.6673,
+      "step": 9480
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.5304133806400975,
+      "learning_rate": 1.683133113175531e-06,
+      "loss": 0.6966,
+      "step": 9481
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.9375636910255628,
+      "learning_rate": 1.6830625578217999e-06,
+      "loss": 0.6895,
+      "step": 9482
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.7744897329136857,
+      "learning_rate": 1.6829919960929852e-06,
+      "loss": 0.6725,
+      "step": 9483
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.710363822398303,
+      "learning_rate": 1.682921427989746e-06,
+      "loss": 0.638,
+      "step": 9484
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8190450684233177,
+      "learning_rate": 1.6828508535127407e-06,
+      "loss": 0.6901,
+      "step": 9485
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.2355217774363692,
+      "learning_rate": 1.6827802726626284e-06,
+      "loss": 0.6934,
+      "step": 9486
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2804217090171095,
+      "learning_rate": 1.682709685440067e-06,
+      "loss": 0.6842,
+      "step": 9487
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.1774272376141146,
+      "learning_rate": 1.6826390918457156e-06,
+      "loss": 0.6725,
+      "step": 9488
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7968430968198743,
+      "learning_rate": 1.6825684918802334e-06,
+      "loss": 0.6517,
+      "step": 9489
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.727738582364127,
+      "learning_rate": 1.6824978855442793e-06,
+      "loss": 0.6738,
+      "step": 9490
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.510988702282659,
+      "learning_rate": 1.6824272728385118e-06,
+      "loss": 0.7064,
+      "step": 9491
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7338102177417876,
+      "learning_rate": 1.6823566537635902e-06,
+      "loss": 0.6803,
+      "step": 9492
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7890686248308613,
+      "learning_rate": 1.6822860283201737e-06,
+      "loss": 0.679,
+      "step": 9493
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.5444183716030784,
+      "learning_rate": 1.6822153965089214e-06,
+      "loss": 0.668,
+      "step": 9494
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8905975786057961,
+      "learning_rate": 1.6821447583304922e-06,
+      "loss": 0.679,
+      "step": 9495
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.005566232584522,
+      "learning_rate": 1.6820741137855462e-06,
+      "loss": 0.6875,
+      "step": 9496
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.151636785115559,
+      "learning_rate": 1.6820034628747417e-06,
+      "loss": 0.6608,
+      "step": 9497
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.908389906764061,
+      "learning_rate": 1.6819328055987386e-06,
+      "loss": 0.6868,
+      "step": 9498
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.0816483072461582,
+      "learning_rate": 1.6818621419581964e-06,
+      "loss": 0.6628,
+      "step": 9499
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7860002157910276,
+      "learning_rate": 1.6817914719537748e-06,
+      "loss": 0.6908,
+      "step": 9500
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.1403342576183664,
+      "learning_rate": 1.6817207955861327e-06,
+      "loss": 0.6693,
+      "step": 9501
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4538150051969574,
+      "learning_rate": 1.6816501128559302e-06,
+      "loss": 0.6823,
+      "step": 9502
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7007248967961794,
+      "learning_rate": 1.6815794237638271e-06,
+      "loss": 0.6816,
+      "step": 9503
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.368701576403718,
+      "learning_rate": 1.681508728310483e-06,
+      "loss": 0.6621,
+      "step": 9504
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.31489975735538,
+      "learning_rate": 1.6814380264965574e-06,
+      "loss": 0.6986,
+      "step": 9505
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8937389275453913,
+      "learning_rate": 1.6813673183227106e-06,
+      "loss": 0.6751,
+      "step": 9506
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7520955674839375,
+      "learning_rate": 1.6812966037896024e-06,
+      "loss": 0.6589,
+      "step": 9507
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6896347064535673,
+      "learning_rate": 1.6812258828978924e-06,
+      "loss": 0.6816,
+      "step": 9508
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.676676351526415,
+      "learning_rate": 1.6811551556482413e-06,
+      "loss": 0.6803,
+      "step": 9509
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.148687662317602,
+      "learning_rate": 1.6810844220413086e-06,
+      "loss": 0.6523,
+      "step": 9510
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9295917333969211,
+      "learning_rate": 1.6810136820777552e-06,
+      "loss": 0.7038,
+      "step": 9511
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 6.070397914437808,
+      "learning_rate": 1.6809429357582405e-06,
+      "loss": 0.6979,
+      "step": 9512
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.1205715248378458,
+      "learning_rate": 1.6808721830834256e-06,
+      "loss": 0.6602,
+      "step": 9513
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7143439998145027,
+      "learning_rate": 1.6808014240539701e-06,
+      "loss": 0.6725,
+      "step": 9514
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.3728416065521893,
+      "learning_rate": 1.6807306586705347e-06,
+      "loss": 0.6576,
+      "step": 9515
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.826664803912482,
+      "learning_rate": 1.68065988693378e-06,
+      "loss": 0.666,
+      "step": 9516
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.2549173782513803,
+      "learning_rate": 1.6805891088443662e-06,
+      "loss": 0.6823,
+      "step": 9517
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8270671202152395,
+      "learning_rate": 1.6805183244029541e-06,
+      "loss": 0.679,
+      "step": 9518
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.071884995927598,
+      "learning_rate": 1.6804475336102044e-06,
+      "loss": 0.6419,
+      "step": 9519
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.1921822155296558,
+      "learning_rate": 1.6803767364667778e-06,
+      "loss": 0.6595,
+      "step": 9520
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4268249283393655,
+      "learning_rate": 1.6803059329733346e-06,
+      "loss": 0.6497,
+      "step": 9521
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.729064493976697,
+      "learning_rate": 1.680235123130536e-06,
+      "loss": 0.7038,
+      "step": 9522
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.5042403478764372,
+      "learning_rate": 1.6801643069390435e-06,
+      "loss": 0.7044,
+      "step": 9523
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.590225790011658,
+      "learning_rate": 1.6800934843995167e-06,
+      "loss": 0.6719,
+      "step": 9524
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.2176712358619817,
+      "learning_rate": 1.6800226555126173e-06,
+      "loss": 0.6986,
+      "step": 9525
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.6768008836468007,
+      "learning_rate": 1.6799518202790064e-06,
+      "loss": 0.6602,
+      "step": 9526
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.2054494416072576,
+      "learning_rate": 1.6798809786993454e-06,
+      "loss": 0.6387,
+      "step": 9527
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.39989150701562,
+      "learning_rate": 1.6798101307742947e-06,
+      "loss": 0.668,
+      "step": 9528
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8740192996012748,
+      "learning_rate": 1.6797392765045156e-06,
+      "loss": 0.6888,
+      "step": 9529
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6984164991486733,
+      "learning_rate": 1.6796684158906705e-06,
+      "loss": 0.6634,
+      "step": 9530
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 7.0219652218048285,
+      "learning_rate": 1.6795975489334193e-06,
+      "loss": 0.6628,
+      "step": 9531
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.45182528376254,
+      "learning_rate": 1.6795266756334244e-06,
+      "loss": 0.6816,
+      "step": 9532
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.378099410512111,
+      "learning_rate": 1.6794557959913468e-06,
+      "loss": 0.6973,
+      "step": 9533
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.456752570052077,
+      "learning_rate": 1.6793849100078485e-06,
+      "loss": 0.6895,
+      "step": 9534
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.301773763546872,
+      "learning_rate": 1.6793140176835904e-06,
+      "loss": 0.6491,
+      "step": 9535
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8644499893119629,
+      "learning_rate": 1.6792431190192348e-06,
+      "loss": 0.6921,
+      "step": 9536
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9007932079597792,
+      "learning_rate": 1.679172214015443e-06,
+      "loss": 0.6725,
+      "step": 9537
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8064218117990218,
+      "learning_rate": 1.6791013026728766e-06,
+      "loss": 0.6895,
+      "step": 9538
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.025246198336835,
+      "learning_rate": 1.6790303849921981e-06,
+      "loss": 0.666,
+      "step": 9539
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0923385463632698,
+      "learning_rate": 1.6789594609740687e-06,
+      "loss": 0.653,
+      "step": 9540
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.723089034627031,
+      "learning_rate": 1.6788885306191504e-06,
+      "loss": 0.6868,
+      "step": 9541
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.3609184016493776,
+      "learning_rate": 1.6788175939281058e-06,
+      "loss": 0.6712,
+      "step": 9542
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 5.694594441518031,
+      "learning_rate": 1.6787466509015964e-06,
+      "loss": 0.7201,
+      "step": 9543
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.7850621354176779,
+      "learning_rate": 1.6786757015402846e-06,
+      "loss": 0.6693,
+      "step": 9544
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.5107322594001444,
+      "learning_rate": 1.6786047458448324e-06,
+      "loss": 0.6589,
+      "step": 9545
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.623678888888987,
+      "learning_rate": 1.678533783815902e-06,
+      "loss": 0.6842,
+      "step": 9546
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.7427515282302597,
+      "learning_rate": 1.6784628154541558e-06,
+      "loss": 0.6699,
+      "step": 9547
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.3715119829295537,
+      "learning_rate": 1.6783918407602563e-06,
+      "loss": 0.6419,
+      "step": 9548
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0744454267967012,
+      "learning_rate": 1.6783208597348657e-06,
+      "loss": 0.6829,
+      "step": 9549
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.7690346538082838,
+      "learning_rate": 1.6782498723786463e-06,
+      "loss": 0.668,
+      "step": 9550
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6091673257773724,
+      "learning_rate": 1.678178878692261e-06,
+      "loss": 0.6582,
+      "step": 9551
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6273066237536522,
+      "learning_rate": 1.6781078786763722e-06,
+      "loss": 0.6855,
+      "step": 9552
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.8373166564547065,
+      "learning_rate": 1.6780368723316428e-06,
+      "loss": 0.6543,
+      "step": 9553
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0507414529695513,
+      "learning_rate": 1.6779658596587352e-06,
+      "loss": 0.6732,
+      "step": 9554
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8343834746425348,
+      "learning_rate": 1.6778948406583125e-06,
+      "loss": 0.6628,
+      "step": 9555
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8935135885596721,
+      "learning_rate": 1.677823815331037e-06,
+      "loss": 0.6374,
+      "step": 9556
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.48577661596623,
+      "learning_rate": 1.677752783677572e-06,
+      "loss": 0.6602,
+      "step": 9557
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6501129204688043,
+      "learning_rate": 1.6776817456985803e-06,
+      "loss": 0.6777,
+      "step": 9558
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7743504590288344,
+      "learning_rate": 1.677610701394725e-06,
+      "loss": 0.6686,
+      "step": 9559
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.9084133981082347,
+      "learning_rate": 1.6775396507666689e-06,
+      "loss": 0.6914,
+      "step": 9560
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9272438050433522,
+      "learning_rate": 1.6774685938150754e-06,
+      "loss": 0.6667,
+      "step": 9561
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.633723655984862,
+      "learning_rate": 1.677397530540608e-06,
+      "loss": 0.6908,
+      "step": 9562
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.9835246389006103,
+      "learning_rate": 1.6773264609439287e-06,
+      "loss": 0.6589,
+      "step": 9563
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.611949815451516,
+      "learning_rate": 1.6772553850257022e-06,
+      "loss": 0.6393,
+      "step": 9564
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7884290154364095,
+      "learning_rate": 1.6771843027865912e-06,
+      "loss": 0.6999,
+      "step": 9565
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.3029032388658894,
+      "learning_rate": 1.6771132142272592e-06,
+      "loss": 0.6595,
+      "step": 9566
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6797538121690359,
+      "learning_rate": 1.6770421193483698e-06,
+      "loss": 0.6934,
+      "step": 9567
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.482809965198993,
+      "learning_rate": 1.6769710181505863e-06,
+      "loss": 0.6934,
+      "step": 9568
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.318838690733213,
+      "learning_rate": 1.6768999106345726e-06,
+      "loss": 0.6634,
+      "step": 9569
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9876597078985407,
+      "learning_rate": 1.6768287968009923e-06,
+      "loss": 0.6491,
+      "step": 9570
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.705540263104218,
+      "learning_rate": 1.6767576766505086e-06,
+      "loss": 0.6823,
+      "step": 9571
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.998442153282063,
+      "learning_rate": 1.6766865501837857e-06,
+      "loss": 0.6582,
+      "step": 9572
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.413893095578076,
+      "learning_rate": 1.6766154174014876e-06,
+      "loss": 0.6771,
+      "step": 9573
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6060252687346341,
+      "learning_rate": 1.6765442783042778e-06,
+      "loss": 0.6589,
+      "step": 9574
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.5287492618693583,
+      "learning_rate": 1.6764731328928203e-06,
+      "loss": 0.6751,
+      "step": 9575
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3538618683915717,
+      "learning_rate": 1.6764019811677797e-06,
+      "loss": 0.6595,
+      "step": 9576
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0648060792179486,
+      "learning_rate": 1.676330823129819e-06,
+      "loss": 0.6543,
+      "step": 9577
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.688173476836239,
+      "learning_rate": 1.6762596587796031e-06,
+      "loss": 0.6576,
+      "step": 9578
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.362489724366235,
+      "learning_rate": 1.6761884881177963e-06,
+      "loss": 0.6875,
+      "step": 9579
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.2523422281552372,
+      "learning_rate": 1.6761173111450624e-06,
+      "loss": 0.679,
+      "step": 9580
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.86204480888129,
+      "learning_rate": 1.6760461278620657e-06,
+      "loss": 0.6673,
+      "step": 9581
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.1242457915406445,
+      "learning_rate": 1.6759749382694707e-06,
+      "loss": 0.694,
+      "step": 9582
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8028932646019833,
+      "learning_rate": 1.6759037423679422e-06,
+      "loss": 0.6927,
+      "step": 9583
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.285348736083158,
+      "learning_rate": 1.6758325401581436e-06,
+      "loss": 0.666,
+      "step": 9584
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.0072473117865912,
+      "learning_rate": 1.6757613316407405e-06,
+      "loss": 0.6875,
+      "step": 9585
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.864085001826599,
+      "learning_rate": 1.6756901168163975e-06,
+      "loss": 0.6712,
+      "step": 9586
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.34249932812976,
+      "learning_rate": 1.6756188956857783e-06,
+      "loss": 0.6849,
+      "step": 9587
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.9789253460393126,
+      "learning_rate": 1.6755476682495486e-06,
+      "loss": 0.6862,
+      "step": 9588
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8578171814383369,
+      "learning_rate": 1.6754764345083728e-06,
+      "loss": 0.679,
+      "step": 9589
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1859304537147928,
+      "learning_rate": 1.6754051944629154e-06,
+      "loss": 0.6712,
+      "step": 9590
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9135208047183097,
+      "learning_rate": 1.6753339481138418e-06,
+      "loss": 0.6439,
+      "step": 9591
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.291172598780422,
+      "learning_rate": 1.6752626954618165e-06,
+      "loss": 0.6673,
+      "step": 9592
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.280402110755857,
+      "learning_rate": 1.675191436507505e-06,
+      "loss": 0.709,
+      "step": 9593
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7131740886402314,
+      "learning_rate": 1.675120171251572e-06,
+      "loss": 0.6693,
+      "step": 9594
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7351501419976892,
+      "learning_rate": 1.6750488996946825e-06,
+      "loss": 0.6777,
+      "step": 9595
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8318229076332505,
+      "learning_rate": 1.6749776218375024e-06,
+      "loss": 0.6595,
+      "step": 9596
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6961597984382141,
+      "learning_rate": 1.6749063376806963e-06,
+      "loss": 0.6966,
+      "step": 9597
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8282434695401781,
+      "learning_rate": 1.6748350472249296e-06,
+      "loss": 0.6862,
+      "step": 9598
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6735552813162105,
+      "learning_rate": 1.6747637504708676e-06,
+      "loss": 0.681,
+      "step": 9599
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.764530279016239,
+      "learning_rate": 1.6746924474191761e-06,
+      "loss": 0.6641,
+      "step": 9600
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.1189664337287515,
+      "learning_rate": 1.67462113807052e-06,
+      "loss": 0.6738,
+      "step": 9601
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.2194363658483989,
+      "learning_rate": 1.6745498224255656e-06,
+      "loss": 0.6816,
+      "step": 9602
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.2920823120334126,
+      "learning_rate": 1.6744785004849778e-06,
+      "loss": 0.679,
+      "step": 9603
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.635593517775977,
+      "learning_rate": 1.6744071722494222e-06,
+      "loss": 0.6803,
+      "step": 9604
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.4033331400555151,
+      "learning_rate": 1.6743358377195654e-06,
+      "loss": 0.6615,
+      "step": 9605
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8232074387664571,
+      "learning_rate": 1.674264496896072e-06,
+      "loss": 0.6562,
+      "step": 9606
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.24869857921295,
+      "learning_rate": 1.674193149779609e-06,
+      "loss": 0.6901,
+      "step": 9607
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1656225018438873,
+      "learning_rate": 1.6741217963708415e-06,
+      "loss": 0.679,
+      "step": 9608
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.542307421521935,
+      "learning_rate": 1.6740504366704355e-06,
+      "loss": 0.6725,
+      "step": 9609
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.6037936362184697,
+      "learning_rate": 1.6739790706790573e-06,
+      "loss": 0.6797,
+      "step": 9610
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8995054730865857,
+      "learning_rate": 1.6739076983973727e-06,
+      "loss": 0.6667,
+      "step": 9611
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9390299011212142,
+      "learning_rate": 1.6738363198260482e-06,
+      "loss": 0.6426,
+      "step": 9612
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.236726693810282,
+      "learning_rate": 1.6737649349657495e-06,
+      "loss": 0.681,
+      "step": 9613
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.8448031134491082,
+      "learning_rate": 1.6736935438171434e-06,
+      "loss": 0.6855,
+      "step": 9614
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.5177577669101523,
+      "learning_rate": 1.6736221463808955e-06,
+      "loss": 0.6758,
+      "step": 9615
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 5.238479662743005,
+      "learning_rate": 1.6735507426576728e-06,
+      "loss": 0.6875,
+      "step": 9616
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.253235804237378,
+      "learning_rate": 1.6734793326481413e-06,
+      "loss": 0.6602,
+      "step": 9617
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.9911227638338227,
+      "learning_rate": 1.6734079163529676e-06,
+      "loss": 0.6758,
+      "step": 9618
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.8380047772462822,
+      "learning_rate": 1.6733364937728185e-06,
+      "loss": 0.6882,
+      "step": 9619
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 4.6722788314606785,
+      "learning_rate": 1.6732650649083599e-06,
+      "loss": 0.6686,
+      "step": 9620
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.9995105601917205,
+      "learning_rate": 1.6731936297602593e-06,
+      "loss": 0.6868,
+      "step": 9621
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.197650016937239,
+      "learning_rate": 1.673122188329183e-06,
+      "loss": 0.6829,
+      "step": 9622
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.34885559222513,
+      "learning_rate": 1.6730507406157976e-06,
+      "loss": 0.6758,
+      "step": 9623
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.106855153171697,
+      "learning_rate": 1.6729792866207703e-06,
+      "loss": 0.6908,
+      "step": 9624
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.0138300644384106,
+      "learning_rate": 1.6729078263447675e-06,
+      "loss": 0.6621,
+      "step": 9625
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1569891601378757,
+      "learning_rate": 1.672836359788457e-06,
+      "loss": 0.6628,
+      "step": 9626
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3199601596592592,
+      "learning_rate": 1.6727648869525046e-06,
+      "loss": 0.6986,
+      "step": 9627
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.1064007874353297,
+      "learning_rate": 1.6726934078375786e-06,
+      "loss": 0.6497,
+      "step": 9628
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8419547809298458,
+      "learning_rate": 1.6726219224443452e-06,
+      "loss": 0.6497,
+      "step": 9629
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.1751370569248705,
+      "learning_rate": 1.672550430773472e-06,
+      "loss": 0.6823,
+      "step": 9630
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.5194894220311874,
+      "learning_rate": 1.672478932825626e-06,
+      "loss": 0.6361,
+      "step": 9631
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.5417294987000405,
+      "learning_rate": 1.6724074286014748e-06,
+      "loss": 0.6562,
+      "step": 9632
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9743426347578505,
+      "learning_rate": 1.6723359181016856e-06,
+      "loss": 0.6452,
+      "step": 9633
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8832296641210833,
+      "learning_rate": 1.672264401326926e-06,
+      "loss": 0.6947,
+      "step": 9634
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8648227017232651,
+      "learning_rate": 1.672192878277863e-06,
+      "loss": 0.696,
+      "step": 9635
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3387160380603584,
+      "learning_rate": 1.6721213489551645e-06,
+      "loss": 0.6641,
+      "step": 9636
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.1410197158347626,
+      "learning_rate": 1.6720498133594977e-06,
+      "loss": 0.6797,
+      "step": 9637
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8910753650028101,
+      "learning_rate": 1.6719782714915312e-06,
+      "loss": 0.6732,
+      "step": 9638
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3793697895166697,
+      "learning_rate": 1.671906723351932e-06,
+      "loss": 0.6719,
+      "step": 9639
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.6085415332503556,
+      "learning_rate": 1.6718351689413676e-06,
+      "loss": 0.6621,
+      "step": 9640
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5305712384094143,
+      "learning_rate": 1.6717636082605062e-06,
+      "loss": 0.6634,
+      "step": 9641
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5750329521977326,
+      "learning_rate": 1.6716920413100156e-06,
+      "loss": 0.6738,
+      "step": 9642
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5051648999715816,
+      "learning_rate": 1.6716204680905638e-06,
+      "loss": 0.6771,
+      "step": 9643
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0030414537916683,
+      "learning_rate": 1.6715488886028188e-06,
+      "loss": 0.6979,
+      "step": 9644
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6811455042521513,
+      "learning_rate": 1.6714773028474487e-06,
+      "loss": 0.6719,
+      "step": 9645
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.564446335632076,
+      "learning_rate": 1.6714057108251214e-06,
+      "loss": 0.6927,
+      "step": 9646
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.2006433670350414,
+      "learning_rate": 1.6713341125365054e-06,
+      "loss": 0.651,
+      "step": 9647
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.2188325001227733,
+      "learning_rate": 1.6712625079822686e-06,
+      "loss": 0.6387,
+      "step": 9648
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8511725164602723,
+      "learning_rate": 1.6711908971630798e-06,
+      "loss": 0.6758,
+      "step": 9649
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.6221917510378905,
+      "learning_rate": 1.6711192800796065e-06,
+      "loss": 0.7025,
+      "step": 9650
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3503540489236965,
+      "learning_rate": 1.6710476567325177e-06,
+      "loss": 0.681,
+      "step": 9651
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9768459997001198,
+      "learning_rate": 1.6709760271224824e-06,
+      "loss": 0.6842,
+      "step": 9652
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7466352274070805,
+      "learning_rate": 1.670904391250168e-06,
+      "loss": 0.6706,
+      "step": 9653
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.1187038399670226,
+      "learning_rate": 1.6708327491162435e-06,
+      "loss": 0.6628,
+      "step": 9654
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.702639473272661,
+      "learning_rate": 1.6707611007213778e-06,
+      "loss": 0.6576,
+      "step": 9655
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3748746727861492,
+      "learning_rate": 1.6706894460662392e-06,
+      "loss": 0.6842,
+      "step": 9656
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4939498880561393,
+      "learning_rate": 1.6706177851514971e-06,
+      "loss": 0.6784,
+      "step": 9657
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.581790806568183,
+      "learning_rate": 1.6705461179778195e-06,
+      "loss": 0.6751,
+      "step": 9658
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.8399195827266945,
+      "learning_rate": 1.6704744445458759e-06,
+      "loss": 0.6745,
+      "step": 9659
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.1846807233389074,
+      "learning_rate": 1.6704027648563348e-06,
+      "loss": 0.6771,
+      "step": 9660
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8724194163547382,
+      "learning_rate": 1.6703310789098656e-06,
+      "loss": 0.6784,
+      "step": 9661
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 5.703113338853615,
+      "learning_rate": 1.6702593867071372e-06,
+      "loss": 0.6797,
+      "step": 9662
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.014899046744529,
+      "learning_rate": 1.6701876882488186e-06,
+      "loss": 0.6771,
+      "step": 9663
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.5310672770706124,
+      "learning_rate": 1.670115983535579e-06,
+      "loss": 0.6771,
+      "step": 9664
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.430963628822894,
+      "learning_rate": 1.6700442725680876e-06,
+      "loss": 0.6562,
+      "step": 9665
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.7877270897711375,
+      "learning_rate": 1.6699725553470141e-06,
+      "loss": 0.679,
+      "step": 9666
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.321511239910974,
+      "learning_rate": 1.669900831873027e-06,
+      "loss": 0.6699,
+      "step": 9667
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4257416858208407,
+      "learning_rate": 1.6698291021467967e-06,
+      "loss": 0.6908,
+      "step": 9668
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2701348782697925,
+      "learning_rate": 1.6697573661689916e-06,
+      "loss": 0.694,
+      "step": 9669
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.2134053399041311,
+      "learning_rate": 1.669685623940282e-06,
+      "loss": 0.6478,
+      "step": 9670
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0728262720315482,
+      "learning_rate": 1.6696138754613375e-06,
+      "loss": 0.7051,
+      "step": 9671
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.7813672192025423,
+      "learning_rate": 1.6695421207328273e-06,
+      "loss": 0.668,
+      "step": 9672
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3095903628187533,
+      "learning_rate": 1.6694703597554213e-06,
+      "loss": 0.6816,
+      "step": 9673
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.7277598376162095,
+      "learning_rate": 1.669398592529789e-06,
+      "loss": 0.6699,
+      "step": 9674
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3496002993294751,
+      "learning_rate": 1.6693268190566007e-06,
+      "loss": 0.6875,
+      "step": 9675
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 6.072181522590317,
+      "learning_rate": 1.6692550393365259e-06,
+      "loss": 0.6855,
+      "step": 9676
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3630186541198854,
+      "learning_rate": 1.6691832533702349e-06,
+      "loss": 0.6914,
+      "step": 9677
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.3958573958675493,
+      "learning_rate": 1.6691114611583969e-06,
+      "loss": 0.6706,
+      "step": 9678
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3720163486162358,
+      "learning_rate": 1.6690396627016828e-06,
+      "loss": 0.653,
+      "step": 9679
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.999575453309043,
+      "learning_rate": 1.6689678580007622e-06,
+      "loss": 0.6562,
+      "step": 9680
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.975215997286789,
+      "learning_rate": 1.6688960470563056e-06,
+      "loss": 0.696,
+      "step": 9681
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3277966540874255,
+      "learning_rate": 1.6688242298689829e-06,
+      "loss": 0.6699,
+      "step": 9682
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.034929807189619,
+      "learning_rate": 1.6687524064394648e-06,
+      "loss": 0.6667,
+      "step": 9683
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.6120732587569009,
+      "learning_rate": 1.668680576768421e-06,
+      "loss": 0.7096,
+      "step": 9684
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.4762152495721406,
+      "learning_rate": 1.6686087408565224e-06,
+      "loss": 0.6667,
+      "step": 9685
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.147645433677212,
+      "learning_rate": 1.6685368987044392e-06,
+      "loss": 0.6927,
+      "step": 9686
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.822945191312777,
+      "learning_rate": 1.6684650503128421e-06,
+      "loss": 0.6667,
+      "step": 9687
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.637141422157643,
+      "learning_rate": 1.6683931956824015e-06,
+      "loss": 0.6836,
+      "step": 9688
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8327189887284647,
+      "learning_rate": 1.668321334813788e-06,
+      "loss": 0.6517,
+      "step": 9689
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1655777979161333,
+      "learning_rate": 1.6682494677076727e-06,
+      "loss": 0.6628,
+      "step": 9690
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.9745096422524133,
+      "learning_rate": 1.6681775943647258e-06,
+      "loss": 0.6725,
+      "step": 9691
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4839107895559898,
+      "learning_rate": 1.6681057147856183e-06,
+      "loss": 0.6758,
+      "step": 9692
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.255368826700643,
+      "learning_rate": 1.668033828971021e-06,
+      "loss": 0.6602,
+      "step": 9693
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.202548805042927,
+      "learning_rate": 1.667961936921605e-06,
+      "loss": 0.6641,
+      "step": 9694
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3362022676111964,
+      "learning_rate": 1.6678900386380414e-06,
+      "loss": 0.6615,
+      "step": 9695
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.301692288876765,
+      "learning_rate": 1.6678181341210006e-06,
+      "loss": 0.6855,
+      "step": 9696
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.587575565235512,
+      "learning_rate": 1.6677462233711544e-06,
+      "loss": 0.6484,
+      "step": 9697
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.8330753834865057,
+      "learning_rate": 1.6676743063891736e-06,
+      "loss": 0.7142,
+      "step": 9698
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8806365219363756,
+      "learning_rate": 1.6676023831757293e-06,
+      "loss": 0.6816,
+      "step": 9699
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8732887130169137,
+      "learning_rate": 1.667530453731493e-06,
+      "loss": 0.668,
+      "step": 9700
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.068346304168517,
+      "learning_rate": 1.6674585180571362e-06,
+      "loss": 0.6803,
+      "step": 9701
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.126979944432927,
+      "learning_rate": 1.6673865761533296e-06,
+      "loss": 0.6693,
+      "step": 9702
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.0921392012856914,
+      "learning_rate": 1.6673146280207453e-06,
+      "loss": 0.6543,
+      "step": 9703
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9359091954711463,
+      "learning_rate": 1.6672426736600542e-06,
+      "loss": 0.6699,
+      "step": 9704
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9518910079676866,
+      "learning_rate": 1.6671707130719288e-06,
+      "loss": 0.651,
+      "step": 9705
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.2425547176816032,
+      "learning_rate": 1.6670987462570398e-06,
+      "loss": 0.6868,
+      "step": 9706
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.247302545809432,
+      "learning_rate": 1.6670267732160592e-06,
+      "loss": 0.6797,
+      "step": 9707
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.844535076921108,
+      "learning_rate": 1.666954793949659e-06,
+      "loss": 0.7103,
+      "step": 9708
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.670003665248686,
+      "learning_rate": 1.6668828084585106e-06,
+      "loss": 0.6621,
+      "step": 9709
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1066171458201315,
+      "learning_rate": 1.6668108167432857e-06,
+      "loss": 0.6309,
+      "step": 9710
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 6.103652096362747,
+      "learning_rate": 1.6667388188046566e-06,
+      "loss": 0.6849,
+      "step": 9711
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.4395702790341298,
+      "learning_rate": 1.6666668146432952e-06,
+      "loss": 0.6992,
+      "step": 9712
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.628926661943596,
+      "learning_rate": 1.6665948042598734e-06,
+      "loss": 0.7174,
+      "step": 9713
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.7294803332650512,
+      "learning_rate": 1.6665227876550633e-06,
+      "loss": 0.679,
+      "step": 9714
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4831239927296727,
+      "learning_rate": 1.6664507648295374e-06,
+      "loss": 0.6589,
+      "step": 9715
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.811136377416195,
+      "learning_rate": 1.6663787357839674e-06,
+      "loss": 0.6882,
+      "step": 9716
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.070162829310352,
+      "learning_rate": 1.6663067005190254e-06,
+      "loss": 0.6589,
+      "step": 9717
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.127900796891792,
+      "learning_rate": 1.6662346590353847e-06,
+      "loss": 0.6836,
+      "step": 9718
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.028414382375511,
+      "learning_rate": 1.6661626113337164e-06,
+      "loss": 0.6699,
+      "step": 9719
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.1186275730037307,
+      "learning_rate": 1.666090557414694e-06,
+      "loss": 0.6582,
+      "step": 9720
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8100493946962505,
+      "learning_rate": 1.6660184972789892e-06,
+      "loss": 0.6953,
+      "step": 9721
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.508338191705204,
+      "learning_rate": 1.665946430927275e-06,
+      "loss": 0.6517,
+      "step": 9722
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.923315843937856,
+      "learning_rate": 1.6658743583602238e-06,
+      "loss": 0.6751,
+      "step": 9723
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7655147113632967,
+      "learning_rate": 1.6658022795785086e-06,
+      "loss": 0.666,
+      "step": 9724
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.9870452894289063,
+      "learning_rate": 1.6657301945828013e-06,
+      "loss": 0.6686,
+      "step": 9725
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.872222606378239,
+      "learning_rate": 1.6656581033737755e-06,
+      "loss": 0.6471,
+      "step": 9726
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9418020036085006,
+      "learning_rate": 1.665586005952104e-06,
+      "loss": 0.7044,
+      "step": 9727
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2798697756572857,
+      "learning_rate": 1.6655139023184592e-06,
+      "loss": 0.668,
+      "step": 9728
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.2683535029402613,
+      "learning_rate": 1.6654417924735142e-06,
+      "loss": 0.7129,
+      "step": 9729
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8225226793740442,
+      "learning_rate": 1.665369676417942e-06,
+      "loss": 0.6875,
+      "step": 9730
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.4816016402818848,
+      "learning_rate": 1.6652975541524162e-06,
+      "loss": 0.6784,
+      "step": 9731
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2042561017312865,
+      "learning_rate": 1.6652254256776092e-06,
+      "loss": 0.6901,
+      "step": 9732
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6045966222998318,
+      "learning_rate": 1.6651532909941945e-06,
+      "loss": 0.651,
+      "step": 9733
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8934264942445005,
+      "learning_rate": 1.6650811501028454e-06,
+      "loss": 0.6784,
+      "step": 9734
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.862258918482296,
+      "learning_rate": 1.6650090030042353e-06,
+      "loss": 0.6797,
+      "step": 9735
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.632831610960232,
+      "learning_rate": 1.664936849699037e-06,
+      "loss": 0.6914,
+      "step": 9736
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 5.022620421420958,
+      "learning_rate": 1.6648646901879242e-06,
+      "loss": 0.681,
+      "step": 9737
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9508298690090705,
+      "learning_rate": 1.6647925244715709e-06,
+      "loss": 0.6777,
+      "step": 9738
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.630931686515698,
+      "learning_rate": 1.6647203525506498e-06,
+      "loss": 0.7005,
+      "step": 9739
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 7.338958859621994,
+      "learning_rate": 1.6646481744258352e-06,
+      "loss": 0.7109,
+      "step": 9740
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.445622919826445,
+      "learning_rate": 1.6645759900978002e-06,
+      "loss": 0.6979,
+      "step": 9741
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4545403913671824,
+      "learning_rate": 1.6645037995672186e-06,
+      "loss": 0.6673,
+      "step": 9742
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4950360031213865,
+      "learning_rate": 1.6644316028347645e-06,
+      "loss": 0.6823,
+      "step": 9743
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.060917012928369,
+      "learning_rate": 1.6643593999011114e-06,
+      "loss": 0.6849,
+      "step": 9744
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3862232824384333,
+      "learning_rate": 1.664287190766933e-06,
+      "loss": 0.694,
+      "step": 9745
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8552087013477551,
+      "learning_rate": 1.664214975432904e-06,
+      "loss": 0.666,
+      "step": 9746
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.49292427998881,
+      "learning_rate": 1.6641427538996977e-06,
+      "loss": 0.6582,
+      "step": 9747
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.4886485421590443,
+      "learning_rate": 1.6640705261679883e-06,
+      "loss": 0.6764,
+      "step": 9748
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.3668286106525467,
+      "learning_rate": 1.6639982922384497e-06,
+      "loss": 0.6829,
+      "step": 9749
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.4584286440700827,
+      "learning_rate": 1.6639260521117566e-06,
+      "loss": 0.6751,
+      "step": 9750
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 4.521497243643481,
+      "learning_rate": 1.6638538057885831e-06,
+      "loss": 0.6706,
+      "step": 9751
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.092617250602266,
+      "learning_rate": 1.6637815532696031e-06,
+      "loss": 0.6868,
+      "step": 9752
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.589137793935203,
+      "learning_rate": 1.663709294555491e-06,
+      "loss": 0.6816,
+      "step": 9753
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.6540304088528517,
+      "learning_rate": 1.6636370296469218e-06,
+      "loss": 0.6738,
+      "step": 9754
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 3.433939872536329,
+      "learning_rate": 1.6635647585445694e-06,
+      "loss": 0.6712,
+      "step": 9755
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4721720743051914,
+      "learning_rate": 1.6634924812491083e-06,
+      "loss": 0.6673,
+      "step": 9756
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9610845680779063,
+      "learning_rate": 1.663420197761213e-06,
+      "loss": 0.6328,
+      "step": 9757
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.3426400746442064,
+      "learning_rate": 1.6633479080815586e-06,
+      "loss": 0.6569,
+      "step": 9758
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.8442052693277415,
+      "learning_rate": 1.6632756122108194e-06,
+      "loss": 0.6465,
+      "step": 9759
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0111606275071159,
+      "learning_rate": 1.66320331014967e-06,
+      "loss": 0.6517,
+      "step": 9760
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.900961578489697,
+      "learning_rate": 1.663131001898786e-06,
+      "loss": 0.6908,
+      "step": 9761
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.1527698410462595,
+      "learning_rate": 1.6630586874588413e-06,
+      "loss": 0.6419,
+      "step": 9762
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.595408651288162,
+      "learning_rate": 1.6629863668305112e-06,
+      "loss": 0.6725,
+      "step": 9763
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4917457039857178,
+      "learning_rate": 1.662914040014471e-06,
+      "loss": 0.6309,
+      "step": 9764
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.7956494108541556,
+      "learning_rate": 1.6628417070113953e-06,
+      "loss": 0.6621,
+      "step": 9765
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4219412268472278,
+      "learning_rate": 1.6627693678219594e-06,
+      "loss": 0.6836,
+      "step": 9766
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6256713855011813,
+      "learning_rate": 1.6626970224468387e-06,
+      "loss": 0.7064,
+      "step": 9767
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4115057913548923,
+      "learning_rate": 1.6626246708867076e-06,
+      "loss": 0.6673,
+      "step": 9768
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.4928122185175834,
+      "learning_rate": 1.6625523131422417e-06,
+      "loss": 0.694,
+      "step": 9769
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4832188413049459,
+      "learning_rate": 1.6624799492141168e-06,
+      "loss": 0.6927,
+      "step": 9770
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0898047812555751,
+      "learning_rate": 1.6624075791030081e-06,
+      "loss": 0.6738,
+      "step": 9771
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0935544220894187,
+      "learning_rate": 1.6623352028095904e-06,
+      "loss": 0.6484,
+      "step": 9772
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.2414110521814226,
+      "learning_rate": 1.6622628203345399e-06,
+      "loss": 0.6986,
+      "step": 9773
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9436869089966068,
+      "learning_rate": 1.6621904316785323e-06,
+      "loss": 0.6699,
+      "step": 9774
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.9764007524960485,
+      "learning_rate": 1.6621180368422423e-06,
+      "loss": 0.6986,
+      "step": 9775
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0421123288871688,
+      "learning_rate": 1.6620456358263463e-06,
+      "loss": 0.6361,
+      "step": 9776
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.708008561247752,
+      "learning_rate": 1.66197322863152e-06,
+      "loss": 0.7038,
+      "step": 9777
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.1160123159416082,
+      "learning_rate": 1.6619008152584387e-06,
+      "loss": 0.6673,
+      "step": 9778
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.122894280507208,
+      "learning_rate": 1.6618283957077787e-06,
+      "loss": 0.6543,
+      "step": 9779
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.2639825148663444,
+      "learning_rate": 1.661755969980216e-06,
+      "loss": 0.7142,
+      "step": 9780
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8968389193266446,
+      "learning_rate": 1.661683538076426e-06,
+      "loss": 0.6647,
+      "step": 9781
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4334992922354743,
+      "learning_rate": 1.6616110999970849e-06,
+      "loss": 0.6797,
+      "step": 9782
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8293062730660452,
+      "learning_rate": 1.6615386557428693e-06,
+      "loss": 0.6947,
+      "step": 9783
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.4134118537157425,
+      "learning_rate": 1.6614662053144546e-06,
+      "loss": 0.6797,
+      "step": 9784
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.9819117817149463,
+      "learning_rate": 1.6613937487125176e-06,
+      "loss": 0.6693,
+      "step": 9785
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.687479821979684,
+      "learning_rate": 1.661321285937734e-06,
+      "loss": 0.6777,
+      "step": 9786
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.549526723829375,
+      "learning_rate": 1.6612488169907803e-06,
+      "loss": 0.6751,
+      "step": 9787
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.059026373136371,
+      "learning_rate": 1.6611763418723333e-06,
+      "loss": 0.6875,
+      "step": 9788
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.0635332246506946,
+      "learning_rate": 1.6611038605830686e-06,
+      "loss": 0.6686,
+      "step": 9789
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0411742030185271,
+      "learning_rate": 1.6610313731236635e-06,
+      "loss": 0.6556,
+      "step": 9790
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.108268372762885,
+      "learning_rate": 1.6609588794947937e-06,
+      "loss": 0.7155,
+      "step": 9791
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.300516292962294,
+      "learning_rate": 1.6608863796971364e-06,
+      "loss": 0.681,
+      "step": 9792
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7469426722475678,
+      "learning_rate": 1.6608138737313681e-06,
+      "loss": 0.681,
+      "step": 9793
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7909849205170547,
+      "learning_rate": 1.6607413615981658e-06,
+      "loss": 0.6816,
+      "step": 9794
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9229890456113778,
+      "learning_rate": 1.6606688432982054e-06,
+      "loss": 0.6836,
+      "step": 9795
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.8294300378244888,
+      "learning_rate": 1.6605963188321646e-06,
+      "loss": 0.6849,
+      "step": 9796
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.2886518756758847,
+      "learning_rate": 1.6605237882007196e-06,
+      "loss": 0.651,
+      "step": 9797
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.312948770034611,
+      "learning_rate": 1.6604512514045481e-06,
+      "loss": 0.6582,
+      "step": 9798
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8412145554445429,
+      "learning_rate": 1.6603787084443265e-06,
+      "loss": 0.6628,
+      "step": 9799
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 4.76515660078146,
+      "learning_rate": 1.660306159320732e-06,
+      "loss": 0.6855,
+      "step": 9800
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.721959848765309,
+      "learning_rate": 1.6602336040344418e-06,
+      "loss": 0.6836,
+      "step": 9801
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.818785489983603,
+      "learning_rate": 1.6601610425861327e-06,
+      "loss": 0.6784,
+      "step": 9802
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.1550146305719124,
+      "learning_rate": 1.6600884749764824e-06,
+      "loss": 0.681,
+      "step": 9803
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.5428615451109757,
+      "learning_rate": 1.660015901206168e-06,
+      "loss": 0.6615,
+      "step": 9804
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.935713691475513,
+      "learning_rate": 1.6599433212758669e-06,
+      "loss": 0.679,
+      "step": 9805
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.2926517887057205,
+      "learning_rate": 1.6598707351862562e-06,
+      "loss": 0.6803,
+      "step": 9806
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.056385892331831,
+      "learning_rate": 1.6597981429380136e-06,
+      "loss": 0.6947,
+      "step": 9807
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9991339031936356,
+      "learning_rate": 1.6597255445318167e-06,
+      "loss": 0.6868,
+      "step": 9808
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.407317626178942,
+      "learning_rate": 1.659652939968343e-06,
+      "loss": 0.6556,
+      "step": 9809
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.272493147860428,
+      "learning_rate": 1.6595803292482699e-06,
+      "loss": 0.6712,
+      "step": 9810
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7958060109711418,
+      "learning_rate": 1.6595077123722754e-06,
+      "loss": 0.6758,
+      "step": 9811
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7251922036259588,
+      "learning_rate": 1.6594350893410371e-06,
+      "loss": 0.6751,
+      "step": 9812
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.009177822464526,
+      "learning_rate": 1.6593624601552327e-06,
+      "loss": 0.7077,
+      "step": 9813
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 6.232954564613191,
+      "learning_rate": 1.6592898248155406e-06,
+      "loss": 0.6901,
+      "step": 9814
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 5.773028733985202,
+      "learning_rate": 1.6592171833226375e-06,
+      "loss": 0.6693,
+      "step": 9815
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.6372812385090216,
+      "learning_rate": 1.6591445356772024e-06,
+      "loss": 0.6803,
+      "step": 9816
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.4128250382131875,
+      "learning_rate": 1.6590718818799134e-06,
+      "loss": 0.6745,
+      "step": 9817
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.6094847791902662,
+      "learning_rate": 1.658999221931448e-06,
+      "loss": 0.6654,
+      "step": 9818
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.1863051559330382,
+      "learning_rate": 1.6589265558324845e-06,
+      "loss": 0.625,
+      "step": 9819
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6887388272900787,
+      "learning_rate": 1.6588538835837014e-06,
+      "loss": 0.679,
+      "step": 9820
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0280499180670137,
+      "learning_rate": 1.6587812051857766e-06,
+      "loss": 0.6745,
+      "step": 9821
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6340914339715475,
+      "learning_rate": 1.6587085206393882e-06,
+      "loss": 0.6803,
+      "step": 9822
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8572458477567645,
+      "learning_rate": 1.6586358299452154e-06,
+      "loss": 0.6673,
+      "step": 9823
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.9431014724381395,
+      "learning_rate": 1.6585631331039363e-06,
+      "loss": 0.6888,
+      "step": 9824
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.097119985369713,
+      "learning_rate": 1.658490430116229e-06,
+      "loss": 0.668,
+      "step": 9825
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6467018113645682,
+      "learning_rate": 1.658417720982772e-06,
+      "loss": 0.7031,
+      "step": 9826
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.318628537004638,
+      "learning_rate": 1.6583450057042443e-06,
+      "loss": 0.7064,
+      "step": 9827
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.0230295806026461,
+      "learning_rate": 1.6582722842813247e-06,
+      "loss": 0.6927,
+      "step": 9828
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.5652656597226042,
+      "learning_rate": 1.6581995567146914e-06,
+      "loss": 0.6895,
+      "step": 9829
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6784488571165563,
+      "learning_rate": 1.6581268230050238e-06,
+      "loss": 0.6641,
+      "step": 9830
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9129400983884061,
+      "learning_rate": 1.6580540831529998e-06,
+      "loss": 0.6836,
+      "step": 9831
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.510801044985163,
+      "learning_rate": 1.6579813371592994e-06,
+      "loss": 0.6764,
+      "step": 9832
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9120823818752061,
+      "learning_rate": 1.6579085850246005e-06,
+      "loss": 0.679,
+      "step": 9833
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7305943356863195,
+      "learning_rate": 1.657835826749583e-06,
+      "loss": 0.6745,
+      "step": 9834
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2243408226157664,
+      "learning_rate": 1.6577630623349252e-06,
+      "loss": 0.6875,
+      "step": 9835
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.7810101553386706,
+      "learning_rate": 1.6576902917813069e-06,
+      "loss": 0.6797,
+      "step": 9836
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.6374878710578666,
+      "learning_rate": 1.657617515089407e-06,
+      "loss": 0.6947,
+      "step": 9837
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.9329461796345964,
+      "learning_rate": 1.6575447322599043e-06,
+      "loss": 0.6719,
+      "step": 9838
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.5044560547857644,
+      "learning_rate": 1.6574719432934787e-06,
+      "loss": 0.6914,
+      "step": 9839
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.412584157791351,
+      "learning_rate": 1.657399148190809e-06,
+      "loss": 0.6699,
+      "step": 9840
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.3819741201165197,
+      "learning_rate": 1.6573263469525754e-06,
+      "loss": 0.6895,
+      "step": 9841
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.3638637328602843,
+      "learning_rate": 1.6572535395794566e-06,
+      "loss": 0.6758,
+      "step": 9842
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9693963761677733,
+      "learning_rate": 1.6571807260721324e-06,
+      "loss": 0.6543,
+      "step": 9843
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9476116931549834,
+      "learning_rate": 1.6571079064312826e-06,
+      "loss": 0.653,
+      "step": 9844
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1106753372242069,
+      "learning_rate": 1.6570350806575864e-06,
+      "loss": 0.6738,
+      "step": 9845
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6881844021402357,
+      "learning_rate": 1.6569622487517236e-06,
+      "loss": 0.6445,
+      "step": 9846
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.677795538091941,
+      "learning_rate": 1.6568894107143741e-06,
+      "loss": 0.6895,
+      "step": 9847
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.994761970390672,
+      "learning_rate": 1.6568165665462176e-06,
+      "loss": 0.666,
+      "step": 9848
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.213324164379337,
+      "learning_rate": 1.656743716247934e-06,
+      "loss": 0.6706,
+      "step": 9849
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.151848591766622,
+      "learning_rate": 1.6566708598202034e-06,
+      "loss": 0.679,
+      "step": 9850
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.431650300702672,
+      "learning_rate": 1.6565979972637055e-06,
+      "loss": 0.6641,
+      "step": 9851
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 7.876689842037546,
+      "learning_rate": 1.6565251285791205e-06,
+      "loss": 0.7018,
+      "step": 9852
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.772023018433608,
+      "learning_rate": 1.6564522537671282e-06,
+      "loss": 0.6719,
+      "step": 9853
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6769198096291962,
+      "learning_rate": 1.656379372828409e-06,
+      "loss": 0.679,
+      "step": 9854
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.1544063784835514,
+      "learning_rate": 1.656306485763643e-06,
+      "loss": 0.6719,
+      "step": 9855
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.0626458482572727,
+      "learning_rate": 1.6562335925735108e-06,
+      "loss": 0.679,
+      "step": 9856
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 6.4171779476564605,
+      "learning_rate": 1.6561606932586925e-06,
+      "loss": 0.6569,
+      "step": 9857
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8426440919988678,
+      "learning_rate": 1.656087787819868e-06,
+      "loss": 0.6549,
+      "step": 9858
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.729220339787862,
+      "learning_rate": 1.6560148762577181e-06,
+      "loss": 0.6914,
+      "step": 9859
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.586584854497965,
+      "learning_rate": 1.655941958572924e-06,
+      "loss": 0.6706,
+      "step": 9860
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.543836915159266,
+      "learning_rate": 1.655869034766165e-06,
+      "loss": 0.6647,
+      "step": 9861
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.733243275553741,
+      "learning_rate": 1.6557961048381225e-06,
+      "loss": 0.6621,
+      "step": 9862
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.8011314364180078,
+      "learning_rate": 1.6557231687894771e-06,
+      "loss": 0.6641,
+      "step": 9863
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.1800549939419775,
+      "learning_rate": 1.6556502266209092e-06,
+      "loss": 0.6693,
+      "step": 9864
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.328023102258037,
+      "learning_rate": 1.6555772783330998e-06,
+      "loss": 0.7311,
+      "step": 9865
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2598830977171613,
+      "learning_rate": 1.6555043239267298e-06,
+      "loss": 0.6849,
+      "step": 9866
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.359746152122676,
+      "learning_rate": 1.6554313634024798e-06,
+      "loss": 0.694,
+      "step": 9867
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.317029029167903,
+      "learning_rate": 1.6553583967610309e-06,
+      "loss": 0.6478,
+      "step": 9868
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.040390999198379,
+      "learning_rate": 1.6552854240030643e-06,
+      "loss": 0.6549,
+      "step": 9869
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9681573107406458,
+      "learning_rate": 1.6552124451292608e-06,
+      "loss": 0.6966,
+      "step": 9870
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8413450400178544,
+      "learning_rate": 1.6551394601403015e-06,
+      "loss": 0.6836,
+      "step": 9871
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9686397598214506,
+      "learning_rate": 1.6550664690368678e-06,
+      "loss": 0.6973,
+      "step": 9872
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7960481332728135,
+      "learning_rate": 1.6549934718196407e-06,
+      "loss": 0.64,
+      "step": 9873
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.260838867142914,
+      "learning_rate": 1.6549204684893019e-06,
+      "loss": 0.6634,
+      "step": 9874
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.319906004149914,
+      "learning_rate": 1.6548474590465323e-06,
+      "loss": 0.7077,
+      "step": 9875
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.958525477853591,
+      "learning_rate": 1.6547744434920134e-06,
+      "loss": 0.681,
+      "step": 9876
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.3444243138149365,
+      "learning_rate": 1.6547014218264265e-06,
+      "loss": 0.6497,
+      "step": 9877
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 5.468791324941726,
+      "learning_rate": 1.6546283940504538e-06,
+      "loss": 0.6862,
+      "step": 9878
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.573313468985222,
+      "learning_rate": 1.654555360164776e-06,
+      "loss": 0.6628,
+      "step": 9879
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 8.922031088090549,
+      "learning_rate": 1.6544823201700753e-06,
+      "loss": 0.6641,
+      "step": 9880
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.08899204304058,
+      "learning_rate": 1.6544092740670333e-06,
+      "loss": 0.6719,
+      "step": 9881
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.5198970624583983,
+      "learning_rate": 1.6543362218563314e-06,
+      "loss": 0.6875,
+      "step": 9882
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7933324011825298,
+      "learning_rate": 1.654263163538652e-06,
+      "loss": 0.6784,
+      "step": 9883
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.004600049751718,
+      "learning_rate": 1.6541900991146764e-06,
+      "loss": 0.6576,
+      "step": 9884
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.8735131876181963,
+      "learning_rate": 1.6541170285850869e-06,
+      "loss": 0.6693,
+      "step": 9885
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.351812946094553,
+      "learning_rate": 1.6540439519505654e-06,
+      "loss": 0.6875,
+      "step": 9886
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.944223494603865,
+      "learning_rate": 1.6539708692117937e-06,
+      "loss": 0.6803,
+      "step": 9887
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 4.486377271610382,
+      "learning_rate": 1.6538977803694541e-06,
+      "loss": 0.7025,
+      "step": 9888
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.156924591873219,
+      "learning_rate": 1.6538246854242284e-06,
+      "loss": 0.6549,
+      "step": 9889
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0708877411193474,
+      "learning_rate": 1.6537515843767995e-06,
+      "loss": 0.6823,
+      "step": 9890
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.113195834709318,
+      "learning_rate": 1.653678477227849e-06,
+      "loss": 0.679,
+      "step": 9891
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.842462057708399,
+      "learning_rate": 1.6536053639780596e-06,
+      "loss": 0.7181,
+      "step": 9892
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.31666780644333,
+      "learning_rate": 1.6535322446281135e-06,
+      "loss": 0.6986,
+      "step": 9893
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.3497186731817528,
+      "learning_rate": 1.6534591191786932e-06,
+      "loss": 0.6973,
+      "step": 9894
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7918966070161075,
+      "learning_rate": 1.653385987630481e-06,
+      "loss": 0.6934,
+      "step": 9895
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.4286206345549182,
+      "learning_rate": 1.65331284998416e-06,
+      "loss": 0.6992,
+      "step": 9896
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.2135903429058477,
+      "learning_rate": 1.653239706240412e-06,
+      "loss": 0.681,
+      "step": 9897
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.316903854663833,
+      "learning_rate": 1.65316655639992e-06,
+      "loss": 0.681,
+      "step": 9898
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.4931124603792463,
+      "learning_rate": 1.653093400463367e-06,
+      "loss": 0.679,
+      "step": 9899
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7284430656698797,
+      "learning_rate": 1.6530202384314356e-06,
+      "loss": 0.6706,
+      "step": 9900
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6657576336003742,
+      "learning_rate": 1.6529470703048084e-06,
+      "loss": 0.6836,
+      "step": 9901
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.603917312095757,
+      "learning_rate": 1.6528738960841682e-06,
+      "loss": 0.6816,
+      "step": 9902
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0329223371212657,
+      "learning_rate": 1.6528007157701986e-06,
+      "loss": 0.6771,
+      "step": 9903
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8323641916797815,
+      "learning_rate": 1.652727529363582e-06,
+      "loss": 0.6745,
+      "step": 9904
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.206245772824931,
+      "learning_rate": 1.6526543368650017e-06,
+      "loss": 0.6758,
+      "step": 9905
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5350552572166755,
+      "learning_rate": 1.652581138275141e-06,
+      "loss": 0.6895,
+      "step": 9906
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.2167629922241874,
+      "learning_rate": 1.6525079335946823e-06,
+      "loss": 0.6908,
+      "step": 9907
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.3924977992347682,
+      "learning_rate": 1.6524347228243096e-06,
+      "loss": 0.6751,
+      "step": 9908
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7471420753179014,
+      "learning_rate": 1.652361505964706e-06,
+      "loss": 0.6914,
+      "step": 9909
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 5.555616987141426,
+      "learning_rate": 1.652288283016555e-06,
+      "loss": 0.681,
+      "step": 9910
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.6314292956600354,
+      "learning_rate": 1.6522150539805392e-06,
+      "loss": 0.6758,
+      "step": 9911
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.636869225829931,
+      "learning_rate": 1.6521418188573431e-06,
+      "loss": 0.6895,
+      "step": 9912
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6521618316540639,
+      "learning_rate": 1.6520685776476495e-06,
+      "loss": 0.6725,
+      "step": 9913
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.5605299245445297,
+      "learning_rate": 1.6519953303521423e-06,
+      "loss": 0.6842,
+      "step": 9914
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 6.024815640352903,
+      "learning_rate": 1.651922076971505e-06,
+      "loss": 0.6758,
+      "step": 9915
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6095034089301409,
+      "learning_rate": 1.6518488175064213e-06,
+      "loss": 0.6706,
+      "step": 9916
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.418683053308214,
+      "learning_rate": 1.651775551957575e-06,
+      "loss": 0.681,
+      "step": 9917
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.5795962204571,
+      "learning_rate": 1.6517022803256495e-06,
+      "loss": 0.6771,
+      "step": 9918
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.3064639931515403,
+      "learning_rate": 1.6516290026113295e-06,
+      "loss": 0.681,
+      "step": 9919
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.625374913063576,
+      "learning_rate": 1.6515557188152979e-06,
+      "loss": 0.679,
+      "step": 9920
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6141781287872546,
+      "learning_rate": 1.6514824289382395e-06,
+      "loss": 0.6803,
+      "step": 9921
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.247251926634323,
+      "learning_rate": 1.6514091329808378e-06,
+      "loss": 0.6882,
+      "step": 9922
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8795045115200982,
+      "learning_rate": 1.6513358309437772e-06,
+      "loss": 0.6719,
+      "step": 9923
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.928579701688441,
+      "learning_rate": 1.6512625228277415e-06,
+      "loss": 0.6562,
+      "step": 9924
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.7578256748138177,
+      "learning_rate": 1.6511892086334155e-06,
+      "loss": 0.7031,
+      "step": 9925
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0991464784939484,
+      "learning_rate": 1.6511158883614823e-06,
+      "loss": 0.6836,
+      "step": 9926
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.2664765643153544,
+      "learning_rate": 1.6510425620126275e-06,
+      "loss": 0.6543,
+      "step": 9927
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.4932884269908633,
+      "learning_rate": 1.6509692295875347e-06,
+      "loss": 0.668,
+      "step": 9928
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6098757625246827,
+      "learning_rate": 1.6508958910868886e-06,
+      "loss": 0.6589,
+      "step": 9929
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.103215373239368,
+      "learning_rate": 1.6508225465113737e-06,
+      "loss": 0.694,
+      "step": 9930
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7107622881144262,
+      "learning_rate": 1.650749195861674e-06,
+      "loss": 0.6862,
+      "step": 9931
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.656338189121676,
+      "learning_rate": 1.6506758391384747e-06,
+      "loss": 0.6712,
+      "step": 9932
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.122563161303288,
+      "learning_rate": 1.6506024763424607e-06,
+      "loss": 0.6901,
+      "step": 9933
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.674901837334565,
+      "learning_rate": 1.6505291074743157e-06,
+      "loss": 0.6667,
+      "step": 9934
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.2183901991678245,
+      "learning_rate": 1.650455732534725e-06,
+      "loss": 0.6992,
+      "step": 9935
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.775127007209287,
+      "learning_rate": 1.6503823515243737e-06,
+      "loss": 0.6576,
+      "step": 9936
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.771424177565759,
+      "learning_rate": 1.6503089644439463e-06,
+      "loss": 0.6667,
+      "step": 9937
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.1410985672506837,
+      "learning_rate": 1.6502355712941278e-06,
+      "loss": 0.6947,
+      "step": 9938
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7927106300957308,
+      "learning_rate": 1.6501621720756033e-06,
+      "loss": 0.6517,
+      "step": 9939
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.956358119415231,
+      "learning_rate": 1.6500887667890578e-06,
+      "loss": 0.6693,
+      "step": 9940
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.402063066913063,
+      "learning_rate": 1.650015355435176e-06,
+      "loss": 0.6953,
+      "step": 9941
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.879040991610949,
+      "learning_rate": 1.6499419380146437e-06,
+      "loss": 0.6712,
+      "step": 9942
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.7678752535632025,
+      "learning_rate": 1.649868514528146e-06,
+      "loss": 0.7096,
+      "step": 9943
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.135139001882899,
+      "learning_rate": 1.6497950849763675e-06,
+      "loss": 0.6862,
+      "step": 9944
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 8.03701122104132,
+      "learning_rate": 1.6497216493599942e-06,
+      "loss": 0.6608,
+      "step": 9945
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.325119180204224,
+      "learning_rate": 1.6496482076797114e-06,
+      "loss": 0.6589,
+      "step": 9946
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.617221420712736,
+      "learning_rate": 1.6495747599362044e-06,
+      "loss": 0.6628,
+      "step": 9947
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.9472259810965098,
+      "learning_rate": 1.6495013061301586e-06,
+      "loss": 0.6673,
+      "step": 9948
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.693456960603709,
+      "learning_rate": 1.6494278462622599e-06,
+      "loss": 0.6784,
+      "step": 9949
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.4792967609242966,
+      "learning_rate": 1.6493543803331933e-06,
+      "loss": 0.6797,
+      "step": 9950
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.8862477843142669,
+      "learning_rate": 1.649280908343645e-06,
+      "loss": 0.6667,
+      "step": 9951
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 4.482084092124688,
+      "learning_rate": 1.6492074302943006e-06,
+      "loss": 0.6797,
+      "step": 9952
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.592123881696566,
+      "learning_rate": 1.649133946185846e-06,
+      "loss": 0.6901,
+      "step": 9953
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.3000686582365106,
+      "learning_rate": 1.6490604560189667e-06,
+      "loss": 0.6784,
+      "step": 9954
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.40554690450808,
+      "learning_rate": 1.6489869597943486e-06,
+      "loss": 0.7044,
+      "step": 9955
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9080005400283278,
+      "learning_rate": 1.6489134575126781e-06,
+      "loss": 0.6764,
+      "step": 9956
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.8666511750151473,
+      "learning_rate": 1.648839949174641e-06,
+      "loss": 0.7025,
+      "step": 9957
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 5.226683752515855,
+      "learning_rate": 1.6487664347809229e-06,
+      "loss": 0.6934,
+      "step": 9958
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.3454855822448273,
+      "learning_rate": 1.6486929143322109e-06,
+      "loss": 0.6465,
+      "step": 9959
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.939556644956294,
+      "learning_rate": 1.64861938782919e-06,
+      "loss": 0.6797,
+      "step": 9960
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.726738417011594,
+      "learning_rate": 1.648545855272547e-06,
+      "loss": 0.679,
+      "step": 9961
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.3468676696892572,
+      "learning_rate": 1.6484723166629684e-06,
+      "loss": 0.6562,
+      "step": 9962
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8794287875959326,
+      "learning_rate": 1.64839877200114e-06,
+      "loss": 0.6471,
+      "step": 9963
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7096241873656556,
+      "learning_rate": 1.648325221287749e-06,
+      "loss": 0.6771,
+      "step": 9964
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.527265138430421,
+      "learning_rate": 1.6482516645234811e-06,
+      "loss": 0.6888,
+      "step": 9965
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.8248149089016414,
+      "learning_rate": 1.6481781017090234e-06,
+      "loss": 0.6628,
+      "step": 9966
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.684900575171601,
+      "learning_rate": 1.6481045328450616e-06,
+      "loss": 0.6829,
+      "step": 9967
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.292888961210333,
+      "learning_rate": 1.6480309579322831e-06,
+      "loss": 0.6849,
+      "step": 9968
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.827013086784271,
+      "learning_rate": 1.6479573769713745e-06,
+      "loss": 0.6706,
+      "step": 9969
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7623454604462017,
+      "learning_rate": 1.6478837899630223e-06,
+      "loss": 0.6797,
+      "step": 9970
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.0522098944438925,
+      "learning_rate": 1.6478101969079135e-06,
+      "loss": 0.6777,
+      "step": 9971
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7822033766690107,
+      "learning_rate": 1.6477365978067347e-06,
+      "loss": 0.6764,
+      "step": 9972
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.276645194078977,
+      "learning_rate": 1.647662992660173e-06,
+      "loss": 0.6816,
+      "step": 9973
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.448456818354612,
+      "learning_rate": 1.6475893814689156e-06,
+      "loss": 0.6751,
+      "step": 9974
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.2967554484691157,
+      "learning_rate": 1.6475157642336489e-06,
+      "loss": 0.6673,
+      "step": 9975
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.830765368045768,
+      "learning_rate": 1.6474421409550602e-06,
+      "loss": 0.6849,
+      "step": 9976
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4111959926118993,
+      "learning_rate": 1.6473685116338368e-06,
+      "loss": 0.6777,
+      "step": 9977
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8366615505089715,
+      "learning_rate": 1.647294876270666e-06,
+      "loss": 0.653,
+      "step": 9978
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8834181216650654,
+      "learning_rate": 1.6472212348662348e-06,
+      "loss": 0.6628,
+      "step": 9979
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.3715062743769715,
+      "learning_rate": 1.6471475874212306e-06,
+      "loss": 0.6986,
+      "step": 9980
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.7074706882837152,
+      "learning_rate": 1.6470739339363406e-06,
+      "loss": 0.6471,
+      "step": 9981
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.006183832735476,
+      "learning_rate": 1.6470002744122526e-06,
+      "loss": 0.6589,
+      "step": 9982
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.35964339250405,
+      "learning_rate": 1.6469266088496537e-06,
+      "loss": 0.6621,
+      "step": 9983
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.2144134580224017,
+      "learning_rate": 1.6468529372492315e-06,
+      "loss": 0.6751,
+      "step": 9984
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.5184373433269736,
+      "learning_rate": 1.646779259611674e-06,
+      "loss": 0.6634,
+      "step": 9985
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.4946966305629332,
+      "learning_rate": 1.646705575937668e-06,
+      "loss": 0.6699,
+      "step": 9986
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.921605648806424,
+      "learning_rate": 1.646631886227902e-06,
+      "loss": 0.6927,
+      "step": 9987
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9563212590951395,
+      "learning_rate": 1.6465581904830627e-06,
+      "loss": 0.6947,
+      "step": 9988
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.363289718233698,
+      "learning_rate": 1.6464844887038394e-06,
+      "loss": 0.6816,
+      "step": 9989
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 4.6834021581060155,
+      "learning_rate": 1.646410780890919e-06,
+      "loss": 0.6667,
+      "step": 9990
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.285496453921681,
+      "learning_rate": 1.6463370670449894e-06,
+      "loss": 0.6901,
+      "step": 9991
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.9144105929026172,
+      "learning_rate": 1.646263347166739e-06,
+      "loss": 0.6738,
+      "step": 9992
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.0826417989560624,
+      "learning_rate": 1.6461896212568557e-06,
+      "loss": 0.6497,
+      "step": 9993
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.05497370429454,
+      "learning_rate": 1.6461158893160275e-06,
+      "loss": 0.6667,
+      "step": 9994
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.4865225041151064,
+      "learning_rate": 1.6460421513449422e-06,
+      "loss": 0.6621,
+      "step": 9995
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8441192050696686,
+      "learning_rate": 1.6459684073442887e-06,
+      "loss": 0.6829,
+      "step": 9996
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.6691485289232972,
+      "learning_rate": 1.6458946573147549e-06,
+      "loss": 0.6771,
+      "step": 9997
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.4942345175125595,
+      "learning_rate": 1.645820901257029e-06,
+      "loss": 0.6608,
+      "step": 9998
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.7330203369105919,
+      "learning_rate": 1.6457471391717999e-06,
+      "loss": 0.6888,
+      "step": 9999
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.180051857444326,
+      "learning_rate": 1.6456733710597553e-06,
+      "loss": 0.7155,
+      "step": 10000
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.5754689181450026,
+      "learning_rate": 1.645599596921584e-06,
+      "loss": 0.6895,
+      "step": 10001
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 6.719709913597436,
+      "learning_rate": 1.6455258167579749e-06,
+      "loss": 0.707,
+      "step": 10002
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8227450681596211,
+      "learning_rate": 1.6454520305696161e-06,
+      "loss": 0.6771,
+      "step": 10003
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7899071314459465,
+      "learning_rate": 1.6453782383571961e-06,
+      "loss": 0.6608,
+      "step": 10004
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.5950688709998109,
+      "learning_rate": 1.6453044401214046e-06,
+      "loss": 0.6764,
+      "step": 10005
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.0574134250647447,
+      "learning_rate": 1.645230635862929e-06,
+      "loss": 0.6771,
+      "step": 10006
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4218660439905266,
+      "learning_rate": 1.6451568255824592e-06,
+      "loss": 0.668,
+      "step": 10007
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8232592648361359,
+      "learning_rate": 1.6450830092806832e-06,
+      "loss": 0.6628,
+      "step": 10008
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4736661645844606,
+      "learning_rate": 1.645009186958291e-06,
+      "loss": 0.6706,
+      "step": 10009
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8600865022452353,
+      "learning_rate": 1.6449353586159709e-06,
+      "loss": 0.6803,
+      "step": 10010
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.3022630605018486,
+      "learning_rate": 1.644861524254412e-06,
+      "loss": 0.6803,
+      "step": 10011
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8118546706850466,
+      "learning_rate": 1.6447876838743034e-06,
+      "loss": 0.6732,
+      "step": 10012
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.2426370023676856,
+      "learning_rate": 1.6447138374763343e-06,
+      "loss": 0.6452,
+      "step": 10013
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9489654785476652,
+      "learning_rate": 1.644639985061194e-06,
+      "loss": 0.6725,
+      "step": 10014
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.7672036043656284,
+      "learning_rate": 1.6445661266295719e-06,
+      "loss": 0.6875,
+      "step": 10015
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.9774213104405796,
+      "learning_rate": 1.644492262182157e-06,
+      "loss": 0.6868,
+      "step": 10016
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.4971003239095444,
+      "learning_rate": 1.6444183917196388e-06,
+      "loss": 0.6556,
+      "step": 10017
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.136063171919282,
+      "learning_rate": 1.6443445152427068e-06,
+      "loss": 0.6868,
+      "step": 10018
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.526107132749558,
+      "learning_rate": 1.6442706327520503e-06,
+      "loss": 0.6589,
+      "step": 10019
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9196094808843973,
+      "learning_rate": 1.644196744248359e-06,
+      "loss": 0.6725,
+      "step": 10020
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.9637141800315325,
+      "learning_rate": 1.6441228497323227e-06,
+      "loss": 0.6901,
+      "step": 10021
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9633403067275358,
+      "learning_rate": 1.644048949204631e-06,
+      "loss": 0.6901,
+      "step": 10022
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.509248435909666,
+      "learning_rate": 1.643975042665973e-06,
+      "loss": 0.6882,
+      "step": 10023
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.8453863580551237,
+      "learning_rate": 1.6439011301170393e-06,
+      "loss": 0.6549,
+      "step": 10024
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.33638815327061,
+      "learning_rate": 1.6438272115585195e-06,
+      "loss": 0.6628,
+      "step": 10025
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.436197485807936,
+      "learning_rate": 1.6437532869911033e-06,
+      "loss": 0.6549,
+      "step": 10026
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.006482610256991,
+      "learning_rate": 1.6436793564154808e-06,
+      "loss": 0.6712,
+      "step": 10027
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7923271002944339,
+      "learning_rate": 1.6436054198323417e-06,
+      "loss": 0.7031,
+      "step": 10028
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.2136392351116996,
+      "learning_rate": 1.6435314772423765e-06,
+      "loss": 0.6797,
+      "step": 10029
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.39066684730858,
+      "learning_rate": 1.643457528646275e-06,
+      "loss": 0.6602,
+      "step": 10030
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.185218772224239,
+      "learning_rate": 1.6433835740447277e-06,
+      "loss": 0.7005,
+      "step": 10031
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.30551345051245,
+      "learning_rate": 1.6433096134384245e-06,
+      "loss": 0.6927,
+      "step": 10032
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.309976160620673,
+      "learning_rate": 1.6432356468280557e-06,
+      "loss": 0.6589,
+      "step": 10033
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.4535421285553503,
+      "learning_rate": 1.6431616742143117e-06,
+      "loss": 0.6608,
+      "step": 10034
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.810293185735206,
+      "learning_rate": 1.6430876955978832e-06,
+      "loss": 0.6973,
+      "step": 10035
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.1280773260600077,
+      "learning_rate": 1.6430137109794601e-06,
+      "loss": 0.6628,
+      "step": 10036
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.2919601527852067,
+      "learning_rate": 1.6429397203597333e-06,
+      "loss": 0.6855,
+      "step": 10037
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.1971019552033257,
+      "learning_rate": 1.6428657237393933e-06,
+      "loss": 0.6497,
+      "step": 10038
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.2538367795125285,
+      "learning_rate": 1.6427917211191304e-06,
+      "loss": 0.6654,
+      "step": 10039
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.2170286076344574,
+      "learning_rate": 1.6427177124996353e-06,
+      "loss": 0.6901,
+      "step": 10040
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.3629576565614965,
+      "learning_rate": 1.6426436978815997e-06,
+      "loss": 0.6745,
+      "step": 10041
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.383037132339359,
+      "learning_rate": 1.6425696772657131e-06,
+      "loss": 0.6855,
+      "step": 10042
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8363338943626104,
+      "learning_rate": 1.642495650652667e-06,
+      "loss": 0.6641,
+      "step": 10043
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7715424454085996,
+      "learning_rate": 1.642421618043152e-06,
+      "loss": 0.6823,
+      "step": 10044
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.4720760281670646,
+      "learning_rate": 1.6423475794378593e-06,
+      "loss": 0.6784,
+      "step": 10045
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.321998821471092,
+      "learning_rate": 1.6422735348374802e-06,
+      "loss": 0.6829,
+      "step": 10046
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3264765373966032,
+      "learning_rate": 1.6421994842427052e-06,
+      "loss": 0.7135,
+      "step": 10047
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0447975938330636,
+      "learning_rate": 1.6421254276542255e-06,
+      "loss": 0.6836,
+      "step": 10048
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.40589979951859,
+      "learning_rate": 1.6420513650727324e-06,
+      "loss": 0.6777,
+      "step": 10049
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.716699374532945,
+      "learning_rate": 1.6419772964989172e-06,
+      "loss": 0.6751,
+      "step": 10050
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 5.627287290569854,
+      "learning_rate": 1.6419032219334712e-06,
+      "loss": 0.6947,
+      "step": 10051
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.2388253448792164,
+      "learning_rate": 1.6418291413770856e-06,
+      "loss": 0.7044,
+      "step": 10052
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.479099720915079,
+      "learning_rate": 1.6417550548304515e-06,
+      "loss": 0.6595,
+      "step": 10053
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.982423700019169,
+      "learning_rate": 1.6416809622942612e-06,
+      "loss": 0.6491,
+      "step": 10054
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0974516225801076,
+      "learning_rate": 1.6416068637692058e-06,
+      "loss": 0.6751,
+      "step": 10055
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8904539272982976,
+      "learning_rate": 1.6415327592559765e-06,
+      "loss": 0.6732,
+      "step": 10056
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.3472401765800934,
+      "learning_rate": 1.641458648755265e-06,
+      "loss": 0.6582,
+      "step": 10057
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.5365743482298653,
+      "learning_rate": 1.6413845322677635e-06,
+      "loss": 0.6882,
+      "step": 10058
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7991107803163578,
+      "learning_rate": 1.6413104097941637e-06,
+      "loss": 0.6439,
+      "step": 10059
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.962610781081156,
+      "learning_rate": 1.6412362813351566e-06,
+      "loss": 0.6973,
+      "step": 10060
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5904427816096207,
+      "learning_rate": 1.641162146891435e-06,
+      "loss": 0.679,
+      "step": 10061
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.9199057114835714,
+      "learning_rate": 1.6410880064636904e-06,
+      "loss": 0.6829,
+      "step": 10062
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.4536112249463886,
+      "learning_rate": 1.6410138600526144e-06,
+      "loss": 0.6986,
+      "step": 10063
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.4282300678268347,
+      "learning_rate": 1.6409397076588993e-06,
+      "loss": 0.6719,
+      "step": 10064
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.296620215139268,
+      "learning_rate": 1.6408655492832378e-06,
+      "loss": 0.6914,
+      "step": 10065
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.971540902464812,
+      "learning_rate": 1.640791384926321e-06,
+      "loss": 0.6979,
+      "step": 10066
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 7.032783747167393,
+      "learning_rate": 1.6407172145888414e-06,
+      "loss": 0.6719,
+      "step": 10067
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6468483667010119,
+      "learning_rate": 1.640643038271492e-06,
+      "loss": 0.6803,
+      "step": 10068
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.251977832249665,
+      "learning_rate": 1.640568855974964e-06,
+      "loss": 0.6875,
+      "step": 10069
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.037096745237455,
+      "learning_rate": 1.6404946676999502e-06,
+      "loss": 0.6719,
+      "step": 10070
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7584541786601974,
+      "learning_rate": 1.6404204734471432e-06,
+      "loss": 0.6712,
+      "step": 10071
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.979719975736982,
+      "learning_rate": 1.640346273217235e-06,
+      "loss": 0.6712,
+      "step": 10072
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7421505319894611,
+      "learning_rate": 1.6402720670109189e-06,
+      "loss": 0.6654,
+      "step": 10073
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.7687563677521991,
+      "learning_rate": 1.6401978548288864e-06,
+      "loss": 0.6849,
+      "step": 10074
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.097166661572905,
+      "learning_rate": 1.6401236366718311e-06,
+      "loss": 0.6517,
+      "step": 10075
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.8262110669750455,
+      "learning_rate": 1.6400494125404452e-06,
+      "loss": 0.6901,
+      "step": 10076
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7908367471956809,
+      "learning_rate": 1.6399751824354215e-06,
+      "loss": 0.6732,
+      "step": 10077
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.1130818154575521,
+      "learning_rate": 1.639900946357453e-06,
+      "loss": 0.6745,
+      "step": 10078
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.7736055528082635,
+      "learning_rate": 1.6398267043072322e-06,
+      "loss": 0.681,
+      "step": 10079
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 4.599054147338766,
+      "learning_rate": 1.639752456285452e-06,
+      "loss": 0.694,
+      "step": 10080
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.6607182288898472,
+      "learning_rate": 1.639678202292806e-06,
+      "loss": 0.6641,
+      "step": 10081
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7879646240307386,
+      "learning_rate": 1.6396039423299863e-06,
+      "loss": 0.6836,
+      "step": 10082
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.845942548443652,
+      "learning_rate": 1.6395296763976866e-06,
+      "loss": 0.6608,
+      "step": 10083
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.268874323932314,
+      "learning_rate": 1.6394554044966001e-06,
+      "loss": 0.6758,
+      "step": 10084
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.535517933697156,
+      "learning_rate": 1.6393811266274196e-06,
+      "loss": 0.6758,
+      "step": 10085
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.1063614438503,
+      "learning_rate": 1.6393068427908386e-06,
+      "loss": 0.6706,
+      "step": 10086
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7837066318283684,
+      "learning_rate": 1.6392325529875502e-06,
+      "loss": 0.6699,
+      "step": 10087
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.1548663166252537,
+      "learning_rate": 1.6391582572182476e-06,
+      "loss": 0.6849,
+      "step": 10088
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.671247082524675,
+      "learning_rate": 1.639083955483625e-06,
+      "loss": 0.6921,
+      "step": 10089
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.7654103599532784,
+      "learning_rate": 1.6390096477843746e-06,
+      "loss": 0.6914,
+      "step": 10090
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.0769563657372365,
+      "learning_rate": 1.6389353341211913e-06,
+      "loss": 0.6882,
+      "step": 10091
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.347365172382965,
+      "learning_rate": 1.638861014494768e-06,
+      "loss": 0.6771,
+      "step": 10092
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1596798275971207,
+      "learning_rate": 1.638786688905798e-06,
+      "loss": 0.6732,
+      "step": 10093
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.6014933709828543,
+      "learning_rate": 1.6387123573549755e-06,
+      "loss": 0.6973,
+      "step": 10094
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9847970477914642,
+      "learning_rate": 1.6386380198429941e-06,
+      "loss": 0.653,
+      "step": 10095
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8043871075806164,
+      "learning_rate": 1.6385636763705475e-06,
+      "loss": 0.6589,
+      "step": 10096
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8109676358289992,
+      "learning_rate": 1.6384893269383294e-06,
+      "loss": 0.7148,
+      "step": 10097
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.0215467735122177,
+      "learning_rate": 1.6384149715470347e-06,
+      "loss": 0.6921,
+      "step": 10098
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.040336323102228,
+      "learning_rate": 1.6383406101973561e-06,
+      "loss": 0.6784,
+      "step": 10099
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.014433289397713,
+      "learning_rate": 1.638266242889988e-06,
+      "loss": 0.7018,
+      "step": 10100
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.781177746666437,
+      "learning_rate": 1.638191869625625e-06,
+      "loss": 0.6732,
+      "step": 10101
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8733977716848446,
+      "learning_rate": 1.6381174904049607e-06,
+      "loss": 0.6647,
+      "step": 10102
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.5264441624701548,
+      "learning_rate": 1.6380431052286894e-06,
+      "loss": 0.696,
+      "step": 10103
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.3508130755595977,
+      "learning_rate": 1.6379687140975054e-06,
+      "loss": 0.653,
+      "step": 10104
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.639809336726616,
+      "learning_rate": 1.6378943170121027e-06,
+      "loss": 0.6888,
+      "step": 10105
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.8663493031781435,
+      "learning_rate": 1.637819913973176e-06,
+      "loss": 0.6986,
+      "step": 10106
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6596002172206183,
+      "learning_rate": 1.63774550498142e-06,
+      "loss": 0.6836,
+      "step": 10107
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.581845764461697,
+      "learning_rate": 1.6376710900375286e-06,
+      "loss": 0.6647,
+      "step": 10108
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.176172332087184,
+      "learning_rate": 1.6375966691421966e-06,
+      "loss": 0.6829,
+      "step": 10109
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6789340835719438,
+      "learning_rate": 1.637522242296118e-06,
+      "loss": 0.6849,
+      "step": 10110
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.5223715617572435,
+      "learning_rate": 1.6374478094999886e-06,
+      "loss": 0.6706,
+      "step": 10111
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7750950535212838,
+      "learning_rate": 1.6373733707545022e-06,
+      "loss": 0.6732,
+      "step": 10112
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6832449039725715,
+      "learning_rate": 1.6372989260603535e-06,
+      "loss": 0.6732,
+      "step": 10113
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.4253886399196152,
+      "learning_rate": 1.6372244754182381e-06,
+      "loss": 0.679,
+      "step": 10114
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.480245927437989,
+      "learning_rate": 1.6371500188288498e-06,
+      "loss": 0.6927,
+      "step": 10115
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.466479191525556,
+      "learning_rate": 1.6370755562928841e-06,
+      "loss": 0.6536,
+      "step": 10116
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7190816948484166,
+      "learning_rate": 1.6370010878110364e-06,
+      "loss": 0.7018,
+      "step": 10117
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.72477046966782,
+      "learning_rate": 1.6369266133840007e-06,
+      "loss": 0.6849,
+      "step": 10118
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.5574491173398703,
+      "learning_rate": 1.6368521330124726e-06,
+      "loss": 0.6667,
+      "step": 10119
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.0578772228369018,
+      "learning_rate": 1.6367776466971475e-06,
+      "loss": 0.6764,
+      "step": 10120
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.816104043851193,
+      "learning_rate": 1.6367031544387202e-06,
+      "loss": 0.6764,
+      "step": 10121
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.087994031616014,
+      "learning_rate": 1.636628656237886e-06,
+      "loss": 0.6758,
+      "step": 10122
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7635054481009311,
+      "learning_rate": 1.6365541520953402e-06,
+      "loss": 0.6667,
+      "step": 10123
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8296801488639648,
+      "learning_rate": 1.6364796420117784e-06,
+      "loss": 0.6953,
+      "step": 10124
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.991915572044242,
+      "learning_rate": 1.6364051259878956e-06,
+      "loss": 0.6732,
+      "step": 10125
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 5.667795456664484,
+      "learning_rate": 1.6363306040243873e-06,
+      "loss": 0.6868,
+      "step": 10126
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.3776165322164236,
+      "learning_rate": 1.6362560761219497e-06,
+      "loss": 0.6816,
+      "step": 10127
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7019908845973406,
+      "learning_rate": 1.6361815422812777e-06,
+      "loss": 0.6934,
+      "step": 10128
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.5593749591665698,
+      "learning_rate": 1.636107002503067e-06,
+      "loss": 0.6667,
+      "step": 10129
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8027352966350878,
+      "learning_rate": 1.6360324567880133e-06,
+      "loss": 0.6914,
+      "step": 10130
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.920752979814105,
+      "learning_rate": 1.635957905136813e-06,
+      "loss": 0.6979,
+      "step": 10131
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8087994504539686,
+      "learning_rate": 1.6358833475501605e-06,
+      "loss": 0.6882,
+      "step": 10132
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.703692391021306,
+      "learning_rate": 1.635808784028753e-06,
+      "loss": 0.6784,
+      "step": 10133
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6539948368190013,
+      "learning_rate": 1.6357342145732858e-06,
+      "loss": 0.6536,
+      "step": 10134
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8343285379500907,
+      "learning_rate": 1.635659639184455e-06,
+      "loss": 0.6719,
+      "step": 10135
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.258557928040255,
+      "learning_rate": 1.6355850578629565e-06,
+      "loss": 0.6589,
+      "step": 10136
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.245022044801823,
+      "learning_rate": 1.6355104706094865e-06,
+      "loss": 0.6693,
+      "step": 10137
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8986973764700903,
+      "learning_rate": 1.635435877424741e-06,
+      "loss": 0.6875,
+      "step": 10138
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6622693848536387,
+      "learning_rate": 1.635361278309416e-06,
+      "loss": 0.666,
+      "step": 10139
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9968342198827762,
+      "learning_rate": 1.6352866732642083e-06,
+      "loss": 0.6471,
+      "step": 10140
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.562092781032756,
+      "learning_rate": 1.635212062289814e-06,
+      "loss": 0.6868,
+      "step": 10141
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.6900232073483618,
+      "learning_rate": 1.6351374453869293e-06,
+      "loss": 0.6803,
+      "step": 10142
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.657407062126761,
+      "learning_rate": 1.6350628225562504e-06,
+      "loss": 0.6934,
+      "step": 10143
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.6101743170838416,
+      "learning_rate": 1.6349881937984743e-06,
+      "loss": 0.6549,
+      "step": 10144
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.405400034768449,
+      "learning_rate": 1.634913559114297e-06,
+      "loss": 0.6549,
+      "step": 10145
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.245549562189062,
+      "learning_rate": 1.6348389185044153e-06,
+      "loss": 0.6823,
+      "step": 10146
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.9402360527455706,
+      "learning_rate": 1.6347642719695263e-06,
+      "loss": 0.668,
+      "step": 10147
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.2455841531889265,
+      "learning_rate": 1.6346896195103256e-06,
+      "loss": 0.6738,
+      "step": 10148
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 4.805486460137932,
+      "learning_rate": 1.6346149611275108e-06,
+      "loss": 0.6699,
+      "step": 10149
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.8653157159292948,
+      "learning_rate": 1.6345402968217785e-06,
+      "loss": 0.6823,
+      "step": 10150
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.2604685800643636,
+      "learning_rate": 1.6344656265938258e-06,
+      "loss": 0.7031,
+      "step": 10151
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.91381360066182,
+      "learning_rate": 1.6343909504443488e-06,
+      "loss": 0.6823,
+      "step": 10152
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1935482621625224,
+      "learning_rate": 1.6343162683740452e-06,
+      "loss": 0.6647,
+      "step": 10153
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.056364066680561,
+      "learning_rate": 1.6342415803836115e-06,
+      "loss": 0.681,
+      "step": 10154
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.5386584322584778,
+      "learning_rate": 1.6341668864737454e-06,
+      "loss": 0.6732,
+      "step": 10155
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.6829998032825328,
+      "learning_rate": 1.6340921866451435e-06,
+      "loss": 0.6628,
+      "step": 10156
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.347483403602712,
+      "learning_rate": 1.6340174808985031e-06,
+      "loss": 0.6882,
+      "step": 10157
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.0834124578010924,
+      "learning_rate": 1.6339427692345216e-06,
+      "loss": 0.6725,
+      "step": 10158
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.1403597280166458,
+      "learning_rate": 1.6338680516538959e-06,
+      "loss": 0.6875,
+      "step": 10159
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.6654887463487547,
+      "learning_rate": 1.6337933281573237e-06,
+      "loss": 0.6582,
+      "step": 10160
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.230748631514442,
+      "learning_rate": 1.6337185987455028e-06,
+      "loss": 0.6771,
+      "step": 10161
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.326215881440545,
+      "learning_rate": 1.6336438634191296e-06,
+      "loss": 0.6966,
+      "step": 10162
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.19701067119091,
+      "learning_rate": 1.6335691221789026e-06,
+      "loss": 0.6829,
+      "step": 10163
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.3342579577739264,
+      "learning_rate": 1.6334943750255187e-06,
+      "loss": 0.6471,
+      "step": 10164
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.642670366293431,
+      "learning_rate": 1.6334196219596756e-06,
+      "loss": 0.6816,
+      "step": 10165
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9358583803554025,
+      "learning_rate": 1.6333448629820714e-06,
+      "loss": 0.6602,
+      "step": 10166
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.469903810800908,
+      "learning_rate": 1.6332700980934039e-06,
+      "loss": 0.6927,
+      "step": 10167
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.4981181192704194,
+      "learning_rate": 1.63319532729437e-06,
+      "loss": 0.6654,
+      "step": 10168
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8977779831329404,
+      "learning_rate": 1.6331205505856685e-06,
+      "loss": 0.6764,
+      "step": 10169
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7831993874528095,
+      "learning_rate": 1.6330457679679968e-06,
+      "loss": 0.651,
+      "step": 10170
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1483964579367503,
+      "learning_rate": 1.6329709794420527e-06,
+      "loss": 0.6992,
+      "step": 10171
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.4936508060349545,
+      "learning_rate": 1.6328961850085348e-06,
+      "loss": 0.679,
+      "step": 10172
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.2545513931674224,
+      "learning_rate": 1.6328213846681408e-06,
+      "loss": 0.6895,
+      "step": 10173
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.133046721875583,
+      "learning_rate": 1.6327465784215685e-06,
+      "loss": 0.6602,
+      "step": 10174
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.2539818861505325,
+      "learning_rate": 1.632671766269517e-06,
+      "loss": 0.6569,
+      "step": 10175
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9064549916681074,
+      "learning_rate": 1.6325969482126834e-06,
+      "loss": 0.6901,
+      "step": 10176
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.5717407957664746,
+      "learning_rate": 1.6325221242517668e-06,
+      "loss": 0.6947,
+      "step": 10177
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8082567738896592,
+      "learning_rate": 1.6324472943874653e-06,
+      "loss": 0.6764,
+      "step": 10178
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.61283141286511,
+      "learning_rate": 1.6323724586204773e-06,
+      "loss": 0.6543,
+      "step": 10179
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8816004523360768,
+      "learning_rate": 1.6322976169515012e-06,
+      "loss": 0.679,
+      "step": 10180
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.52387466362683,
+      "learning_rate": 1.6322227693812355e-06,
+      "loss": 0.6908,
+      "step": 10181
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1459706374655456,
+      "learning_rate": 1.6321479159103786e-06,
+      "loss": 0.6829,
+      "step": 10182
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.754960502830345,
+      "learning_rate": 1.6320730565396295e-06,
+      "loss": 0.6973,
+      "step": 10183
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.8752969844217806,
+      "learning_rate": 1.6319981912696867e-06,
+      "loss": 0.6719,
+      "step": 10184
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.9849654555690641,
+      "learning_rate": 1.6319233201012487e-06,
+      "loss": 0.6953,
+      "step": 10185
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.824950459430911,
+      "learning_rate": 1.6318484430350145e-06,
+      "loss": 0.6953,
+      "step": 10186
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.02955346632862,
+      "learning_rate": 1.6317735600716831e-06,
+      "loss": 0.679,
+      "step": 10187
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.535491523559198,
+      "learning_rate": 1.6316986712119532e-06,
+      "loss": 0.6882,
+      "step": 10188
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.722841135211233,
+      "learning_rate": 1.6316237764565237e-06,
+      "loss": 0.6927,
+      "step": 10189
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7416073064439076,
+      "learning_rate": 1.6315488758060934e-06,
+      "loss": 0.6758,
+      "step": 10190
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.3282814774312617,
+      "learning_rate": 1.6314739692613618e-06,
+      "loss": 0.6784,
+      "step": 10191
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6436925990803667,
+      "learning_rate": 1.6313990568230275e-06,
+      "loss": 0.6628,
+      "step": 10192
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.936867254264143,
+      "learning_rate": 1.6313241384917903e-06,
+      "loss": 0.6673,
+      "step": 10193
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.4963431331490813,
+      "learning_rate": 1.631249214268349e-06,
+      "loss": 0.6738,
+      "step": 10194
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.2126435386532717,
+      "learning_rate": 1.631174284153403e-06,
+      "loss": 0.6862,
+      "step": 10195
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.8099546592695273,
+      "learning_rate": 1.6310993481476515e-06,
+      "loss": 0.6706,
+      "step": 10196
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.7826347127334707,
+      "learning_rate": 1.6310244062517945e-06,
+      "loss": 0.6673,
+      "step": 10197
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.74578378026671,
+      "learning_rate": 1.6309494584665303e-06,
+      "loss": 0.6934,
+      "step": 10198
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.425658762701524,
+      "learning_rate": 1.6308745047925587e-06,
+      "loss": 0.6712,
+      "step": 10199
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.7237724959516856,
+      "learning_rate": 1.6307995452305802e-06,
+      "loss": 0.6673,
+      "step": 10200
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5236453477809615,
+      "learning_rate": 1.6307245797812937e-06,
+      "loss": 0.6712,
+      "step": 10201
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6154711134080868,
+      "learning_rate": 1.6306496084453986e-06,
+      "loss": 0.7031,
+      "step": 10202
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.665408669274428,
+      "learning_rate": 1.6305746312235952e-06,
+      "loss": 0.6842,
+      "step": 10203
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.878489482167477,
+      "learning_rate": 1.630499648116583e-06,
+      "loss": 0.6738,
+      "step": 10204
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.2721275679175412,
+      "learning_rate": 1.6304246591250617e-06,
+      "loss": 0.6895,
+      "step": 10205
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6836061066473168,
+      "learning_rate": 1.6303496642497312e-06,
+      "loss": 0.6725,
+      "step": 10206
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6507515114941774,
+      "learning_rate": 1.6302746634912916e-06,
+      "loss": 0.6628,
+      "step": 10207
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.859361787743219,
+      "learning_rate": 1.630199656850443e-06,
+      "loss": 0.6829,
+      "step": 10208
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.5987164002225263,
+      "learning_rate": 1.630124644327885e-06,
+      "loss": 0.6764,
+      "step": 10209
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.6694358461119305,
+      "learning_rate": 1.6300496259243181e-06,
+      "loss": 0.6855,
+      "step": 10210
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1996335294224652,
+      "learning_rate": 1.6299746016404426e-06,
+      "loss": 0.6771,
+      "step": 10211
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.6404601380253476,
+      "learning_rate": 1.6298995714769582e-06,
+      "loss": 0.668,
+      "step": 10212
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 6.8769489161103845,
+      "learning_rate": 1.6298245354345654e-06,
+      "loss": 0.6999,
+      "step": 10213
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.2371408091309215,
+      "learning_rate": 1.6297494935139645e-06,
+      "loss": 0.7116,
+      "step": 10214
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.0834647452488033,
+      "learning_rate": 1.6296744457158557e-06,
+      "loss": 0.6686,
+      "step": 10215
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 5.160371435697365,
+      "learning_rate": 1.6295993920409398e-06,
+      "loss": 0.6576,
+      "step": 10216
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.620089602980297,
+      "learning_rate": 1.6295243324899173e-06,
+      "loss": 0.6699,
+      "step": 10217
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 4.227889594259028,
+      "learning_rate": 1.6294492670634882e-06,
+      "loss": 0.6706,
+      "step": 10218
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.324766123449667,
+      "learning_rate": 1.6293741957623537e-06,
+      "loss": 0.6797,
+      "step": 10219
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.713342542915228,
+      "learning_rate": 1.6292991185872138e-06,
+      "loss": 0.6647,
+      "step": 10220
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.942652588558279,
+      "learning_rate": 1.6292240355387698e-06,
+      "loss": 0.6862,
+      "step": 10221
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.9525528950705553,
+      "learning_rate": 1.6291489466177222e-06,
+      "loss": 0.6673,
+      "step": 10222
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.744284789644142,
+      "learning_rate": 1.629073851824772e-06,
+      "loss": 0.668,
+      "step": 10223
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.1758313500768423,
+      "learning_rate": 1.6289987511606195e-06,
+      "loss": 0.6758,
+      "step": 10224
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.667445527121641,
+      "learning_rate": 1.6289236446259667e-06,
+      "loss": 0.6582,
+      "step": 10225
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8082587166088344,
+      "learning_rate": 1.6288485322215134e-06,
+      "loss": 0.6862,
+      "step": 10226
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9862166431497015,
+      "learning_rate": 1.6287734139479611e-06,
+      "loss": 0.6751,
+      "step": 10227
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.2718635105011007,
+      "learning_rate": 1.6286982898060111e-06,
+      "loss": 0.6771,
+      "step": 10228
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.962802085706511,
+      "learning_rate": 1.6286231597963643e-06,
+      "loss": 0.7129,
+      "step": 10229
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9680763093820512,
+      "learning_rate": 1.628548023919722e-06,
+      "loss": 0.6751,
+      "step": 10230
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.3150360441942235,
+      "learning_rate": 1.6284728821767851e-06,
+      "loss": 0.681,
+      "step": 10231
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9218702634692069,
+      "learning_rate": 1.6283977345682558e-06,
+      "loss": 0.6693,
+      "step": 10232
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8337874253079368,
+      "learning_rate": 1.6283225810948347e-06,
+      "loss": 0.6836,
+      "step": 10233
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.0916383511693777,
+      "learning_rate": 1.6282474217572232e-06,
+      "loss": 0.6504,
+      "step": 10234
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.715499050650289,
+      "learning_rate": 1.6281722565561232e-06,
+      "loss": 0.6621,
+      "step": 10235
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.318202975018482,
+      "learning_rate": 1.6280970854922355e-06,
+      "loss": 0.679,
+      "step": 10236
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.4426900915138532,
+      "learning_rate": 1.6280219085662625e-06,
+      "loss": 0.6777,
+      "step": 10237
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.199743093987751,
+      "learning_rate": 1.6279467257789056e-06,
+      "loss": 0.7135,
+      "step": 10238
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.015986455541469,
+      "learning_rate": 1.627871537130866e-06,
+      "loss": 0.6732,
+      "step": 10239
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.8683534167201508,
+      "learning_rate": 1.627796342622846e-06,
+      "loss": 0.6849,
+      "step": 10240
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.3983919213151395,
+      "learning_rate": 1.6277211422555474e-06,
+      "loss": 0.6797,
+      "step": 10241
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.5341524821475971,
+      "learning_rate": 1.6276459360296716e-06,
+      "loss": 0.6569,
+      "step": 10242
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.551259473386159,
+      "learning_rate": 1.627570723945921e-06,
+      "loss": 0.6816,
+      "step": 10243
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.59239142106943,
+      "learning_rate": 1.6274955060049972e-06,
+      "loss": 0.6849,
+      "step": 10244
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.678634048143368,
+      "learning_rate": 1.6274202822076024e-06,
+      "loss": 0.6725,
+      "step": 10245
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.5182998445144045,
+      "learning_rate": 1.6273450525544386e-06,
+      "loss": 0.6556,
+      "step": 10246
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7173293665131881,
+      "learning_rate": 1.6272698170462077e-06,
+      "loss": 0.6745,
+      "step": 10247
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7211050351314325,
+      "learning_rate": 1.6271945756836125e-06,
+      "loss": 0.6758,
+      "step": 10248
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.4628521836165356,
+      "learning_rate": 1.6271193284673545e-06,
+      "loss": 0.7116,
+      "step": 10249
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.954219297355041,
+      "learning_rate": 1.6270440753981368e-06,
+      "loss": 0.6868,
+      "step": 10250
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.0968153447687723,
+      "learning_rate": 1.626968816476661e-06,
+      "loss": 0.6803,
+      "step": 10251
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7529045146585827,
+      "learning_rate": 1.6268935517036297e-06,
+      "loss": 0.6517,
+      "step": 10252
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.8802577719851397,
+      "learning_rate": 1.6268182810797458e-06,
+      "loss": 0.6803,
+      "step": 10253
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.810232527936292,
+      "learning_rate": 1.626743004605711e-06,
+      "loss": 0.666,
+      "step": 10254
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.750125218198303,
+      "learning_rate": 1.6266677222822287e-06,
+      "loss": 0.7064,
+      "step": 10255
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.1385962040382558,
+      "learning_rate": 1.626592434110001e-06,
+      "loss": 0.6634,
+      "step": 10256
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.3698431921831062,
+      "learning_rate": 1.6265171400897306e-06,
+      "loss": 0.6556,
+      "step": 10257
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.142701844875042,
+      "learning_rate": 1.6264418402221206e-06,
+      "loss": 0.6836,
+      "step": 10258
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.643140119537721,
+      "learning_rate": 1.626366534507873e-06,
+      "loss": 0.668,
+      "step": 10259
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.3189399298953113,
+      "learning_rate": 1.6262912229476915e-06,
+      "loss": 0.6523,
+      "step": 10260
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7935042729905852,
+      "learning_rate": 1.6262159055422785e-06,
+      "loss": 0.6576,
+      "step": 10261
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8061731753004324,
+      "learning_rate": 1.6261405822923373e-06,
+      "loss": 0.6784,
+      "step": 10262
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8363536723769096,
+      "learning_rate": 1.6260652531985706e-06,
+      "loss": 0.6562,
+      "step": 10263
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.4502414668921044,
+      "learning_rate": 1.6259899182616813e-06,
+      "loss": 0.6875,
+      "step": 10264
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7463163408981366,
+      "learning_rate": 1.6259145774823728e-06,
+      "loss": 0.6706,
+      "step": 10265
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.764122146655114,
+      "learning_rate": 1.6258392308613484e-06,
+      "loss": 0.6888,
+      "step": 10266
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.0865046850817155,
+      "learning_rate": 1.625763878399311e-06,
+      "loss": 0.6706,
+      "step": 10267
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.4621657263317278,
+      "learning_rate": 1.625688520096964e-06,
+      "loss": 0.7005,
+      "step": 10268
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.2590185105464897,
+      "learning_rate": 1.6256131559550105e-06,
+      "loss": 0.6882,
+      "step": 10269
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9253857550091246,
+      "learning_rate": 1.6255377859741545e-06,
+      "loss": 0.6751,
+      "step": 10270
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.9315871619080514,
+      "learning_rate": 1.6254624101550987e-06,
+      "loss": 0.6777,
+      "step": 10271
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.2262371059121546,
+      "learning_rate": 1.625387028498547e-06,
+      "loss": 0.6816,
+      "step": 10272
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7818965720107574,
+      "learning_rate": 1.625311641005203e-06,
+      "loss": 0.6849,
+      "step": 10273
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.3743926992465485,
+      "learning_rate": 1.6252362476757697e-06,
+      "loss": 0.6562,
+      "step": 10274
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8283077736629882,
+      "learning_rate": 1.6251608485109519e-06,
+      "loss": 0.6999,
+      "step": 10275
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.2735100480789725,
+      "learning_rate": 1.625085443511452e-06,
+      "loss": 0.6725,
+      "step": 10276
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.378212324952378,
+      "learning_rate": 1.6250100326779746e-06,
+      "loss": 0.7018,
+      "step": 10277
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9177241480908265,
+      "learning_rate": 1.6249346160112233e-06,
+      "loss": 0.6654,
+      "step": 10278
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.9846389662203885,
+      "learning_rate": 1.6248591935119022e-06,
+      "loss": 0.6556,
+      "step": 10279
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.317685696027806,
+      "learning_rate": 1.6247837651807147e-06,
+      "loss": 0.6842,
+      "step": 10280
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.406709687639959,
+      "learning_rate": 1.6247083310183649e-06,
+      "loss": 0.7031,
+      "step": 10281
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.057843366874547,
+      "learning_rate": 1.6246328910255574e-06,
+      "loss": 0.6719,
+      "step": 10282
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.2834437230059264,
+      "learning_rate": 1.6245574452029954e-06,
+      "loss": 0.6921,
+      "step": 10283
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 5.140388752226303,
+      "learning_rate": 1.624481993551384e-06,
+      "loss": 0.6829,
+      "step": 10284
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.836295121526413,
+      "learning_rate": 1.6244065360714267e-06,
+      "loss": 0.7012,
+      "step": 10285
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.785551326687787,
+      "learning_rate": 1.6243310727638282e-06,
+      "loss": 0.6738,
+      "step": 10286
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.442611807000822,
+      "learning_rate": 1.6242556036292924e-06,
+      "loss": 0.6367,
+      "step": 10287
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7092779912331559,
+      "learning_rate": 1.6241801286685239e-06,
+      "loss": 0.6797,
+      "step": 10288
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6807523592277771,
+      "learning_rate": 1.624104647882227e-06,
+      "loss": 0.6589,
+      "step": 10289
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9640673804602893,
+      "learning_rate": 1.6240291612711063e-06,
+      "loss": 0.6549,
+      "step": 10290
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.768386547971458,
+      "learning_rate": 1.6239536688358662e-06,
+      "loss": 0.6686,
+      "step": 10291
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.1268106982083896,
+      "learning_rate": 1.6238781705772113e-06,
+      "loss": 0.6862,
+      "step": 10292
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8007950386447416,
+      "learning_rate": 1.6238026664958465e-06,
+      "loss": 0.6445,
+      "step": 10293
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7380814479108405,
+      "learning_rate": 1.6237271565924759e-06,
+      "loss": 0.6895,
+      "step": 10294
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.332159870948773,
+      "learning_rate": 1.6236516408678047e-06,
+      "loss": 0.7064,
+      "step": 10295
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9272970314236602,
+      "learning_rate": 1.623576119322538e-06,
+      "loss": 0.681,
+      "step": 10296
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.4212373615672518,
+      "learning_rate": 1.6235005919573794e-06,
+      "loss": 0.6784,
+      "step": 10297
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1946902643028976,
+      "learning_rate": 1.6234250587730351e-06,
+      "loss": 0.6654,
+      "step": 10298
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.4818673091315688,
+      "learning_rate": 1.6233495197702097e-06,
+      "loss": 0.6706,
+      "step": 10299
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.0568815813564705,
+      "learning_rate": 1.623273974949608e-06,
+      "loss": 0.681,
+      "step": 10300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.1749948570188,
+      "learning_rate": 1.6231984243119351e-06,
+      "loss": 0.6849,
+      "step": 10301
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8902852780322059,
+      "learning_rate": 1.6231228678578962e-06,
+      "loss": 0.6595,
+      "step": 10302
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.413186285748763,
+      "learning_rate": 1.6230473055881965e-06,
+      "loss": 0.6712,
+      "step": 10303
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.5654114174447125,
+      "learning_rate": 1.6229717375035414e-06,
+      "loss": 0.6855,
+      "step": 10304
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 6.260165424807542,
+      "learning_rate": 1.622896163604636e-06,
+      "loss": 0.668,
+      "step": 10305
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7561792209704052,
+      "learning_rate": 1.622820583892185e-06,
+      "loss": 0.6712,
+      "step": 10306
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9566194838320212,
+      "learning_rate": 1.622744998366895e-06,
+      "loss": 0.6595,
+      "step": 10307
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8920073461346159,
+      "learning_rate": 1.6226694070294706e-06,
+      "loss": 0.6855,
+      "step": 10308
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.3236681698489543,
+      "learning_rate": 1.6225938098806178e-06,
+      "loss": 0.6654,
+      "step": 10309
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8041724735418005,
+      "learning_rate": 1.622518206921042e-06,
+      "loss": 0.681,
+      "step": 10310
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8624253069370428,
+      "learning_rate": 1.6224425981514483e-06,
+      "loss": 0.696,
+      "step": 10311
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.5655836959200857,
+      "learning_rate": 1.622366983572543e-06,
+      "loss": 0.6582,
+      "step": 10312
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.3023426686781667,
+      "learning_rate": 1.6222913631850314e-06,
+      "loss": 0.653,
+      "step": 10313
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.1456280570902195,
+      "learning_rate": 1.6222157369896197e-06,
+      "loss": 0.6367,
+      "step": 10314
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.668001587279381,
+      "learning_rate": 1.6221401049870133e-06,
+      "loss": 0.6628,
+      "step": 10315
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.7439507443593811,
+      "learning_rate": 1.6220644671779187e-06,
+      "loss": 0.6413,
+      "step": 10316
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.327873032148467,
+      "learning_rate": 1.6219888235630413e-06,
+      "loss": 0.7018,
+      "step": 10317
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9829974326608202,
+      "learning_rate": 1.6219131741430868e-06,
+      "loss": 0.6628,
+      "step": 10318
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1637074003834662,
+      "learning_rate": 1.6218375189187618e-06,
+      "loss": 0.6901,
+      "step": 10319
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.05436967926895,
+      "learning_rate": 1.6217618578907723e-06,
+      "loss": 0.6732,
+      "step": 10320
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.069736830476203,
+      "learning_rate": 1.6216861910598247e-06,
+      "loss": 0.696,
+      "step": 10321
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6785577335906408,
+      "learning_rate": 1.6216105184266248e-06,
+      "loss": 0.653,
+      "step": 10322
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.164265326103103,
+      "learning_rate": 1.6215348399918787e-06,
+      "loss": 0.6868,
+      "step": 10323
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.943248145300726,
+      "learning_rate": 1.6214591557562929e-06,
+      "loss": 0.7005,
+      "step": 10324
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.8102200847432615,
+      "learning_rate": 1.6213834657205738e-06,
+      "loss": 0.6602,
+      "step": 10325
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9982958292995333,
+      "learning_rate": 1.6213077698854283e-06,
+      "loss": 0.6862,
+      "step": 10326
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.016349713808062,
+      "learning_rate": 1.6212320682515621e-06,
+      "loss": 0.6543,
+      "step": 10327
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.019242970468254,
+      "learning_rate": 1.6211563608196822e-06,
+      "loss": 0.6738,
+      "step": 10328
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.58131986323554,
+      "learning_rate": 1.621080647590495e-06,
+      "loss": 0.6934,
+      "step": 10329
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.8016794903308844,
+      "learning_rate": 1.6210049285647074e-06,
+      "loss": 0.6842,
+      "step": 10330
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.3104563031084524,
+      "learning_rate": 1.6209292037430255e-06,
+      "loss": 0.6602,
+      "step": 10331
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.9500694719970144,
+      "learning_rate": 1.6208534731261568e-06,
+      "loss": 0.6628,
+      "step": 10332
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.815335827733258,
+      "learning_rate": 1.6207777367148076e-06,
+      "loss": 0.6901,
+      "step": 10333
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6106602085952437,
+      "learning_rate": 1.6207019945096847e-06,
+      "loss": 0.6641,
+      "step": 10334
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9335654738618404,
+      "learning_rate": 1.6206262465114954e-06,
+      "loss": 0.668,
+      "step": 10335
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.9871359402541584,
+      "learning_rate": 1.6205504927209463e-06,
+      "loss": 0.6875,
+      "step": 10336
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.4486935626367714,
+      "learning_rate": 1.6204747331387448e-06,
+      "loss": 0.6628,
+      "step": 10337
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.746217134066712,
+      "learning_rate": 1.6203989677655977e-06,
+      "loss": 0.6673,
+      "step": 10338
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.5443834761754296,
+      "learning_rate": 1.6203231966022119e-06,
+      "loss": 0.6953,
+      "step": 10339
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.7000358389088794,
+      "learning_rate": 1.620247419649295e-06,
+      "loss": 0.6693,
+      "step": 10340
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.04412248752864,
+      "learning_rate": 1.6201716369075543e-06,
+      "loss": 0.6816,
+      "step": 10341
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.2586514283544883,
+      "learning_rate": 1.6200958483776966e-06,
+      "loss": 0.6595,
+      "step": 10342
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.889012669517161,
+      "learning_rate": 1.6200200540604295e-06,
+      "loss": 0.6595,
+      "step": 10343
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.8976146040744,
+      "learning_rate": 1.6199442539564606e-06,
+      "loss": 0.681,
+      "step": 10344
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.5545503481266008,
+      "learning_rate": 1.6198684480664968e-06,
+      "loss": 0.6719,
+      "step": 10345
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.540908788680958,
+      "learning_rate": 1.6197926363912461e-06,
+      "loss": 0.6855,
+      "step": 10346
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.95318830408814,
+      "learning_rate": 1.619716818931416e-06,
+      "loss": 0.6628,
+      "step": 10347
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.981416715286128,
+      "learning_rate": 1.6196409956877142e-06,
+      "loss": 0.6621,
+      "step": 10348
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.73716482164375,
+      "learning_rate": 1.619565166660848e-06,
+      "loss": 0.6673,
+      "step": 10349
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.122998877901755,
+      "learning_rate": 1.6194893318515255e-06,
+      "loss": 0.6654,
+      "step": 10350
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 5.198227964604503,
+      "learning_rate": 1.619413491260454e-06,
+      "loss": 0.6797,
+      "step": 10351
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.750103007225608,
+      "learning_rate": 1.6193376448883418e-06,
+      "loss": 0.6797,
+      "step": 10352
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1045226776613553,
+      "learning_rate": 1.6192617927358963e-06,
+      "loss": 0.6608,
+      "step": 10353
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.2814242860591682,
+      "learning_rate": 1.6191859348038261e-06,
+      "loss": 0.6562,
+      "step": 10354
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.14377832476718,
+      "learning_rate": 1.6191100710928387e-06,
+      "loss": 0.6615,
+      "step": 10355
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.6870690612340367,
+      "learning_rate": 1.6190342016036422e-06,
+      "loss": 0.6797,
+      "step": 10356
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.3209144662399384,
+      "learning_rate": 1.6189583263369445e-06,
+      "loss": 0.6517,
+      "step": 10357
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.5603025809848408,
+      "learning_rate": 1.6188824452934542e-06,
+      "loss": 0.6862,
+      "step": 10358
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.351726551970594,
+      "learning_rate": 1.6188065584738795e-06,
+      "loss": 0.6849,
+      "step": 10359
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.192022690367436,
+      "learning_rate": 1.6187306658789282e-06,
+      "loss": 0.6458,
+      "step": 10360
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.788007555369339,
+      "learning_rate": 1.6186547675093093e-06,
+      "loss": 0.7051,
+      "step": 10361
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.504266123160117,
+      "learning_rate": 1.6185788633657303e-06,
+      "loss": 0.6673,
+      "step": 10362
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6865134965610802,
+      "learning_rate": 1.6185029534489e-06,
+      "loss": 0.6797,
+      "step": 10363
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9341424785068296,
+      "learning_rate": 1.6184270377595272e-06,
+      "loss": 0.6647,
+      "step": 10364
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.0815860615065533,
+      "learning_rate": 1.6183511162983203e-06,
+      "loss": 0.6647,
+      "step": 10365
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.5945081871429245,
+      "learning_rate": 1.6182751890659876e-06,
+      "loss": 0.6908,
+      "step": 10366
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8763319451401799,
+      "learning_rate": 1.6181992560632379e-06,
+      "loss": 0.6901,
+      "step": 10367
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.741498208980225,
+      "learning_rate": 1.6181233172907796e-06,
+      "loss": 0.6888,
+      "step": 10368
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8143414862425047,
+      "learning_rate": 1.618047372749322e-06,
+      "loss": 0.6621,
+      "step": 10369
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.329865665458528,
+      "learning_rate": 1.6179714224395736e-06,
+      "loss": 0.7077,
+      "step": 10370
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.5759771294002705,
+      "learning_rate": 1.6178954663622432e-06,
+      "loss": 0.6725,
+      "step": 10371
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.214199554003566,
+      "learning_rate": 1.6178195045180397e-06,
+      "loss": 0.6764,
+      "step": 10372
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1235622170957353,
+      "learning_rate": 1.6177435369076722e-06,
+      "loss": 0.7038,
+      "step": 10373
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.5525366613637863,
+      "learning_rate": 1.6176675635318496e-06,
+      "loss": 0.6523,
+      "step": 10374
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.581232288898547,
+      "learning_rate": 1.617591584391281e-06,
+      "loss": 0.6706,
+      "step": 10375
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.3223048462399847,
+      "learning_rate": 1.6175155994866758e-06,
+      "loss": 0.6589,
+      "step": 10376
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.72471007887352,
+      "learning_rate": 1.6174396088187425e-06,
+      "loss": 0.6849,
+      "step": 10377
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.330658386130743,
+      "learning_rate": 1.6173636123881909e-06,
+      "loss": 0.6686,
+      "step": 10378
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8001712346950672,
+      "learning_rate": 1.6172876101957302e-06,
+      "loss": 0.6615,
+      "step": 10379
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 5.383317877560884,
+      "learning_rate": 1.6172116022420696e-06,
+      "loss": 0.6712,
+      "step": 10380
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.5328668184902223,
+      "learning_rate": 1.6171355885279182e-06,
+      "loss": 0.6673,
+      "step": 10381
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.174624306618927,
+      "learning_rate": 1.6170595690539864e-06,
+      "loss": 0.6634,
+      "step": 10382
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.2616399382115104,
+      "learning_rate": 1.6169835438209827e-06,
+      "loss": 0.6836,
+      "step": 10383
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.4267215406453206,
+      "learning_rate": 1.616907512829617e-06,
+      "loss": 0.6895,
+      "step": 10384
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.131815108711594,
+      "learning_rate": 1.616831476080599e-06,
+      "loss": 0.6589,
+      "step": 10385
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.1992642015711996,
+      "learning_rate": 1.6167554335746383e-06,
+      "loss": 0.6829,
+      "step": 10386
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.069198644927804,
+      "learning_rate": 1.6166793853124446e-06,
+      "loss": 0.6608,
+      "step": 10387
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8047024208391222,
+      "learning_rate": 1.6166033312947277e-06,
+      "loss": 0.6582,
+      "step": 10388
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.896511442813455,
+      "learning_rate": 1.6165272715221971e-06,
+      "loss": 0.6895,
+      "step": 10389
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.0310259108127324,
+      "learning_rate": 1.6164512059955634e-06,
+      "loss": 0.6908,
+      "step": 10390
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7994083102015301,
+      "learning_rate": 1.6163751347155356e-06,
+      "loss": 0.6784,
+      "step": 10391
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.9506858265245428,
+      "learning_rate": 1.6162990576828243e-06,
+      "loss": 0.651,
+      "step": 10392
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.4007130857528476,
+      "learning_rate": 1.6162229748981395e-06,
+      "loss": 0.6803,
+      "step": 10393
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.798553516880961,
+      "learning_rate": 1.6161468863621912e-06,
+      "loss": 0.6758,
+      "step": 10394
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 4.206464170282407,
+      "learning_rate": 1.6160707920756896e-06,
+      "loss": 0.6764,
+      "step": 10395
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.9388289846270292,
+      "learning_rate": 1.6159946920393446e-06,
+      "loss": 0.6764,
+      "step": 10396
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8489396704147095,
+      "learning_rate": 1.6159185862538667e-06,
+      "loss": 0.6706,
+      "step": 10397
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.1757773705151244,
+      "learning_rate": 1.6158424747199663e-06,
+      "loss": 0.7038,
+      "step": 10398
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.2249919493466384,
+      "learning_rate": 1.6157663574383538e-06,
+      "loss": 0.6562,
+      "step": 10399
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 7.365167283942752,
+      "learning_rate": 1.615690234409739e-06,
+      "loss": 0.6569,
+      "step": 10400
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.369963269935487,
+      "learning_rate": 1.6156141056348326e-06,
+      "loss": 0.681,
+      "step": 10401
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.650694683702199,
+      "learning_rate": 1.6155379711143459e-06,
+      "loss": 0.6738,
+      "step": 10402
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8692712755574945,
+      "learning_rate": 1.6154618308489888e-06,
+      "loss": 0.6699,
+      "step": 10403
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.5486848130848325,
+      "learning_rate": 1.6153856848394719e-06,
+      "loss": 0.6426,
+      "step": 10404
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.3769222281434983,
+      "learning_rate": 1.6153095330865058e-06,
+      "loss": 0.6784,
+      "step": 10405
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.3887591117591862,
+      "learning_rate": 1.6152333755908017e-06,
+      "loss": 0.6725,
+      "step": 10406
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.410583961766482,
+      "learning_rate": 1.6151572123530698e-06,
+      "loss": 0.6641,
+      "step": 10407
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.176482983639999,
+      "learning_rate": 1.6150810433740215e-06,
+      "loss": 0.6706,
+      "step": 10408
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.2539466590024766,
+      "learning_rate": 1.6150048686543673e-06,
+      "loss": 0.6777,
+      "step": 10409
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.9741635487264337,
+      "learning_rate": 1.6149286881948183e-06,
+      "loss": 0.6888,
+      "step": 10410
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.8209258110351887,
+      "learning_rate": 1.6148525019960855e-06,
+      "loss": 0.6784,
+      "step": 10411
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1262730757025194,
+      "learning_rate": 1.6147763100588798e-06,
+      "loss": 0.6536,
+      "step": 10412
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.229446980983648,
+      "learning_rate": 1.6147001123839123e-06,
+      "loss": 0.6419,
+      "step": 10413
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.1248102650360874,
+      "learning_rate": 1.6146239089718948e-06,
+      "loss": 0.6712,
+      "step": 10414
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.883114823114334,
+      "learning_rate": 1.6145476998235376e-06,
+      "loss": 0.6816,
+      "step": 10415
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 6.456584065868253,
+      "learning_rate": 1.6144714849395525e-06,
+      "loss": 0.6868,
+      "step": 10416
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.349353327782616,
+      "learning_rate": 1.6143952643206506e-06,
+      "loss": 0.666,
+      "step": 10417
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.2332579356050517,
+      "learning_rate": 1.6143190379675438e-06,
+      "loss": 0.6816,
+      "step": 10418
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0158465766087084,
+      "learning_rate": 1.6142428058809425e-06,
+      "loss": 0.6706,
+      "step": 10419
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0892002413418176,
+      "learning_rate": 1.614166568061559e-06,
+      "loss": 0.6764,
+      "step": 10420
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.4195318732827586,
+      "learning_rate": 1.6140903245101046e-06,
+      "loss": 0.6764,
+      "step": 10421
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.061846028194124,
+      "learning_rate": 1.614014075227291e-06,
+      "loss": 0.6758,
+      "step": 10422
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9899584425524364,
+      "learning_rate": 1.6139378202138295e-06,
+      "loss": 0.6634,
+      "step": 10423
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.955512872978637,
+      "learning_rate": 1.613861559470432e-06,
+      "loss": 0.6855,
+      "step": 10424
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.7157983782957074,
+      "learning_rate": 1.6137852929978107e-06,
+      "loss": 0.7077,
+      "step": 10425
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1569380509452167,
+      "learning_rate": 1.6137090207966766e-06,
+      "loss": 0.6895,
+      "step": 10426
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0958081136112015,
+      "learning_rate": 1.6136327428677421e-06,
+      "loss": 0.6432,
+      "step": 10427
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.8908274638195035,
+      "learning_rate": 1.613556459211719e-06,
+      "loss": 0.6432,
+      "step": 10428
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9807690588725883,
+      "learning_rate": 1.613480169829319e-06,
+      "loss": 0.6849,
+      "step": 10429
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.7634343014871123,
+      "learning_rate": 1.6134038747212544e-06,
+      "loss": 0.6504,
+      "step": 10430
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8422792012989204,
+      "learning_rate": 1.6133275738882373e-06,
+      "loss": 0.651,
+      "step": 10431
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9802672663108236,
+      "learning_rate": 1.6132512673309794e-06,
+      "loss": 0.6634,
+      "step": 10432
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.769275077767195,
+      "learning_rate": 1.6131749550501937e-06,
+      "loss": 0.6973,
+      "step": 10433
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.618417782396874,
+      "learning_rate": 1.6130986370465914e-06,
+      "loss": 0.6771,
+      "step": 10434
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.2837721955170236,
+      "learning_rate": 1.6130223133208857e-06,
+      "loss": 0.6484,
+      "step": 10435
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8728846283375432,
+      "learning_rate": 1.6129459838737885e-06,
+      "loss": 0.6719,
+      "step": 10436
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.6443744587216584,
+      "learning_rate": 1.6128696487060118e-06,
+      "loss": 0.6686,
+      "step": 10437
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.414006712068593,
+      "learning_rate": 1.612793307818269e-06,
+      "loss": 0.6842,
+      "step": 10438
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.880968944746577,
+      "learning_rate": 1.6127169612112717e-06,
+      "loss": 0.6868,
+      "step": 10439
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.924138063683255,
+      "learning_rate": 1.6126406088857332e-06,
+      "loss": 0.6647,
+      "step": 10440
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.8105705087558817,
+      "learning_rate": 1.6125642508423653e-06,
+      "loss": 0.7051,
+      "step": 10441
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.8269868290764024,
+      "learning_rate": 1.6124878870818812e-06,
+      "loss": 0.6764,
+      "step": 10442
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8412735988853681,
+      "learning_rate": 1.6124115176049935e-06,
+      "loss": 0.6751,
+      "step": 10443
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.29874572931978,
+      "learning_rate": 1.612335142412415e-06,
+      "loss": 0.7038,
+      "step": 10444
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.2465152967572943,
+      "learning_rate": 1.6122587615048585e-06,
+      "loss": 0.666,
+      "step": 10445
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.673608730443005,
+      "learning_rate": 1.6121823748830367e-06,
+      "loss": 0.6725,
+      "step": 10446
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.003444823455845,
+      "learning_rate": 1.6121059825476628e-06,
+      "loss": 0.6641,
+      "step": 10447
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.278584169801551,
+      "learning_rate": 1.6120295844994495e-06,
+      "loss": 0.6712,
+      "step": 10448
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.701231225212683,
+      "learning_rate": 1.6119531807391101e-06,
+      "loss": 0.6855,
+      "step": 10449
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.5803940278464905,
+      "learning_rate": 1.6118767712673576e-06,
+      "loss": 0.6855,
+      "step": 10450
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0860018128550684,
+      "learning_rate": 1.6118003560849047e-06,
+      "loss": 0.6901,
+      "step": 10451
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.753730623367058,
+      "learning_rate": 1.6117239351924654e-06,
+      "loss": 0.6497,
+      "step": 10452
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.3078542400341875,
+      "learning_rate": 1.6116475085907522e-06,
+      "loss": 0.6947,
+      "step": 10453
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9325891366307528,
+      "learning_rate": 1.611571076280479e-06,
+      "loss": 0.6927,
+      "step": 10454
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.343129013444942,
+      "learning_rate": 1.6114946382623587e-06,
+      "loss": 0.6849,
+      "step": 10455
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7728169738890908,
+      "learning_rate": 1.6114181945371047e-06,
+      "loss": 0.6439,
+      "step": 10456
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.387393004323564,
+      "learning_rate": 1.6113417451054309e-06,
+      "loss": 0.6699,
+      "step": 10457
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.5437382887293705,
+      "learning_rate": 1.6112652899680504e-06,
+      "loss": 0.679,
+      "step": 10458
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.1233492622932046,
+      "learning_rate": 1.6111888291256768e-06,
+      "loss": 0.6784,
+      "step": 10459
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.490913850229282,
+      "learning_rate": 1.611112362579024e-06,
+      "loss": 0.6582,
+      "step": 10460
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6004067589826256,
+      "learning_rate": 1.6110358903288056e-06,
+      "loss": 0.6751,
+      "step": 10461
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.678892270608283,
+      "learning_rate": 1.6109594123757348e-06,
+      "loss": 0.6784,
+      "step": 10462
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.76935470413852,
+      "learning_rate": 1.6108829287205258e-06,
+      "loss": 0.6582,
+      "step": 10463
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.372485025747459,
+      "learning_rate": 1.6108064393638926e-06,
+      "loss": 0.6641,
+      "step": 10464
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7081553294186177,
+      "learning_rate": 1.6107299443065491e-06,
+      "loss": 0.6582,
+      "step": 10465
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.314668946029745,
+      "learning_rate": 1.6106534435492088e-06,
+      "loss": 0.6699,
+      "step": 10466
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3340896894664265,
+      "learning_rate": 1.6105769370925857e-06,
+      "loss": 0.6719,
+      "step": 10467
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.9546058309904315,
+      "learning_rate": 1.6105004249373946e-06,
+      "loss": 0.7064,
+      "step": 10468
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.569812357202883,
+      "learning_rate": 1.6104239070843485e-06,
+      "loss": 0.681,
+      "step": 10469
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8295144847152031,
+      "learning_rate": 1.6103473835341622e-06,
+      "loss": 0.681,
+      "step": 10470
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3793225498629087,
+      "learning_rate": 1.61027085428755e-06,
+      "loss": 0.6621,
+      "step": 10471
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.5435876234694472,
+      "learning_rate": 1.6101943193452257e-06,
+      "loss": 0.6927,
+      "step": 10472
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7012742021446208,
+      "learning_rate": 1.610117778707904e-06,
+      "loss": 0.6608,
+      "step": 10473
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9449859337811843,
+      "learning_rate": 1.6100412323762992e-06,
+      "loss": 0.6829,
+      "step": 10474
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.1959659774631164,
+      "learning_rate": 1.6099646803511253e-06,
+      "loss": 0.6764,
+      "step": 10475
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7992540534575984,
+      "learning_rate": 1.6098881226330973e-06,
+      "loss": 0.6602,
+      "step": 10476
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.6723373752643966,
+      "learning_rate": 1.6098115592229295e-06,
+      "loss": 0.6712,
+      "step": 10477
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7289105751564607,
+      "learning_rate": 1.6097349901213367e-06,
+      "loss": 0.6686,
+      "step": 10478
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 4.0699256618476465,
+      "learning_rate": 1.609658415329033e-06,
+      "loss": 0.6719,
+      "step": 10479
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7942858259121693,
+      "learning_rate": 1.6095818348467333e-06,
+      "loss": 0.6517,
+      "step": 10480
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.6811791266427902,
+      "learning_rate": 1.6095052486751528e-06,
+      "loss": 0.6882,
+      "step": 10481
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.840642315135,
+      "learning_rate": 1.6094286568150058e-06,
+      "loss": 0.6302,
+      "step": 10482
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9069955900969143,
+      "learning_rate": 1.609352059267007e-06,
+      "loss": 0.6862,
+      "step": 10483
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.5729942039322156,
+      "learning_rate": 1.609275456031872e-06,
+      "loss": 0.6966,
+      "step": 10484
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 5.059352259643909,
+      "learning_rate": 1.609198847110315e-06,
+      "loss": 0.6888,
+      "step": 10485
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0541187815473374,
+      "learning_rate": 1.6091222325030513e-06,
+      "loss": 0.6934,
+      "step": 10486
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.9819551252196392,
+      "learning_rate": 1.6090456122107958e-06,
+      "loss": 0.6999,
+      "step": 10487
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.0696898906949532,
+      "learning_rate": 1.6089689862342642e-06,
+      "loss": 0.6257,
+      "step": 10488
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9021072518924748,
+      "learning_rate": 1.6088923545741708e-06,
+      "loss": 0.6953,
+      "step": 10489
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.2898816622882154,
+      "learning_rate": 1.6088157172312314e-06,
+      "loss": 0.6784,
+      "step": 10490
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8671226129773821,
+      "learning_rate": 1.608739074206161e-06,
+      "loss": 0.6836,
+      "step": 10491
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9630630986460217,
+      "learning_rate": 1.6086624254996748e-06,
+      "loss": 0.6673,
+      "step": 10492
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9534212601201989,
+      "learning_rate": 1.6085857711124886e-06,
+      "loss": 0.6751,
+      "step": 10493
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8063841834616705,
+      "learning_rate": 1.6085091110453176e-06,
+      "loss": 0.653,
+      "step": 10494
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9418307657614774,
+      "learning_rate": 1.6084324452988768e-06,
+      "loss": 0.6589,
+      "step": 10495
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.966916627813492,
+      "learning_rate": 1.6083557738738827e-06,
+      "loss": 0.6276,
+      "step": 10496
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0025156114418055,
+      "learning_rate": 1.6082790967710502e-06,
+      "loss": 0.6641,
+      "step": 10497
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.1493269929461927,
+      "learning_rate": 1.608202413991095e-06,
+      "loss": 0.6497,
+      "step": 10498
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8445852254329119,
+      "learning_rate": 1.6081257255347331e-06,
+      "loss": 0.668,
+      "step": 10499
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8219434825447665,
+      "learning_rate": 1.6080490314026799e-06,
+      "loss": 0.6641,
+      "step": 10500
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.056938920929201,
+      "learning_rate": 1.6079723315956516e-06,
+      "loss": 0.6654,
+      "step": 10501
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7341374336036666,
+      "learning_rate": 1.6078956261143631e-06,
+      "loss": 0.64,
+      "step": 10502
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.5803581192531526,
+      "learning_rate": 1.6078189149595313e-06,
+      "loss": 0.6589,
+      "step": 10503
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.44271908965502,
+      "learning_rate": 1.6077421981318719e-06,
+      "loss": 0.6719,
+      "step": 10504
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.4986974029924798,
+      "learning_rate": 1.6076654756321008e-06,
+      "loss": 0.6576,
+      "step": 10505
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.6179058766408483,
+      "learning_rate": 1.6075887474609342e-06,
+      "loss": 0.679,
+      "step": 10506
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.761890268668514,
+      "learning_rate": 1.6075120136190878e-06,
+      "loss": 0.6732,
+      "step": 10507
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0889639481834938,
+      "learning_rate": 1.6074352741072783e-06,
+      "loss": 0.6641,
+      "step": 10508
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0308724621585843,
+      "learning_rate": 1.6073585289262214e-06,
+      "loss": 0.6185,
+      "step": 10509
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.3667034155828235,
+      "learning_rate": 1.6072817780766336e-06,
+      "loss": 0.6862,
+      "step": 10510
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.4876183823846394,
+      "learning_rate": 1.6072050215592316e-06,
+      "loss": 0.6576,
+      "step": 10511
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.562099297514402,
+      "learning_rate": 1.6071282593747313e-06,
+      "loss": 0.7012,
+      "step": 10512
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.1425871169952027,
+      "learning_rate": 1.6070514915238494e-06,
+      "loss": 0.6361,
+      "step": 10513
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.568115755048413,
+      "learning_rate": 1.606974718007302e-06,
+      "loss": 0.6621,
+      "step": 10514
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.268286407262168,
+      "learning_rate": 1.6068979388258058e-06,
+      "loss": 0.6667,
+      "step": 10515
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.230876487428838,
+      "learning_rate": 1.6068211539800779e-06,
+      "loss": 0.7012,
+      "step": 10516
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.5971326087346056,
+      "learning_rate": 1.6067443634708342e-06,
+      "loss": 0.6673,
+      "step": 10517
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.8659130673162037,
+      "learning_rate": 1.6066675672987917e-06,
+      "loss": 0.694,
+      "step": 10518
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.4778930943929722,
+      "learning_rate": 1.6065907654646673e-06,
+      "loss": 0.6608,
+      "step": 10519
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.1795504921000597,
+      "learning_rate": 1.6065139579691773e-06,
+      "loss": 0.6738,
+      "step": 10520
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.2139336278060138,
+      "learning_rate": 1.6064371448130394e-06,
+      "loss": 0.6895,
+      "step": 10521
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.688882433914158,
+      "learning_rate": 1.6063603259969696e-06,
+      "loss": 0.6647,
+      "step": 10522
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.571281083275598,
+      "learning_rate": 1.6062835015216854e-06,
+      "loss": 0.6654,
+      "step": 10523
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.347959191372006,
+      "learning_rate": 1.6062066713879037e-06,
+      "loss": 0.6725,
+      "step": 10524
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.991300195623274,
+      "learning_rate": 1.6061298355963414e-06,
+      "loss": 0.6849,
+      "step": 10525
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.7359229731184926,
+      "learning_rate": 1.6060529941477157e-06,
+      "loss": 0.7012,
+      "step": 10526
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.9510153699113486,
+      "learning_rate": 1.605976147042744e-06,
+      "loss": 0.6777,
+      "step": 10527
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.497766326398461,
+      "learning_rate": 1.605899294282143e-06,
+      "loss": 0.6751,
+      "step": 10528
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8911976721575352,
+      "learning_rate": 1.6058224358666307e-06,
+      "loss": 0.6413,
+      "step": 10529
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8074764355196067,
+      "learning_rate": 1.6057455717969237e-06,
+      "loss": 0.6725,
+      "step": 10530
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.232085591843132,
+      "learning_rate": 1.60566870207374e-06,
+      "loss": 0.6816,
+      "step": 10531
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2507801545933415,
+      "learning_rate": 1.6055918266977969e-06,
+      "loss": 0.6595,
+      "step": 10532
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.1183630046251014,
+      "learning_rate": 1.6055149456698112e-06,
+      "loss": 0.7292,
+      "step": 10533
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 4.111647265753686,
+      "learning_rate": 1.6054380589905012e-06,
+      "loss": 0.6673,
+      "step": 10534
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.39721277268262,
+      "learning_rate": 1.6053611666605841e-06,
+      "loss": 0.6667,
+      "step": 10535
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.8732491309080523,
+      "learning_rate": 1.6052842686807782e-06,
+      "loss": 0.6465,
+      "step": 10536
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8850797768324148,
+      "learning_rate": 1.6052073650518003e-06,
+      "loss": 0.6901,
+      "step": 10537
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7271359044288235,
+      "learning_rate": 1.6051304557743683e-06,
+      "loss": 0.6673,
+      "step": 10538
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.17085331479119,
+      "learning_rate": 1.6050535408492008e-06,
+      "loss": 0.6842,
+      "step": 10539
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.5805172998029935,
+      "learning_rate": 1.6049766202770147e-06,
+      "loss": 0.6901,
+      "step": 10540
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0262321421689766,
+      "learning_rate": 1.6048996940585285e-06,
+      "loss": 0.6764,
+      "step": 10541
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.994189389193963,
+      "learning_rate": 1.60482276219446e-06,
+      "loss": 0.6419,
+      "step": 10542
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.2393789170433736,
+      "learning_rate": 1.6047458246855267e-06,
+      "loss": 0.6582,
+      "step": 10543
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.662197849008676,
+      "learning_rate": 1.6046688815324478e-06,
+      "loss": 0.6823,
+      "step": 10544
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2414422723840777,
+      "learning_rate": 1.6045919327359406e-06,
+      "loss": 0.6862,
+      "step": 10545
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.345804822999088,
+      "learning_rate": 1.604514978296723e-06,
+      "loss": 0.6914,
+      "step": 10546
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6791962668558984,
+      "learning_rate": 1.6044380182155139e-06,
+      "loss": 0.6634,
+      "step": 10547
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6795543597212046,
+      "learning_rate": 1.6043610524930316e-06,
+      "loss": 0.6823,
+      "step": 10548
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.925246111511015,
+      "learning_rate": 1.6042840811299937e-06,
+      "loss": 0.6803,
+      "step": 10549
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.869999670301417,
+      "learning_rate": 1.6042071041271195e-06,
+      "loss": 0.6882,
+      "step": 10550
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.6950603424681123,
+      "learning_rate": 1.6041301214851264e-06,
+      "loss": 0.7077,
+      "step": 10551
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.9226251768241063,
+      "learning_rate": 1.6040531332047337e-06,
+      "loss": 0.6523,
+      "step": 10552
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.328540241169398,
+      "learning_rate": 1.6039761392866598e-06,
+      "loss": 0.7018,
+      "step": 10553
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 6.422729052141753,
+      "learning_rate": 1.6038991397316232e-06,
+      "loss": 0.7077,
+      "step": 10554
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.3660643045467609,
+      "learning_rate": 1.6038221345403425e-06,
+      "loss": 0.6784,
+      "step": 10555
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6911833034906448,
+      "learning_rate": 1.6037451237135362e-06,
+      "loss": 0.666,
+      "step": 10556
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.0362565916142636,
+      "learning_rate": 1.6036681072519235e-06,
+      "loss": 0.6751,
+      "step": 10557
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.480279198683516,
+      "learning_rate": 1.6035910851562229e-06,
+      "loss": 0.6953,
+      "step": 10558
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.3578972504825313,
+      "learning_rate": 1.6035140574271533e-06,
+      "loss": 0.6673,
+      "step": 10559
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.9030141254474753,
+      "learning_rate": 1.6034370240654335e-06,
+      "loss": 0.6842,
+      "step": 10560
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.5051463536826981,
+      "learning_rate": 1.6033599850717828e-06,
+      "loss": 0.6523,
+      "step": 10561
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.7094218003369117,
+      "learning_rate": 1.60328294044692e-06,
+      "loss": 0.6738,
+      "step": 10562
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6958792909678274,
+      "learning_rate": 1.6032058901915638e-06,
+      "loss": 0.6706,
+      "step": 10563
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.029556403930683,
+      "learning_rate": 1.6031288343064344e-06,
+      "loss": 0.6589,
+      "step": 10564
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0649091403655686,
+      "learning_rate": 1.6030517727922497e-06,
+      "loss": 0.6738,
+      "step": 10565
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.288391926346313,
+      "learning_rate": 1.6029747056497293e-06,
+      "loss": 0.6862,
+      "step": 10566
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.134857556255149,
+      "learning_rate": 1.6028976328795933e-06,
+      "loss": 0.6914,
+      "step": 10567
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7599892704219504,
+      "learning_rate": 1.6028205544825601e-06,
+      "loss": 0.6829,
+      "step": 10568
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.769851129747811,
+      "learning_rate": 1.602743470459349e-06,
+      "loss": 0.6712,
+      "step": 10569
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6890045963794594,
+      "learning_rate": 1.6026663808106803e-06,
+      "loss": 0.6523,
+      "step": 10570
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1401793961356061,
+      "learning_rate": 1.6025892855372727e-06,
+      "loss": 0.6745,
+      "step": 10571
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.0200274592024288,
+      "learning_rate": 1.602512184639846e-06,
+      "loss": 0.6387,
+      "step": 10572
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.5301738127992706,
+      "learning_rate": 1.6024350781191198e-06,
+      "loss": 0.6712,
+      "step": 10573
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.573280576143697,
+      "learning_rate": 1.602357965975814e-06,
+      "loss": 0.7012,
+      "step": 10574
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.1351852365840585,
+      "learning_rate": 1.6022808482106476e-06,
+      "loss": 0.6458,
+      "step": 10575
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7801267586196359,
+      "learning_rate": 1.602203724824341e-06,
+      "loss": 0.681,
+      "step": 10576
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9120230164791362,
+      "learning_rate": 1.602126595817614e-06,
+      "loss": 0.6947,
+      "step": 10577
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.011129982992828,
+      "learning_rate": 1.6020494611911857e-06,
+      "loss": 0.6764,
+      "step": 10578
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.3188877881084622,
+      "learning_rate": 1.601972320945777e-06,
+      "loss": 0.681,
+      "step": 10579
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7891343293961847,
+      "learning_rate": 1.6018951750821072e-06,
+      "loss": 0.6693,
+      "step": 10580
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7677286112890448,
+      "learning_rate": 1.6018180236008965e-06,
+      "loss": 0.666,
+      "step": 10581
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 5.793699944805356,
+      "learning_rate": 1.601740866502865e-06,
+      "loss": 0.6797,
+      "step": 10582
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.5913012597313587,
+      "learning_rate": 1.601663703788733e-06,
+      "loss": 0.696,
+      "step": 10583
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.810590512131866,
+      "learning_rate": 1.6015865354592202e-06,
+      "loss": 0.6615,
+      "step": 10584
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.980934617927779,
+      "learning_rate": 1.601509361515047e-06,
+      "loss": 0.6289,
+      "step": 10585
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.9154921264179414,
+      "learning_rate": 1.6014321819569338e-06,
+      "loss": 0.6484,
+      "step": 10586
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.4783851137350486,
+      "learning_rate": 1.6013549967856009e-06,
+      "loss": 0.6699,
+      "step": 10587
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0182967173849495,
+      "learning_rate": 1.6012778060017684e-06,
+      "loss": 0.6719,
+      "step": 10588
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.7791925551726,
+      "learning_rate": 1.6012006096061575e-06,
+      "loss": 0.6979,
+      "step": 10589
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0593376391074572,
+      "learning_rate": 1.6011234075994877e-06,
+      "loss": 0.6647,
+      "step": 10590
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.222938338093076,
+      "learning_rate": 1.60104619998248e-06,
+      "loss": 0.6257,
+      "step": 10591
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1081541329480593,
+      "learning_rate": 1.600968986755855e-06,
+      "loss": 0.6849,
+      "step": 10592
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.829697359572362,
+      "learning_rate": 1.6008917679203338e-06,
+      "loss": 0.6549,
+      "step": 10593
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.776584122502331,
+      "learning_rate": 1.6008145434766357e-06,
+      "loss": 0.6712,
+      "step": 10594
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7777194891030494,
+      "learning_rate": 1.600737313425483e-06,
+      "loss": 0.6562,
+      "step": 10595
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.059008514204396,
+      "learning_rate": 1.6006600777675953e-06,
+      "loss": 0.6868,
+      "step": 10596
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.0984679018450267,
+      "learning_rate": 1.6005828365036947e-06,
+      "loss": 0.6497,
+      "step": 10597
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.423789444159164,
+      "learning_rate": 1.6005055896345008e-06,
+      "loss": 0.6699,
+      "step": 10598
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.011301914623093,
+      "learning_rate": 1.6004283371607352e-06,
+      "loss": 0.6432,
+      "step": 10599
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.4182904796778697,
+      "learning_rate": 1.600351079083119e-06,
+      "loss": 0.6471,
+      "step": 10600
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7262612373146613,
+      "learning_rate": 1.600273815402373e-06,
+      "loss": 0.6738,
+      "step": 10601
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6397212747798364,
+      "learning_rate": 1.6001965461192183e-06,
+      "loss": 0.6439,
+      "step": 10602
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.9626154413982173,
+      "learning_rate": 1.6001192712343764e-06,
+      "loss": 0.7005,
+      "step": 10603
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.2771636034846354,
+      "learning_rate": 1.6000419907485681e-06,
+      "loss": 0.7174,
+      "step": 10604
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.657170592516209,
+      "learning_rate": 1.5999647046625149e-06,
+      "loss": 0.6549,
+      "step": 10605
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.9985807661005817,
+      "learning_rate": 1.5998874129769377e-06,
+      "loss": 0.6797,
+      "step": 10606
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.494396830085671,
+      "learning_rate": 1.5998101156925588e-06,
+      "loss": 0.6283,
+      "step": 10607
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.3853377739016746,
+      "learning_rate": 1.5997328128100986e-06,
+      "loss": 0.7174,
+      "step": 10608
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.967399143866023,
+      "learning_rate": 1.5996555043302794e-06,
+      "loss": 0.6882,
+      "step": 10609
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.595210671247009,
+      "learning_rate": 1.599578190253822e-06,
+      "loss": 0.6693,
+      "step": 10610
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.9700704433892473,
+      "learning_rate": 1.5995008705814486e-06,
+      "loss": 0.6836,
+      "step": 10611
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.685540473503203,
+      "learning_rate": 1.5994235453138804e-06,
+      "loss": 0.666,
+      "step": 10612
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 6.404822826004761,
+      "learning_rate": 1.5993462144518396e-06,
+      "loss": 0.6816,
+      "step": 10613
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1426517716359084,
+      "learning_rate": 1.5992688779960467e-06,
+      "loss": 0.6699,
+      "step": 10614
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1688884010138954,
+      "learning_rate": 1.599191535947225e-06,
+      "loss": 0.653,
+      "step": 10615
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.1527693762638274,
+      "learning_rate": 1.5991141883060958e-06,
+      "loss": 0.6829,
+      "step": 10616
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.426717734221776,
+      "learning_rate": 1.5990368350733808e-06,
+      "loss": 0.7298,
+      "step": 10617
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.021500784005617,
+      "learning_rate": 1.5989594762498017e-06,
+      "loss": 0.6803,
+      "step": 10618
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9417019948029158,
+      "learning_rate": 1.598882111836081e-06,
+      "loss": 0.6536,
+      "step": 10619
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.3413359916624525,
+      "learning_rate": 1.5988047418329408e-06,
+      "loss": 0.6699,
+      "step": 10620
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 4.937748628096285,
+      "learning_rate": 1.5987273662411028e-06,
+      "loss": 0.6901,
+      "step": 10621
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7889605400185694,
+      "learning_rate": 1.5986499850612896e-06,
+      "loss": 0.6406,
+      "step": 10622
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.1073054624065533,
+      "learning_rate": 1.5985725982942228e-06,
+      "loss": 0.6706,
+      "step": 10623
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.3816757168926435,
+      "learning_rate": 1.5984952059406251e-06,
+      "loss": 0.6764,
+      "step": 10624
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.2824976092225695,
+      "learning_rate": 1.598417808001219e-06,
+      "loss": 0.6693,
+      "step": 10625
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7026981551250129,
+      "learning_rate": 1.5983404044767265e-06,
+      "loss": 0.668,
+      "step": 10626
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8771504860245654,
+      "learning_rate": 1.59826299536787e-06,
+      "loss": 0.6582,
+      "step": 10627
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7108004689071977,
+      "learning_rate": 1.5981855806753718e-06,
+      "loss": 0.6608,
+      "step": 10628
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7714838366315084,
+      "learning_rate": 1.598108160399955e-06,
+      "loss": 0.6712,
+      "step": 10629
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8367612259249367,
+      "learning_rate": 1.5980307345423415e-06,
+      "loss": 0.6641,
+      "step": 10630
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8966157648471349,
+      "learning_rate": 1.5979533031032547e-06,
+      "loss": 0.6497,
+      "step": 10631
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8331620198486638,
+      "learning_rate": 1.5978758660834165e-06,
+      "loss": 0.6712,
+      "step": 10632
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.771180123256769,
+      "learning_rate": 1.5977984234835502e-06,
+      "loss": 0.6608,
+      "step": 10633
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.5516637302637508,
+      "learning_rate": 1.5977209753043783e-06,
+      "loss": 0.6784,
+      "step": 10634
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3017273692137188,
+      "learning_rate": 1.5976435215466236e-06,
+      "loss": 0.6738,
+      "step": 10635
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.403537993157435,
+      "learning_rate": 1.5975660622110091e-06,
+      "loss": 0.681,
+      "step": 10636
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.9574350781853074,
+      "learning_rate": 1.5974885972982577e-06,
+      "loss": 0.6849,
+      "step": 10637
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9292851676999485,
+      "learning_rate": 1.597411126809092e-06,
+      "loss": 0.6699,
+      "step": 10638
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8730967398121867,
+      "learning_rate": 1.597333650744236e-06,
+      "loss": 0.6803,
+      "step": 10639
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.456041282839581,
+      "learning_rate": 1.597256169104412e-06,
+      "loss": 0.6719,
+      "step": 10640
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.281112859466946,
+      "learning_rate": 1.5971786818903434e-06,
+      "loss": 0.6536,
+      "step": 10641
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0074577082982474,
+      "learning_rate": 1.5971011891027532e-06,
+      "loss": 0.6771,
+      "step": 10642
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8523547660667197,
+      "learning_rate": 1.5970236907423652e-06,
+      "loss": 0.6562,
+      "step": 10643
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.033143600878901,
+      "learning_rate": 1.5969461868099017e-06,
+      "loss": 0.6576,
+      "step": 10644
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.409255167182134,
+      "learning_rate": 1.5968686773060869e-06,
+      "loss": 0.6654,
+      "step": 10645
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.622157846534333,
+      "learning_rate": 1.596791162231644e-06,
+      "loss": 0.7253,
+      "step": 10646
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9096765650283816,
+      "learning_rate": 1.5967136415872966e-06,
+      "loss": 0.6895,
+      "step": 10647
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.5532586403075905,
+      "learning_rate": 1.596636115373768e-06,
+      "loss": 0.6426,
+      "step": 10648
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7730724549930181,
+      "learning_rate": 1.5965585835917814e-06,
+      "loss": 0.6673,
+      "step": 10649
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.1498719377281543,
+      "learning_rate": 1.596481046242061e-06,
+      "loss": 0.6882,
+      "step": 10650
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.6424530245592868,
+      "learning_rate": 1.5964035033253303e-06,
+      "loss": 0.679,
+      "step": 10651
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.467615551990316,
+      "learning_rate": 1.5963259548423126e-06,
+      "loss": 0.6465,
+      "step": 10652
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.6528650991326326,
+      "learning_rate": 1.5962484007937324e-06,
+      "loss": 0.6621,
+      "step": 10653
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6246412450045586,
+      "learning_rate": 1.5961708411803133e-06,
+      "loss": 0.6751,
+      "step": 10654
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.267480013579725,
+      "learning_rate": 1.5960932760027785e-06,
+      "loss": 0.7083,
+      "step": 10655
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.9951273662726403,
+      "learning_rate": 1.5960157052618528e-06,
+      "loss": 0.6934,
+      "step": 10656
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.030775258616441,
+      "learning_rate": 1.5959381289582597e-06,
+      "loss": 0.6706,
+      "step": 10657
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.665651312529856,
+      "learning_rate": 1.5958605470927234e-06,
+      "loss": 0.6686,
+      "step": 10658
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.313112068091202,
+      "learning_rate": 1.5957829596659677e-06,
+      "loss": 0.6992,
+      "step": 10659
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.033066050832694,
+      "learning_rate": 1.5957053666787173e-06,
+      "loss": 0.6888,
+      "step": 10660
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9297313249052698,
+      "learning_rate": 1.5956277681316958e-06,
+      "loss": 0.6764,
+      "step": 10661
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0708667391061506,
+      "learning_rate": 1.5955501640256277e-06,
+      "loss": 0.6549,
+      "step": 10662
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.570193347516926,
+      "learning_rate": 1.5954725543612373e-06,
+      "loss": 0.7259,
+      "step": 10663
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.680448370569042,
+      "learning_rate": 1.595394939139249e-06,
+      "loss": 0.6842,
+      "step": 10664
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.4437611236176404,
+      "learning_rate": 1.5953173183603868e-06,
+      "loss": 0.6673,
+      "step": 10665
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.318647775293041,
+      "learning_rate": 1.5952396920253756e-06,
+      "loss": 0.6764,
+      "step": 10666
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.306790190559772,
+      "learning_rate": 1.5951620601349398e-06,
+      "loss": 0.6855,
+      "step": 10667
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6867292376386129,
+      "learning_rate": 1.5950844226898035e-06,
+      "loss": 0.6849,
+      "step": 10668
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.0049917527696812,
+      "learning_rate": 1.5950067796906922e-06,
+      "loss": 0.6901,
+      "step": 10669
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.0026955809263685,
+      "learning_rate": 1.5949291311383295e-06,
+      "loss": 0.6569,
+      "step": 10670
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6503777245222949,
+      "learning_rate": 1.5948514770334409e-06,
+      "loss": 0.6654,
+      "step": 10671
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.507385231468381,
+      "learning_rate": 1.5947738173767507e-06,
+      "loss": 0.6862,
+      "step": 10672
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 5.124630378002624,
+      "learning_rate": 1.5946961521689839e-06,
+      "loss": 0.7168,
+      "step": 10673
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.5247507404687974,
+      "learning_rate": 1.5946184814108654e-06,
+      "loss": 0.6803,
+      "step": 10674
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.0422876594629236,
+      "learning_rate": 1.59454080510312e-06,
+      "loss": 0.6738,
+      "step": 10675
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.453949133921399,
+      "learning_rate": 1.5944631232464727e-06,
+      "loss": 0.681,
+      "step": 10676
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8134794439606867,
+      "learning_rate": 1.5943854358416485e-06,
+      "loss": 0.6829,
+      "step": 10677
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.6545994086370523,
+      "learning_rate": 1.5943077428893724e-06,
+      "loss": 0.6673,
+      "step": 10678
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.9192731800212324,
+      "learning_rate": 1.5942300443903696e-06,
+      "loss": 0.6895,
+      "step": 10679
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.4495670578153312,
+      "learning_rate": 1.5941523403453652e-06,
+      "loss": 0.6719,
+      "step": 10680
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8132524103752882,
+      "learning_rate": 1.5940746307550845e-06,
+      "loss": 0.6927,
+      "step": 10681
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.94279390186323,
+      "learning_rate": 1.5939969156202526e-06,
+      "loss": 0.6862,
+      "step": 10682
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.2486886393640533,
+      "learning_rate": 1.5939191949415953e-06,
+      "loss": 0.6667,
+      "step": 10683
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.301575910197921,
+      "learning_rate": 1.5938414687198374e-06,
+      "loss": 0.681,
+      "step": 10684
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3174056340125417,
+      "learning_rate": 1.5937637369557044e-06,
+      "loss": 0.6712,
+      "step": 10685
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9660387727673612,
+      "learning_rate": 1.593685999649922e-06,
+      "loss": 0.6966,
+      "step": 10686
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.340156427698749,
+      "learning_rate": 1.593608256803216e-06,
+      "loss": 0.6842,
+      "step": 10687
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.608520264510315,
+      "learning_rate": 1.5935305084163114e-06,
+      "loss": 0.6973,
+      "step": 10688
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.6728076513367613,
+      "learning_rate": 1.593452754489934e-06,
+      "loss": 0.6777,
+      "step": 10689
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.4438963766426107,
+      "learning_rate": 1.59337499502481e-06,
+      "loss": 0.6686,
+      "step": 10690
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.7284689620129163,
+      "learning_rate": 1.593297230021664e-06,
+      "loss": 0.6686,
+      "step": 10691
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.924935696238743,
+      "learning_rate": 1.5932194594812232e-06,
+      "loss": 0.6751,
+      "step": 10692
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 4.691111700898623,
+      "learning_rate": 1.593141683404212e-06,
+      "loss": 0.6803,
+      "step": 10693
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.593891751759368,
+      "learning_rate": 1.5930639017913573e-06,
+      "loss": 0.6712,
+      "step": 10694
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7253660964438966,
+      "learning_rate": 1.592986114643385e-06,
+      "loss": 0.681,
+      "step": 10695
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6097844679411142,
+      "learning_rate": 1.592908321961021e-06,
+      "loss": 0.6751,
+      "step": 10696
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.6158702998651684,
+      "learning_rate": 1.5928305237449906e-06,
+      "loss": 0.696,
+      "step": 10697
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.745575638458016,
+      "learning_rate": 1.5927527199960207e-06,
+      "loss": 0.6615,
+      "step": 10698
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1314246081152448,
+      "learning_rate": 1.5926749107148374e-06,
+      "loss": 0.6901,
+      "step": 10699
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.0707266450240085,
+      "learning_rate": 1.5925970959021667e-06,
+      "loss": 0.694,
+      "step": 10700
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.788200541215602,
+      "learning_rate": 1.5925192755587346e-06,
+      "loss": 0.6934,
+      "step": 10701
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.006777770856037,
+      "learning_rate": 1.592441449685268e-06,
+      "loss": 0.6602,
+      "step": 10702
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9666629648458682,
+      "learning_rate": 1.592363618282493e-06,
+      "loss": 0.6751,
+      "step": 10703
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.43487497401501,
+      "learning_rate": 1.592285781351136e-06,
+      "loss": 0.6849,
+      "step": 10704
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9746821281149624,
+      "learning_rate": 1.5922079388919231e-06,
+      "loss": 0.6725,
+      "step": 10705
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.554987932295611,
+      "learning_rate": 1.5921300909055814e-06,
+      "loss": 0.6576,
+      "step": 10706
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 5.720227953642612,
+      "learning_rate": 1.5920522373928372e-06,
+      "loss": 0.6862,
+      "step": 10707
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.67744336402437,
+      "learning_rate": 1.591974378354417e-06,
+      "loss": 0.6751,
+      "step": 10708
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.328357537237208,
+      "learning_rate": 1.5918965137910478e-06,
+      "loss": 0.6895,
+      "step": 10709
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.088685418858503,
+      "learning_rate": 1.5918186437034557e-06,
+      "loss": 0.7012,
+      "step": 10710
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.74788435758057,
+      "learning_rate": 1.5917407680923681e-06,
+      "loss": 0.6901,
+      "step": 10711
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.02055177013486,
+      "learning_rate": 1.5916628869585118e-06,
+      "loss": 0.6771,
+      "step": 10712
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.3472880575100725,
+      "learning_rate": 1.5915850003026133e-06,
+      "loss": 0.6686,
+      "step": 10713
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.0629192619985117,
+      "learning_rate": 1.5915071081253993e-06,
+      "loss": 0.6895,
+      "step": 10714
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.1463311461267063,
+      "learning_rate": 1.5914292104275973e-06,
+      "loss": 0.6999,
+      "step": 10715
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.563892733234567,
+      "learning_rate": 1.5913513072099348e-06,
+      "loss": 0.6771,
+      "step": 10716
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1228198770737108,
+      "learning_rate": 1.5912733984731375e-06,
+      "loss": 0.6849,
+      "step": 10717
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.390781599933482,
+      "learning_rate": 1.5911954842179336e-06,
+      "loss": 0.6641,
+      "step": 10718
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9398167749610081,
+      "learning_rate": 1.5911175644450498e-06,
+      "loss": 0.6595,
+      "step": 10719
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9282566888158597,
+      "learning_rate": 1.5910396391552137e-06,
+      "loss": 0.6621,
+      "step": 10720
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.422812918261563,
+      "learning_rate": 1.5909617083491522e-06,
+      "loss": 0.6732,
+      "step": 10721
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9727448175650432,
+      "learning_rate": 1.590883772027593e-06,
+      "loss": 0.6673,
+      "step": 10722
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.84949478348144,
+      "learning_rate": 1.5908058301912633e-06,
+      "loss": 0.6706,
+      "step": 10723
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.2801541163470995,
+      "learning_rate": 1.5907278828408903e-06,
+      "loss": 0.6595,
+      "step": 10724
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.2282261751697985,
+      "learning_rate": 1.5906499299772018e-06,
+      "loss": 0.6875,
+      "step": 10725
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7894510526407118,
+      "learning_rate": 1.5905719716009253e-06,
+      "loss": 0.6901,
+      "step": 10726
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7106367162166289,
+      "learning_rate": 1.5904940077127884e-06,
+      "loss": 0.694,
+      "step": 10727
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2551500201796133,
+      "learning_rate": 1.5904160383135189e-06,
+      "loss": 0.6504,
+      "step": 10728
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7432296116665978,
+      "learning_rate": 1.5903380634038439e-06,
+      "loss": 0.6836,
+      "step": 10729
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.9873200881372637,
+      "learning_rate": 1.5902600829844919e-06,
+      "loss": 0.6777,
+      "step": 10730
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.268509414764705,
+      "learning_rate": 1.59018209705619e-06,
+      "loss": 0.6556,
+      "step": 10731
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.888306897847804,
+      "learning_rate": 1.5901041056196669e-06,
+      "loss": 0.7064,
+      "step": 10732
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.0660517219801737,
+      "learning_rate": 1.5900261086756495e-06,
+      "loss": 0.6784,
+      "step": 10733
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.937547280617474,
+      "learning_rate": 1.5899481062248663e-06,
+      "loss": 0.6654,
+      "step": 10734
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2588070975410581,
+      "learning_rate": 1.5898700982680454e-06,
+      "loss": 0.6888,
+      "step": 10735
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.5352139096429003,
+      "learning_rate": 1.5897920848059147e-06,
+      "loss": 0.6654,
+      "step": 10736
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.116683518067418,
+      "learning_rate": 1.5897140658392022e-06,
+      "loss": 0.6699,
+      "step": 10737
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7095881932046281,
+      "learning_rate": 1.5896360413686365e-06,
+      "loss": 0.6569,
+      "step": 10738
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.8883852315106506,
+      "learning_rate": 1.5895580113949452e-06,
+      "loss": 0.681,
+      "step": 10739
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.1007755771851004,
+      "learning_rate": 1.589479975918857e-06,
+      "loss": 0.6712,
+      "step": 10740
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.61274264136697,
+      "learning_rate": 1.5894019349410998e-06,
+      "loss": 0.6543,
+      "step": 10741
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8317078484030825,
+      "learning_rate": 1.5893238884624025e-06,
+      "loss": 0.6641,
+      "step": 10742
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.560262629643608,
+      "learning_rate": 1.5892458364834932e-06,
+      "loss": 0.6771,
+      "step": 10743
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.3555370783367344,
+      "learning_rate": 1.5891677790051e-06,
+      "loss": 0.6602,
+      "step": 10744
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.4929734821333795,
+      "learning_rate": 1.5890897160279524e-06,
+      "loss": 0.6582,
+      "step": 10745
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.5992400376050881,
+      "learning_rate": 1.589011647552778e-06,
+      "loss": 0.6725,
+      "step": 10746
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.4744135664287725,
+      "learning_rate": 1.5889335735803057e-06,
+      "loss": 0.6608,
+      "step": 10747
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.3110259343483635,
+      "learning_rate": 1.5888554941112646e-06,
+      "loss": 0.6686,
+      "step": 10748
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.070564526440586,
+      "learning_rate": 1.5887774091463832e-06,
+      "loss": 0.6745,
+      "step": 10749
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.9745304571720066,
+      "learning_rate": 1.5886993186863898e-06,
+      "loss": 0.6523,
+      "step": 10750
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8897892429336022,
+      "learning_rate": 1.5886212227320137e-06,
+      "loss": 0.6628,
+      "step": 10751
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.4851566701575993,
+      "learning_rate": 1.5885431212839836e-06,
+      "loss": 0.6758,
+      "step": 10752
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 4.460264021534012,
+      "learning_rate": 1.5884650143430287e-06,
+      "loss": 0.696,
+      "step": 10753
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.008518071483874,
+      "learning_rate": 1.5883869019098776e-06,
+      "loss": 0.6751,
+      "step": 10754
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.320455374493063,
+      "learning_rate": 1.5883087839852597e-06,
+      "loss": 0.6797,
+      "step": 10755
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 6.263873494971331,
+      "learning_rate": 1.5882306605699038e-06,
+      "loss": 0.6751,
+      "step": 10756
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.7519917864491772,
+      "learning_rate": 1.5881525316645392e-06,
+      "loss": 0.6484,
+      "step": 10757
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.804009674418962,
+      "learning_rate": 1.5880743972698948e-06,
+      "loss": 0.679,
+      "step": 10758
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9464000078837473,
+      "learning_rate": 1.5879962573867004e-06,
+      "loss": 0.6634,
+      "step": 10759
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.989751078430376,
+      "learning_rate": 1.5879181120156849e-06,
+      "loss": 0.6803,
+      "step": 10760
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.984977823520119,
+      "learning_rate": 1.5878399611575772e-06,
+      "loss": 0.6771,
+      "step": 10761
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8424366550717994,
+      "learning_rate": 1.5877618048131076e-06,
+      "loss": 0.6725,
+      "step": 10762
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.4436700013960118,
+      "learning_rate": 1.5876836429830046e-06,
+      "loss": 0.6732,
+      "step": 10763
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.510815761183359,
+      "learning_rate": 1.5876054756679987e-06,
+      "loss": 0.6504,
+      "step": 10764
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.4326630572519288,
+      "learning_rate": 1.587527302868819e-06,
+      "loss": 0.6829,
+      "step": 10765
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.0053335511658967,
+      "learning_rate": 1.5874491245861946e-06,
+      "loss": 0.6888,
+      "step": 10766
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.987335684911339,
+      "learning_rate": 1.5873709408208557e-06,
+      "loss": 0.7168,
+      "step": 10767
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.8842837370971095,
+      "learning_rate": 1.5872927515735318e-06,
+      "loss": 0.6999,
+      "step": 10768
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.826600319711791,
+      "learning_rate": 1.5872145568449532e-06,
+      "loss": 0.6829,
+      "step": 10769
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.44416003557016,
+      "learning_rate": 1.5871363566358483e-06,
+      "loss": 0.6673,
+      "step": 10770
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7770266677490889,
+      "learning_rate": 1.5870581509469486e-06,
+      "loss": 0.6667,
+      "step": 10771
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.5102882188211637,
+      "learning_rate": 1.5869799397789831e-06,
+      "loss": 0.6647,
+      "step": 10772
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.975335184941419,
+      "learning_rate": 1.5869017231326818e-06,
+      "loss": 0.6504,
+      "step": 10773
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.5435798002658043,
+      "learning_rate": 1.5868235010087748e-06,
+      "loss": 0.6699,
+      "step": 10774
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.935605089526534,
+      "learning_rate": 1.5867452734079925e-06,
+      "loss": 0.6595,
+      "step": 10775
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 5.339411442489267,
+      "learning_rate": 1.5866670403310644e-06,
+      "loss": 0.6641,
+      "step": 10776
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.329686457943396,
+      "learning_rate": 1.5865888017787207e-06,
+      "loss": 0.6641,
+      "step": 10777
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.498784453455429,
+      "learning_rate": 1.5865105577516924e-06,
+      "loss": 0.7129,
+      "step": 10778
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7789620399115162,
+      "learning_rate": 1.5864323082507087e-06,
+      "loss": 0.6758,
+      "step": 10779
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9356948104704966,
+      "learning_rate": 1.5863540532765008e-06,
+      "loss": 0.6868,
+      "step": 10780
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9203659334465277,
+      "learning_rate": 1.5862757928297984e-06,
+      "loss": 0.6504,
+      "step": 10781
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8643464585613446,
+      "learning_rate": 1.5861975269113321e-06,
+      "loss": 0.6628,
+      "step": 10782
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8539930337497562,
+      "learning_rate": 1.5861192555218327e-06,
+      "loss": 0.6602,
+      "step": 10783
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.1983933400382965,
+      "learning_rate": 1.58604097866203e-06,
+      "loss": 0.679,
+      "step": 10784
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8606396505430961,
+      "learning_rate": 1.5859626963326558e-06,
+      "loss": 0.6693,
+      "step": 10785
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.415935133179074,
+      "learning_rate": 1.5858844085344392e-06,
+      "loss": 0.696,
+      "step": 10786
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.0670825538669866,
+      "learning_rate": 1.585806115268112e-06,
+      "loss": 0.6608,
+      "step": 10787
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.5796433344825997,
+      "learning_rate": 1.5857278165344043e-06,
+      "loss": 0.6712,
+      "step": 10788
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8560322375196245,
+      "learning_rate": 1.5856495123340473e-06,
+      "loss": 0.6706,
+      "step": 10789
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.355464693331478,
+      "learning_rate": 1.5855712026677714e-06,
+      "loss": 0.6901,
+      "step": 10790
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9036202297966078,
+      "learning_rate": 1.5854928875363078e-06,
+      "loss": 0.6686,
+      "step": 10791
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.3082789436045046,
+      "learning_rate": 1.5854145669403874e-06,
+      "loss": 0.6608,
+      "step": 10792
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8185196913128634,
+      "learning_rate": 1.5853362408807409e-06,
+      "loss": 0.681,
+      "step": 10793
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.0814875935971164,
+      "learning_rate": 1.5852579093580995e-06,
+      "loss": 0.6647,
+      "step": 10794
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.5072810013207376,
+      "learning_rate": 1.5851795723731946e-06,
+      "loss": 0.6673,
+      "step": 10795
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8338129255290356,
+      "learning_rate": 1.5851012299267565e-06,
+      "loss": 0.6738,
+      "step": 10796
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.5785368659353956,
+      "learning_rate": 1.5850228820195172e-06,
+      "loss": 0.6764,
+      "step": 10797
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.922504041496011,
+      "learning_rate": 1.5849445286522076e-06,
+      "loss": 0.6445,
+      "step": 10798
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.2077574509629128,
+      "learning_rate": 1.5848661698255588e-06,
+      "loss": 0.6999,
+      "step": 10799
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.181838320551803,
+      "learning_rate": 1.5847878055403024e-06,
+      "loss": 0.6634,
+      "step": 10800
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9090246468365284,
+      "learning_rate": 1.5847094357971697e-06,
+      "loss": 0.6927,
+      "step": 10801
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.160670703723923,
+      "learning_rate": 1.5846310605968923e-06,
+      "loss": 0.6829,
+      "step": 10802
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.5521866857602546,
+      "learning_rate": 1.5845526799402013e-06,
+      "loss": 0.6582,
+      "step": 10803
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1524373554235046,
+      "learning_rate": 1.5844742938278284e-06,
+      "loss": 0.7025,
+      "step": 10804
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.459830967953604,
+      "learning_rate": 1.5843959022605054e-06,
+      "loss": 0.6751,
+      "step": 10805
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9204885100904291,
+      "learning_rate": 1.5843175052389638e-06,
+      "loss": 0.6816,
+      "step": 10806
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9154824729524065,
+      "learning_rate": 1.5842391027639348e-06,
+      "loss": 0.6764,
+      "step": 10807
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.0530523004041314,
+      "learning_rate": 1.5841606948361513e-06,
+      "loss": 0.6654,
+      "step": 10808
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.399015165149739,
+      "learning_rate": 1.5840822814563439e-06,
+      "loss": 0.6855,
+      "step": 10809
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9700485861753707,
+      "learning_rate": 1.584003862625245e-06,
+      "loss": 0.6576,
+      "step": 10810
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 4.157575594843804,
+      "learning_rate": 1.5839254383435866e-06,
+      "loss": 0.6751,
+      "step": 10811
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8208198363976994,
+      "learning_rate": 1.5838470086121001e-06,
+      "loss": 0.6504,
+      "step": 10812
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.701890409286022,
+      "learning_rate": 1.5837685734315178e-06,
+      "loss": 0.6361,
+      "step": 10813
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.318702008932996,
+      "learning_rate": 1.5836901328025717e-06,
+      "loss": 0.627,
+      "step": 10814
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8840069938854406,
+      "learning_rate": 1.5836116867259945e-06,
+      "loss": 0.6797,
+      "step": 10815
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9457181562743554,
+      "learning_rate": 1.5835332352025177e-06,
+      "loss": 0.6589,
+      "step": 10816
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9511584922581308,
+      "learning_rate": 1.5834547782328733e-06,
+      "loss": 0.6569,
+      "step": 10817
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.838789389001184,
+      "learning_rate": 1.5833763158177941e-06,
+      "loss": 0.6855,
+      "step": 10818
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.2298633252801436,
+      "learning_rate": 1.5832978479580121e-06,
+      "loss": 0.6641,
+      "step": 10819
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.0685196309837743,
+      "learning_rate": 1.5832193746542596e-06,
+      "loss": 0.6706,
+      "step": 10820
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.6796460528194717,
+      "learning_rate": 1.5831408959072694e-06,
+      "loss": 0.6699,
+      "step": 10821
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.0022230524563915,
+      "learning_rate": 1.5830624117177732e-06,
+      "loss": 0.6823,
+      "step": 10822
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7144892447726907,
+      "learning_rate": 1.5829839220865041e-06,
+      "loss": 0.6608,
+      "step": 10823
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1293658431918006,
+      "learning_rate": 1.5829054270141948e-06,
+      "loss": 0.6602,
+      "step": 10824
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 3.9747387722455483,
+      "learning_rate": 1.5828269265015775e-06,
+      "loss": 0.6823,
+      "step": 10825
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.0157743125964502,
+      "learning_rate": 1.5827484205493847e-06,
+      "loss": 0.6543,
+      "step": 10826
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.669339999355637,
+      "learning_rate": 1.5826699091583493e-06,
+      "loss": 0.6849,
+      "step": 10827
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.1894414658972168,
+      "learning_rate": 1.5825913923292044e-06,
+      "loss": 0.7025,
+      "step": 10828
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.0801933180660597,
+      "learning_rate": 1.5825128700626823e-06,
+      "loss": 0.6803,
+      "step": 10829
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.322973495327259,
+      "learning_rate": 1.5824343423595163e-06,
+      "loss": 0.6836,
+      "step": 10830
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8848285910613736,
+      "learning_rate": 1.582355809220439e-06,
+      "loss": 0.6686,
+      "step": 10831
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.3239464082898382,
+      "learning_rate": 1.5822772706461832e-06,
+      "loss": 0.6569,
+      "step": 10832
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9466892314774642,
+      "learning_rate": 1.5821987266374826e-06,
+      "loss": 0.6686,
+      "step": 10833
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0459272382200813,
+      "learning_rate": 1.5821201771950695e-06,
+      "loss": 0.6615,
+      "step": 10834
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9967695945441474,
+      "learning_rate": 1.5820416223196773e-06,
+      "loss": 0.6836,
+      "step": 10835
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8770053486898904,
+      "learning_rate": 1.581963062012039e-06,
+      "loss": 0.6595,
+      "step": 10836
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.910773594725433,
+      "learning_rate": 1.5818844962728881e-06,
+      "loss": 0.6536,
+      "step": 10837
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.5074522117771987,
+      "learning_rate": 1.581805925102958e-06,
+      "loss": 0.6875,
+      "step": 10838
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.9596419648770977,
+      "learning_rate": 1.5817273485029814e-06,
+      "loss": 0.6641,
+      "step": 10839
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0510076392205372,
+      "learning_rate": 1.581648766473692e-06,
+      "loss": 0.6523,
+      "step": 10840
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9817154594639532,
+      "learning_rate": 1.5815701790158236e-06,
+      "loss": 0.6751,
+      "step": 10841
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.301259661785676,
+      "learning_rate": 1.5814915861301091e-06,
+      "loss": 0.6595,
+      "step": 10842
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.0114616294354986,
+      "learning_rate": 1.581412987817282e-06,
+      "loss": 0.679,
+      "step": 10843
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.572221454907409,
+      "learning_rate": 1.5813343840780763e-06,
+      "loss": 0.6491,
+      "step": 10844
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.845910476239507,
+      "learning_rate": 1.5812557749132257e-06,
+      "loss": 0.6777,
+      "step": 10845
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.956480241470625,
+      "learning_rate": 1.581177160323463e-06,
+      "loss": 0.679,
+      "step": 10846
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.4073459833751856,
+      "learning_rate": 1.5810985403095223e-06,
+      "loss": 0.6777,
+      "step": 10847
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3430517690590422,
+      "learning_rate": 1.5810199148721382e-06,
+      "loss": 0.6836,
+      "step": 10848
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.9588820669757094,
+      "learning_rate": 1.5809412840120433e-06,
+      "loss": 0.6914,
+      "step": 10849
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.1009034027596245,
+      "learning_rate": 1.5808626477299725e-06,
+      "loss": 0.6556,
+      "step": 10850
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0830545971914356,
+      "learning_rate": 1.580784006026659e-06,
+      "loss": 0.6868,
+      "step": 10851
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.0278537749453733,
+      "learning_rate": 1.5807053589028367e-06,
+      "loss": 0.6602,
+      "step": 10852
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9222050747071919,
+      "learning_rate": 1.5806267063592405e-06,
+      "loss": 0.6823,
+      "step": 10853
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.186732253476257,
+      "learning_rate": 1.5805480483966034e-06,
+      "loss": 0.6641,
+      "step": 10854
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9094514025724776,
+      "learning_rate": 1.5804693850156605e-06,
+      "loss": 0.6562,
+      "step": 10855
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.6634288653517064,
+      "learning_rate": 1.580390716217145e-06,
+      "loss": 0.6335,
+      "step": 10856
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.0525260277492583,
+      "learning_rate": 1.5803120420017919e-06,
+      "loss": 0.6732,
+      "step": 10857
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8136244536584318,
+      "learning_rate": 1.5802333623703351e-06,
+      "loss": 0.6745,
+      "step": 10858
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.8010558419343305,
+      "learning_rate": 1.5801546773235091e-06,
+      "loss": 0.6556,
+      "step": 10859
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.629618943054998,
+      "learning_rate": 1.5800759868620483e-06,
+      "loss": 0.6465,
+      "step": 10860
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.4128627436945658,
+      "learning_rate": 1.5799972909866867e-06,
+      "loss": 0.6478,
+      "step": 10861
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.689569587139332,
+      "learning_rate": 1.5799185896981596e-06,
+      "loss": 0.7064,
+      "step": 10862
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.7000922590271332,
+      "learning_rate": 1.5798398829972004e-06,
+      "loss": 0.6556,
+      "step": 10863
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9423234281970452,
+      "learning_rate": 1.5797611708845447e-06,
+      "loss": 0.6615,
+      "step": 10864
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.25358674047133,
+      "learning_rate": 1.5796824533609265e-06,
+      "loss": 0.6732,
+      "step": 10865
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.506514262054172,
+      "learning_rate": 1.5796037304270809e-06,
+      "loss": 0.6934,
+      "step": 10866
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.4669980573473138,
+      "learning_rate": 1.5795250020837423e-06,
+      "loss": 0.6914,
+      "step": 10867
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.4573506142577413,
+      "learning_rate": 1.5794462683316456e-06,
+      "loss": 0.6699,
+      "step": 10868
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.797288241815941,
+      "learning_rate": 1.5793675291715256e-06,
+      "loss": 0.6432,
+      "step": 10869
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3393245995853789,
+      "learning_rate": 1.5792887846041173e-06,
+      "loss": 0.7018,
+      "step": 10870
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.905258121767591,
+      "learning_rate": 1.5792100346301555e-06,
+      "loss": 0.6745,
+      "step": 10871
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.2455463319774505,
+      "learning_rate": 1.5791312792503754e-06,
+      "loss": 0.6751,
+      "step": 10872
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.099197399590721,
+      "learning_rate": 1.5790525184655117e-06,
+      "loss": 0.6927,
+      "step": 10873
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.4544582753929207,
+      "learning_rate": 1.5789737522762997e-06,
+      "loss": 0.6673,
+      "step": 10874
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9706640604385935,
+      "learning_rate": 1.5788949806834743e-06,
+      "loss": 0.6901,
+      "step": 10875
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5160975865259367,
+      "learning_rate": 1.5788162036877709e-06,
+      "loss": 0.7135,
+      "step": 10876
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.69208507040437,
+      "learning_rate": 1.5787374212899247e-06,
+      "loss": 0.6921,
+      "step": 10877
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8855290135376344,
+      "learning_rate": 1.5786586334906711e-06,
+      "loss": 0.6823,
+      "step": 10878
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.679696473697665,
+      "learning_rate": 1.5785798402907453e-06,
+      "loss": 0.6693,
+      "step": 10879
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 8.204966584821817,
+      "learning_rate": 1.5785010416908823e-06,
+      "loss": 0.7025,
+      "step": 10880
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.136604135695541,
+      "learning_rate": 1.5784222376918183e-06,
+      "loss": 0.653,
+      "step": 10881
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.369216482913039,
+      "learning_rate": 1.5783434282942882e-06,
+      "loss": 0.6712,
+      "step": 10882
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.1298008047460024,
+      "learning_rate": 1.5782646134990278e-06,
+      "loss": 0.6738,
+      "step": 10883
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.816092444896434,
+      "learning_rate": 1.5781857933067731e-06,
+      "loss": 0.7012,
+      "step": 10884
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.444683204721409,
+      "learning_rate": 1.5781069677182585e-06,
+      "loss": 0.6842,
+      "step": 10885
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 6.632631276018449,
+      "learning_rate": 1.578028136734221e-06,
+      "loss": 0.7109,
+      "step": 10886
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 5.219793411889018,
+      "learning_rate": 1.5779493003553954e-06,
+      "loss": 0.6986,
+      "step": 10887
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7814650339840221,
+      "learning_rate": 1.577870458582518e-06,
+      "loss": 0.7025,
+      "step": 10888
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.220889433551564,
+      "learning_rate": 1.5777916114163247e-06,
+      "loss": 0.6908,
+      "step": 10889
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.5407129468413965,
+      "learning_rate": 1.577712758857551e-06,
+      "loss": 0.6908,
+      "step": 10890
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.090404316393218,
+      "learning_rate": 1.577633900906933e-06,
+      "loss": 0.6784,
+      "step": 10891
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8495097335816986,
+      "learning_rate": 1.5775550375652067e-06,
+      "loss": 0.6784,
+      "step": 10892
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8969914788994002,
+      "learning_rate": 1.5774761688331082e-06,
+      "loss": 0.6784,
+      "step": 10893
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.042289938802985,
+      "learning_rate": 1.577397294711374e-06,
+      "loss": 0.6478,
+      "step": 10894
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.872418807566418,
+      "learning_rate": 1.577318415200739e-06,
+      "loss": 0.6699,
+      "step": 10895
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6379590771404006,
+      "learning_rate": 1.5772395303019405e-06,
+      "loss": 0.6602,
+      "step": 10896
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.0122191409371633,
+      "learning_rate": 1.5771606400157146e-06,
+      "loss": 0.6836,
+      "step": 10897
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0859701688342407,
+      "learning_rate": 1.5770817443427974e-06,
+      "loss": 0.6706,
+      "step": 10898
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.405799645696073,
+      "learning_rate": 1.577002843283925e-06,
+      "loss": 0.696,
+      "step": 10899
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.5968189217224458,
+      "learning_rate": 1.576923936839834e-06,
+      "loss": 0.6549,
+      "step": 10900
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.303799362696809,
+      "learning_rate": 1.5768450250112611e-06,
+      "loss": 0.6908,
+      "step": 10901
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 8.031690518433326,
+      "learning_rate": 1.5767661077989423e-06,
+      "loss": 0.6888,
+      "step": 10902
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.111060782740673,
+      "learning_rate": 1.5766871852036147e-06,
+      "loss": 0.6732,
+      "step": 10903
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7046592465576236,
+      "learning_rate": 1.5766082572260145e-06,
+      "loss": 0.6751,
+      "step": 10904
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5682048330347854,
+      "learning_rate": 1.5765293238668783e-06,
+      "loss": 0.6745,
+      "step": 10905
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.196006997753249,
+      "learning_rate": 1.5764503851269431e-06,
+      "loss": 0.6536,
+      "step": 10906
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.4240489425377034,
+      "learning_rate": 1.5763714410069456e-06,
+      "loss": 0.6712,
+      "step": 10907
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7190492386183969,
+      "learning_rate": 1.5762924915076224e-06,
+      "loss": 0.6803,
+      "step": 10908
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.5589222745559774,
+      "learning_rate": 1.5762135366297101e-06,
+      "loss": 0.6836,
+      "step": 10909
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.753186264839048,
+      "learning_rate": 1.5761345763739463e-06,
+      "loss": 0.7142,
+      "step": 10910
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.950976156585306,
+      "learning_rate": 1.5760556107410674e-06,
+      "loss": 0.6764,
+      "step": 10911
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.8750083653396317,
+      "learning_rate": 1.5759766397318103e-06,
+      "loss": 0.6986,
+      "step": 10912
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7156017502585692,
+      "learning_rate": 1.5758976633469128e-06,
+      "loss": 0.6829,
+      "step": 10913
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.116291418460826,
+      "learning_rate": 1.575818681587111e-06,
+      "loss": 0.6699,
+      "step": 10914
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.034432795264781,
+      "learning_rate": 1.575739694453143e-06,
+      "loss": 0.6823,
+      "step": 10915
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6924086205399211,
+      "learning_rate": 1.5756607019457452e-06,
+      "loss": 0.6895,
+      "step": 10916
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.846144029242367,
+      "learning_rate": 1.5755817040656551e-06,
+      "loss": 0.6797,
+      "step": 10917
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.8709558181365127,
+      "learning_rate": 1.57550270081361e-06,
+      "loss": 0.6725,
+      "step": 10918
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9721618694161916,
+      "learning_rate": 1.5754236921903475e-06,
+      "loss": 0.6921,
+      "step": 10919
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6979067835663101,
+      "learning_rate": 1.5753446781966049e-06,
+      "loss": 0.6751,
+      "step": 10920
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.872306573131073,
+      "learning_rate": 1.5752656588331193e-06,
+      "loss": 0.6784,
+      "step": 10921
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6956777739750108,
+      "learning_rate": 1.5751866341006283e-06,
+      "loss": 0.6855,
+      "step": 10922
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1897887993478684,
+      "learning_rate": 1.5751076039998698e-06,
+      "loss": 0.6693,
+      "step": 10923
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.34467964753188,
+      "learning_rate": 1.5750285685315814e-06,
+      "loss": 0.6484,
+      "step": 10924
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7549339994270977,
+      "learning_rate": 1.5749495276964999e-06,
+      "loss": 0.6725,
+      "step": 10925
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.075183602347167,
+      "learning_rate": 1.5748704814953643e-06,
+      "loss": 0.6725,
+      "step": 10926
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.6118381754313562,
+      "learning_rate": 1.5747914299289113e-06,
+      "loss": 0.6706,
+      "step": 10927
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 7.348179859933883,
+      "learning_rate": 1.574712372997879e-06,
+      "loss": 0.6771,
+      "step": 10928
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.550081507486053,
+      "learning_rate": 1.5746333107030054e-06,
+      "loss": 0.666,
+      "step": 10929
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9218505298261529,
+      "learning_rate": 1.5745542430450285e-06,
+      "loss": 0.681,
+      "step": 10930
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9136566605186969,
+      "learning_rate": 1.5744751700246856e-06,
+      "loss": 0.6875,
+      "step": 10931
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8948483462206522,
+      "learning_rate": 1.5743960916427153e-06,
+      "loss": 0.6836,
+      "step": 10932
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0234674194939835,
+      "learning_rate": 1.5743170078998555e-06,
+      "loss": 0.6615,
+      "step": 10933
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9943207425774195,
+      "learning_rate": 1.5742379187968444e-06,
+      "loss": 0.6777,
+      "step": 10934
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.3740746814502571,
+      "learning_rate": 1.5741588243344199e-06,
+      "loss": 0.6693,
+      "step": 10935
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.0456323142121864,
+      "learning_rate": 1.5740797245133204e-06,
+      "loss": 0.6361,
+      "step": 10936
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.570121096311766,
+      "learning_rate": 1.574000619334284e-06,
+      "loss": 0.6868,
+      "step": 10937
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.109992810350273,
+      "learning_rate": 1.573921508798049e-06,
+      "loss": 0.6654,
+      "step": 10938
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.724521926961035,
+      "learning_rate": 1.5738423929053538e-06,
+      "loss": 0.6934,
+      "step": 10939
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 5.213900177645598,
+      "learning_rate": 1.573763271656937e-06,
+      "loss": 0.6868,
+      "step": 10940
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.814279246307477,
+      "learning_rate": 1.5736841450535368e-06,
+      "loss": 0.6758,
+      "step": 10941
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1725572902559482,
+      "learning_rate": 1.5736050130958915e-06,
+      "loss": 0.6426,
+      "step": 10942
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.974960599269594,
+      "learning_rate": 1.57352587578474e-06,
+      "loss": 0.696,
+      "step": 10943
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8639996485870586,
+      "learning_rate": 1.573446733120821e-06,
+      "loss": 0.6816,
+      "step": 10944
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.3657074449299011,
+      "learning_rate": 1.5733675851048727e-06,
+      "loss": 0.6621,
+      "step": 10945
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.1340741836927337,
+      "learning_rate": 1.5732884317376339e-06,
+      "loss": 0.6829,
+      "step": 10946
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.3840299963219964,
+      "learning_rate": 1.5732092730198438e-06,
+      "loss": 0.6686,
+      "step": 10947
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9149744869998353,
+      "learning_rate": 1.5731301089522408e-06,
+      "loss": 0.6771,
+      "step": 10948
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.3611126839719732,
+      "learning_rate": 1.5730509395355635e-06,
+      "loss": 0.6719,
+      "step": 10949
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0077093364365355,
+      "learning_rate": 1.5729717647705517e-06,
+      "loss": 0.6855,
+      "step": 10950
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5681658388419473,
+      "learning_rate": 1.572892584657943e-06,
+      "loss": 0.6934,
+      "step": 10951
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.230216131839852,
+      "learning_rate": 1.5728133991984776e-06,
+      "loss": 0.6706,
+      "step": 10952
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.5449229934003084,
+      "learning_rate": 1.5727342083928941e-06,
+      "loss": 0.6576,
+      "step": 10953
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 4.437802055893595,
+      "learning_rate": 1.5726550122419313e-06,
+      "loss": 0.6725,
+      "step": 10954
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.082945808799576,
+      "learning_rate": 1.5725758107463287e-06,
+      "loss": 0.6621,
+      "step": 10955
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9964887351024893,
+      "learning_rate": 1.5724966039068255e-06,
+      "loss": 0.6855,
+      "step": 10956
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.0159438346265537,
+      "learning_rate": 1.5724173917241611e-06,
+      "loss": 0.6882,
+      "step": 10957
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.1281041897269204,
+      "learning_rate": 1.5723381741990745e-06,
+      "loss": 0.6803,
+      "step": 10958
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.5770437752823474,
+      "learning_rate": 1.5722589513323046e-06,
+      "loss": 0.6823,
+      "step": 10959
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.9219254068289353,
+      "learning_rate": 1.572179723124592e-06,
+      "loss": 0.653,
+      "step": 10960
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7983515852505811,
+      "learning_rate": 1.572100489576675e-06,
+      "loss": 0.6784,
+      "step": 10961
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.831478458549315,
+      "learning_rate": 1.5720212506892935e-06,
+      "loss": 0.6842,
+      "step": 10962
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.528216391443293,
+      "learning_rate": 1.5719420064631871e-06,
+      "loss": 0.6855,
+      "step": 10963
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9589306300799871,
+      "learning_rate": 1.5718627568990955e-06,
+      "loss": 0.7051,
+      "step": 10964
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.2410105370776954,
+      "learning_rate": 1.5717835019977583e-06,
+      "loss": 0.6667,
+      "step": 10965
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.3701702737774113,
+      "learning_rate": 1.5717042417599151e-06,
+      "loss": 0.6576,
+      "step": 10966
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7117827525348709,
+      "learning_rate": 1.5716249761863056e-06,
+      "loss": 0.6784,
+      "step": 10967
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.541445167817156,
+      "learning_rate": 1.5715457052776695e-06,
+      "loss": 0.6576,
+      "step": 10968
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.128723950159252,
+      "learning_rate": 1.5714664290347468e-06,
+      "loss": 0.6842,
+      "step": 10969
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1238091725276853,
+      "learning_rate": 1.5713871474582776e-06,
+      "loss": 0.6634,
+      "step": 10970
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.927287928866984,
+      "learning_rate": 1.5713078605490015e-06,
+      "loss": 0.6836,
+      "step": 10971
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.0782073517569946,
+      "learning_rate": 1.5712285683076586e-06,
+      "loss": 0.7129,
+      "step": 10972
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.6506574136045127,
+      "learning_rate": 1.5711492707349892e-06,
+      "loss": 0.6712,
+      "step": 10973
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7899508058710776,
+      "learning_rate": 1.5710699678317327e-06,
+      "loss": 0.6836,
+      "step": 10974
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0889368380581284,
+      "learning_rate": 1.57099065959863e-06,
+      "loss": 0.6725,
+      "step": 10975
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.4821343769582394,
+      "learning_rate": 1.5709113460364209e-06,
+      "loss": 0.6465,
+      "step": 10976
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1628228811368406,
+      "learning_rate": 1.5708320271458458e-06,
+      "loss": 0.6849,
+      "step": 10977
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.376097379563005,
+      "learning_rate": 1.5707527029276447e-06,
+      "loss": 0.6641,
+      "step": 10978
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1831951499265478,
+      "learning_rate": 1.5706733733825583e-06,
+      "loss": 0.6921,
+      "step": 10979
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.4223063622582006,
+      "learning_rate": 1.5705940385113267e-06,
+      "loss": 0.6966,
+      "step": 10980
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1467797957680779,
+      "learning_rate": 1.5705146983146907e-06,
+      "loss": 0.6842,
+      "step": 10981
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.499514511518596,
+      "learning_rate": 1.5704353527933903e-06,
+      "loss": 0.6608,
+      "step": 10982
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.193462364197924,
+      "learning_rate": 1.5703560019481668e-06,
+      "loss": 0.6589,
+      "step": 10983
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0985943897379518,
+      "learning_rate": 1.5702766457797596e-06,
+      "loss": 0.6771,
+      "step": 10984
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.926762124295069,
+      "learning_rate": 1.5701972842889104e-06,
+      "loss": 0.6465,
+      "step": 10985
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.068431986437767,
+      "learning_rate": 1.5701179174763596e-06,
+      "loss": 0.6654,
+      "step": 10986
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.0228241407132734,
+      "learning_rate": 1.5700385453428478e-06,
+      "loss": 0.6504,
+      "step": 10987
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.7526396309636663,
+      "learning_rate": 1.5699591678891157e-06,
+      "loss": 0.6862,
+      "step": 10988
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8663143342100569,
+      "learning_rate": 1.5698797851159044e-06,
+      "loss": 0.6803,
+      "step": 10989
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8010232033255492,
+      "learning_rate": 1.5698003970239546e-06,
+      "loss": 0.6836,
+      "step": 10990
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5314561438062217,
+      "learning_rate": 1.5697210036140074e-06,
+      "loss": 0.6849,
+      "step": 10991
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.452700519343828,
+      "learning_rate": 1.5696416048868036e-06,
+      "loss": 0.6732,
+      "step": 10992
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.2005256212753086,
+      "learning_rate": 1.5695622008430846e-06,
+      "loss": 0.6439,
+      "step": 10993
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.836220200999536,
+      "learning_rate": 1.569482791483591e-06,
+      "loss": 0.6882,
+      "step": 10994
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9426487619067886,
+      "learning_rate": 1.5694033768090638e-06,
+      "loss": 0.638,
+      "step": 10995
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.853220524231923,
+      "learning_rate": 1.569323956820245e-06,
+      "loss": 0.6882,
+      "step": 10996
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5894680463171693,
+      "learning_rate": 1.5692445315178753e-06,
+      "loss": 0.6836,
+      "step": 10997
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.6641968795992494,
+      "learning_rate": 1.569165100902696e-06,
+      "loss": 0.6589,
+      "step": 10998
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.3024425999022622,
+      "learning_rate": 1.5690856649754482e-06,
+      "loss": 0.6823,
+      "step": 10999
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.4018028229974293,
+      "learning_rate": 1.569006223736874e-06,
+      "loss": 0.6797,
+      "step": 11000
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.3558012450356627,
+      "learning_rate": 1.568926777187714e-06,
+      "loss": 0.666,
+      "step": 11001
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5966894607211775,
+      "learning_rate": 1.5688473253287104e-06,
+      "loss": 0.6543,
+      "step": 11002
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.7342646754833875,
+      "learning_rate": 1.5687678681606043e-06,
+      "loss": 0.6549,
+      "step": 11003
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8996206334838268,
+      "learning_rate": 1.5686884056841372e-06,
+      "loss": 0.6712,
+      "step": 11004
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5978410243115855,
+      "learning_rate": 1.568608937900051e-06,
+      "loss": 0.6745,
+      "step": 11005
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9627013908844201,
+      "learning_rate": 1.5685294648090876e-06,
+      "loss": 0.6732,
+      "step": 11006
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.0233960594705365,
+      "learning_rate": 1.5684499864119883e-06,
+      "loss": 0.6829,
+      "step": 11007
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.60832674034636,
+      "learning_rate": 1.5683705027094947e-06,
+      "loss": 0.6868,
+      "step": 11008
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9133052609936653,
+      "learning_rate": 1.5682910137023493e-06,
+      "loss": 0.6471,
+      "step": 11009
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.929251407178253,
+      "learning_rate": 1.5682115193912935e-06,
+      "loss": 0.6458,
+      "step": 11010
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.920223959074569,
+      "learning_rate": 1.5681320197770692e-06,
+      "loss": 0.6465,
+      "step": 11011
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.2506515716931528,
+      "learning_rate": 1.5680525148604184e-06,
+      "loss": 0.6784,
+      "step": 11012
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.19978538271458,
+      "learning_rate": 1.5679730046420838e-06,
+      "loss": 0.6823,
+      "step": 11013
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1315186159684851,
+      "learning_rate": 1.5678934891228065e-06,
+      "loss": 0.6582,
+      "step": 11014
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.0583909406202303,
+      "learning_rate": 1.5678139683033292e-06,
+      "loss": 0.6712,
+      "step": 11015
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8821155846606155,
+      "learning_rate": 1.5677344421843936e-06,
+      "loss": 0.6549,
+      "step": 11016
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.2782099014572355,
+      "learning_rate": 1.5676549107667424e-06,
+      "loss": 0.6751,
+      "step": 11017
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8713202447661705,
+      "learning_rate": 1.5675753740511179e-06,
+      "loss": 0.7161,
+      "step": 11018
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.8669465406688337,
+      "learning_rate": 1.5674958320382623e-06,
+      "loss": 0.6634,
+      "step": 11019
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.071454471754388,
+      "learning_rate": 1.5674162847289176e-06,
+      "loss": 0.6751,
+      "step": 11020
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.651945493672988,
+      "learning_rate": 1.5673367321238267e-06,
+      "loss": 0.6738,
+      "step": 11021
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5139825774792492,
+      "learning_rate": 1.567257174223732e-06,
+      "loss": 0.6719,
+      "step": 11022
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1338159157343408,
+      "learning_rate": 1.5671776110293757e-06,
+      "loss": 0.6315,
+      "step": 11023
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.2665135610964597,
+      "learning_rate": 1.5670980425415006e-06,
+      "loss": 0.6901,
+      "step": 11024
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.326544767457685,
+      "learning_rate": 1.5670184687608497e-06,
+      "loss": 0.6966,
+      "step": 11025
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 4.856435482137852,
+      "learning_rate": 1.566938889688165e-06,
+      "loss": 0.6322,
+      "step": 11026
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.3812227492927738,
+      "learning_rate": 1.5668593053241894e-06,
+      "loss": 0.679,
+      "step": 11027
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.8116644450458828,
+      "learning_rate": 1.566779715669666e-06,
+      "loss": 0.6797,
+      "step": 11028
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.024558192933994,
+      "learning_rate": 1.5667001207253372e-06,
+      "loss": 0.6725,
+      "step": 11029
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7460470047728045,
+      "learning_rate": 1.5666205204919462e-06,
+      "loss": 0.6771,
+      "step": 11030
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1559757087039901,
+      "learning_rate": 1.5665409149702357e-06,
+      "loss": 0.6927,
+      "step": 11031
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.3104958745537196,
+      "learning_rate": 1.5664613041609487e-06,
+      "loss": 0.651,
+      "step": 11032
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.7116229562079615,
+      "learning_rate": 1.5663816880648283e-06,
+      "loss": 0.6764,
+      "step": 11033
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.0851606118462764,
+      "learning_rate": 1.5663020666826175e-06,
+      "loss": 0.6842,
+      "step": 11034
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.6236287283413167,
+      "learning_rate": 1.5662224400150596e-06,
+      "loss": 0.6855,
+      "step": 11035
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.752063287548609,
+      "learning_rate": 1.5661428080628975e-06,
+      "loss": 0.6823,
+      "step": 11036
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.842951190285467,
+      "learning_rate": 1.5660631708268744e-06,
+      "loss": 0.6829,
+      "step": 11037
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1116644534545757,
+      "learning_rate": 1.5659835283077335e-06,
+      "loss": 0.6875,
+      "step": 11038
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.251929075048706,
+      "learning_rate": 1.5659038805062189e-06,
+      "loss": 0.6719,
+      "step": 11039
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8644715149599564,
+      "learning_rate": 1.5658242274230728e-06,
+      "loss": 0.6777,
+      "step": 11040
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7924832400664795,
+      "learning_rate": 1.5657445690590391e-06,
+      "loss": 0.6751,
+      "step": 11041
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8258657749569108,
+      "learning_rate": 1.5656649054148617e-06,
+      "loss": 0.681,
+      "step": 11042
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.0103920740215293,
+      "learning_rate": 1.5655852364912835e-06,
+      "loss": 0.6712,
+      "step": 11043
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1591239295166853,
+      "learning_rate": 1.565505562289048e-06,
+      "loss": 0.6439,
+      "step": 11044
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.4546565091174037,
+      "learning_rate": 1.5654258828088995e-06,
+      "loss": 0.6543,
+      "step": 11045
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.2132332252496725,
+      "learning_rate": 1.5653461980515813e-06,
+      "loss": 0.6979,
+      "step": 11046
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8351220578941677,
+      "learning_rate": 1.5652665080178367e-06,
+      "loss": 0.6764,
+      "step": 11047
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.1169301677372152,
+      "learning_rate": 1.5651868127084097e-06,
+      "loss": 0.6829,
+      "step": 11048
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.499217816718475,
+      "learning_rate": 1.5651071121240444e-06,
+      "loss": 0.6406,
+      "step": 11049
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.3506902595446095,
+      "learning_rate": 1.5650274062654844e-06,
+      "loss": 0.666,
+      "step": 11050
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.063937252141609,
+      "learning_rate": 1.564947695133474e-06,
+      "loss": 0.6803,
+      "step": 11051
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9412156869288277,
+      "learning_rate": 1.5648679787287562e-06,
+      "loss": 0.6875,
+      "step": 11052
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.5726772251929617,
+      "learning_rate": 1.5647882570520761e-06,
+      "loss": 0.6868,
+      "step": 11053
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.2362289612781736,
+      "learning_rate": 1.5647085301041768e-06,
+      "loss": 0.653,
+      "step": 11054
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.311084409902412,
+      "learning_rate": 1.5646287978858033e-06,
+      "loss": 0.6497,
+      "step": 11055
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.150765599392005,
+      "learning_rate": 1.5645490603976992e-06,
+      "loss": 0.668,
+      "step": 11056
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7884654060186447,
+      "learning_rate": 1.5644693176406085e-06,
+      "loss": 0.6562,
+      "step": 11057
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9591915050447417,
+      "learning_rate": 1.564389569615276e-06,
+      "loss": 0.6576,
+      "step": 11058
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.6783593336631135,
+      "learning_rate": 1.5643098163224458e-06,
+      "loss": 0.6693,
+      "step": 11059
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8203033351723013,
+      "learning_rate": 1.564230057762862e-06,
+      "loss": 0.6426,
+      "step": 11060
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9620683386555,
+      "learning_rate": 1.564150293937269e-06,
+      "loss": 0.6693,
+      "step": 11061
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5890499889393723,
+      "learning_rate": 1.5640705248464116e-06,
+      "loss": 0.6751,
+      "step": 11062
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.7196768432680012,
+      "learning_rate": 1.5639907504910344e-06,
+      "loss": 0.6686,
+      "step": 11063
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.1058862697909686,
+      "learning_rate": 1.5639109708718815e-06,
+      "loss": 0.6758,
+      "step": 11064
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.427161859140091,
+      "learning_rate": 1.5638311859896973e-06,
+      "loss": 0.6712,
+      "step": 11065
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.9360148694600396,
+      "learning_rate": 1.563751395845227e-06,
+      "loss": 0.6771,
+      "step": 11066
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.4477488741804418,
+      "learning_rate": 1.563671600439215e-06,
+      "loss": 0.6738,
+      "step": 11067
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.861333825873844,
+      "learning_rate": 1.5635917997724061e-06,
+      "loss": 0.6823,
+      "step": 11068
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.536272174530692,
+      "learning_rate": 1.5635119938455455e-06,
+      "loss": 0.6706,
+      "step": 11069
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.203602998783429,
+      "learning_rate": 1.563432182659377e-06,
+      "loss": 0.7018,
+      "step": 11070
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 4.077625181998982,
+      "learning_rate": 1.5633523662146466e-06,
+      "loss": 0.6523,
+      "step": 11071
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.131325894859596,
+      "learning_rate": 1.5632725445120984e-06,
+      "loss": 0.6823,
+      "step": 11072
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.45626780470299,
+      "learning_rate": 1.563192717552478e-06,
+      "loss": 0.7083,
+      "step": 11073
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.1524919513253162,
+      "learning_rate": 1.5631128853365302e-06,
+      "loss": 0.6491,
+      "step": 11074
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8078167857482328,
+      "learning_rate": 1.5630330478649996e-06,
+      "loss": 0.6654,
+      "step": 11075
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.8282487770443367,
+      "learning_rate": 1.562953205138632e-06,
+      "loss": 0.679,
+      "step": 11076
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.023022217354098,
+      "learning_rate": 1.5628733571581725e-06,
+      "loss": 0.6667,
+      "step": 11077
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.2012741142241046,
+      "learning_rate": 1.5627935039243661e-06,
+      "loss": 0.6465,
+      "step": 11078
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.030772911437229,
+      "learning_rate": 1.562713645437958e-06,
+      "loss": 0.6426,
+      "step": 11079
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.6750047031520667,
+      "learning_rate": 1.562633781699694e-06,
+      "loss": 0.6751,
+      "step": 11080
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 5.41012661519084,
+      "learning_rate": 1.5625539127103188e-06,
+      "loss": 0.7064,
+      "step": 11081
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.3003711784676077,
+      "learning_rate": 1.5624740384705786e-06,
+      "loss": 0.6589,
+      "step": 11082
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.932268940614518,
+      "learning_rate": 1.5623941589812183e-06,
+      "loss": 0.666,
+      "step": 11083
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.65945318973393,
+      "learning_rate": 1.5623142742429834e-06,
+      "loss": 0.6536,
+      "step": 11084
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.3785038805379024,
+      "learning_rate": 1.5622343842566197e-06,
+      "loss": 0.6895,
+      "step": 11085
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9878834860819865,
+      "learning_rate": 1.5621544890228729e-06,
+      "loss": 0.709,
+      "step": 11086
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.004694455330313,
+      "learning_rate": 1.5620745885424884e-06,
+      "loss": 0.681,
+      "step": 11087
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.732089184788964,
+      "learning_rate": 1.561994682816212e-06,
+      "loss": 0.6699,
+      "step": 11088
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.8114382488991914,
+      "learning_rate": 1.5619147718447895e-06,
+      "loss": 0.6816,
+      "step": 11089
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.0372131310317352,
+      "learning_rate": 1.5618348556289672e-06,
+      "loss": 0.7109,
+      "step": 11090
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6027335244257528,
+      "learning_rate": 1.56175493416949e-06,
+      "loss": 0.6745,
+      "step": 11091
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 6.275199910327199,
+      "learning_rate": 1.5616750074671049e-06,
+      "loss": 0.6771,
+      "step": 11092
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9261426125492687,
+      "learning_rate": 1.5615950755225567e-06,
+      "loss": 0.681,
+      "step": 11093
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8316102335710438,
+      "learning_rate": 1.561515138336592e-06,
+      "loss": 0.6686,
+      "step": 11094
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9435841722912126,
+      "learning_rate": 1.561435195909957e-06,
+      "loss": 0.6582,
+      "step": 11095
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.9223064363168515,
+      "learning_rate": 1.561355248243398e-06,
+      "loss": 0.6719,
+      "step": 11096
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.122132726388421,
+      "learning_rate": 1.5612752953376604e-06,
+      "loss": 0.6849,
+      "step": 11097
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.001133518578907,
+      "learning_rate": 1.561195337193491e-06,
+      "loss": 0.6504,
+      "step": 11098
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4183217726393211,
+      "learning_rate": 1.5611153738116358e-06,
+      "loss": 0.6764,
+      "step": 11099
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.7425669600752007,
+      "learning_rate": 1.5610354051928414e-06,
+      "loss": 0.6602,
+      "step": 11100
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.3608260319266887,
+      "learning_rate": 1.5609554313378537e-06,
+      "loss": 0.6686,
+      "step": 11101
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8470761079870077,
+      "learning_rate": 1.5608754522474194e-06,
+      "loss": 0.6517,
+      "step": 11102
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.5083174480901356,
+      "learning_rate": 1.560795467922285e-06,
+      "loss": 0.6777,
+      "step": 11103
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.71007261260778,
+      "learning_rate": 1.560715478363197e-06,
+      "loss": 0.6849,
+      "step": 11104
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.751906016717194,
+      "learning_rate": 1.5606354835709017e-06,
+      "loss": 0.6621,
+      "step": 11105
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8791067113128842,
+      "learning_rate": 1.5605554835461456e-06,
+      "loss": 0.6719,
+      "step": 11106
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.9857240836415744,
+      "learning_rate": 1.560475478289676e-06,
+      "loss": 0.6842,
+      "step": 11107
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.597611651840696,
+      "learning_rate": 1.5603954678022388e-06,
+      "loss": 0.7194,
+      "step": 11108
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.063099189850777,
+      "learning_rate": 1.5603154520845818e-06,
+      "loss": 0.6999,
+      "step": 11109
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.7325004768469654,
+      "learning_rate": 1.5602354311374506e-06,
+      "loss": 0.6901,
+      "step": 11110
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.911838305781167,
+      "learning_rate": 1.5601554049615926e-06,
+      "loss": 0.6758,
+      "step": 11111
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.9741657679860791,
+      "learning_rate": 1.5600753735577547e-06,
+      "loss": 0.666,
+      "step": 11112
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0125759982496674,
+      "learning_rate": 1.559995336926684e-06,
+      "loss": 0.6927,
+      "step": 11113
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8795404938851586,
+      "learning_rate": 1.5599152950691272e-06,
+      "loss": 0.6732,
+      "step": 11114
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.2697395108479121,
+      "learning_rate": 1.5598352479858315e-06,
+      "loss": 0.6855,
+      "step": 11115
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8334039664760695,
+      "learning_rate": 1.5597551956775439e-06,
+      "loss": 0.6732,
+      "step": 11116
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6602132150451303,
+      "learning_rate": 1.5596751381450113e-06,
+      "loss": 0.6771,
+      "step": 11117
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5852376725644413,
+      "learning_rate": 1.5595950753889813e-06,
+      "loss": 0.6445,
+      "step": 11118
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.321383805495227,
+      "learning_rate": 1.559515007410201e-06,
+      "loss": 0.6797,
+      "step": 11119
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.5302405894117355,
+      "learning_rate": 1.5594349342094178e-06,
+      "loss": 0.7057,
+      "step": 11120
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.4270004202918782,
+      "learning_rate": 1.5593548557873786e-06,
+      "loss": 0.6719,
+      "step": 11121
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.068045237997582,
+      "learning_rate": 1.5592747721448314e-06,
+      "loss": 0.6569,
+      "step": 11122
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.063520420142707,
+      "learning_rate": 1.5591946832825229e-06,
+      "loss": 0.6908,
+      "step": 11123
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.438264524722734,
+      "learning_rate": 1.5591145892012011e-06,
+      "loss": 0.6589,
+      "step": 11124
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.618236437279899,
+      "learning_rate": 1.5590344899016137e-06,
+      "loss": 0.6654,
+      "step": 11125
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.779222525978523,
+      "learning_rate": 1.5589543853845077e-06,
+      "loss": 0.6751,
+      "step": 11126
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7173139365475074,
+      "learning_rate": 1.558874275650631e-06,
+      "loss": 0.6842,
+      "step": 11127
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.2776254548913242,
+      "learning_rate": 1.558794160700731e-06,
+      "loss": 0.696,
+      "step": 11128
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.905946180288359,
+      "learning_rate": 1.5587140405355564e-06,
+      "loss": 0.6947,
+      "step": 11129
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.353843384014234,
+      "learning_rate": 1.5586339151558537e-06,
+      "loss": 0.6986,
+      "step": 11130
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.3058526087639761,
+      "learning_rate": 1.5585537845623713e-06,
+      "loss": 0.6764,
+      "step": 11131
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.9818970223889079,
+      "learning_rate": 1.5584736487558572e-06,
+      "loss": 0.681,
+      "step": 11132
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8782354708294189,
+      "learning_rate": 1.558393507737059e-06,
+      "loss": 0.6719,
+      "step": 11133
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.3926594740751774,
+      "learning_rate": 1.5583133615067248e-06,
+      "loss": 0.6699,
+      "step": 11134
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.9831800708714202,
+      "learning_rate": 1.5582332100656029e-06,
+      "loss": 0.6836,
+      "step": 11135
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.242828900256094,
+      "learning_rate": 1.5581530534144408e-06,
+      "loss": 0.6699,
+      "step": 11136
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0960550586108269,
+      "learning_rate": 1.5580728915539866e-06,
+      "loss": 0.6738,
+      "step": 11137
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9043122303698935,
+      "learning_rate": 1.5579927244849893e-06,
+      "loss": 0.6803,
+      "step": 11138
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.9757104192032737,
+      "learning_rate": 1.5579125522081964e-06,
+      "loss": 0.6706,
+      "step": 11139
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7337610447817694,
+      "learning_rate": 1.5578323747243562e-06,
+      "loss": 0.679,
+      "step": 11140
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.599755047666309,
+      "learning_rate": 1.5577521920342171e-06,
+      "loss": 0.6868,
+      "step": 11141
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 6.631720114669484,
+      "learning_rate": 1.5576720041385278e-06,
+      "loss": 0.6882,
+      "step": 11142
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.658903085481322,
+      "learning_rate": 1.5575918110380362e-06,
+      "loss": 0.7103,
+      "step": 11143
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5919183143270976,
+      "learning_rate": 1.5575116127334908e-06,
+      "loss": 0.6745,
+      "step": 11144
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.511230867470934,
+      "learning_rate": 1.5574314092256403e-06,
+      "loss": 0.7031,
+      "step": 11145
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.855088150747014,
+      "learning_rate": 1.5573512005152331e-06,
+      "loss": 0.651,
+      "step": 11146
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 5.226299711028827,
+      "learning_rate": 1.5572709866030179e-06,
+      "loss": 0.6947,
+      "step": 11147
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.3924223996856537,
+      "learning_rate": 1.5571907674897434e-06,
+      "loss": 0.6673,
+      "step": 11148
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.0814797437997457,
+      "learning_rate": 1.557110543176158e-06,
+      "loss": 0.6719,
+      "step": 11149
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7429360919593815,
+      "learning_rate": 1.5570303136630108e-06,
+      "loss": 0.681,
+      "step": 11150
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 4.2572656932113135,
+      "learning_rate": 1.5569500789510501e-06,
+      "loss": 0.6973,
+      "step": 11151
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.41948355950848,
+      "learning_rate": 1.5568698390410252e-06,
+      "loss": 0.6862,
+      "step": 11152
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.178215831361386,
+      "learning_rate": 1.556789593933685e-06,
+      "loss": 0.6751,
+      "step": 11153
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7355337428854247,
+      "learning_rate": 1.5567093436297783e-06,
+      "loss": 0.6608,
+      "step": 11154
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.667750129575242,
+      "learning_rate": 1.556629088130054e-06,
+      "loss": 0.651,
+      "step": 11155
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.4081301076731987,
+      "learning_rate": 1.556548827435261e-06,
+      "loss": 0.6712,
+      "step": 11156
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.454432380138162,
+      "learning_rate": 1.5564685615461489e-06,
+      "loss": 0.6934,
+      "step": 11157
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.8797490240989851,
+      "learning_rate": 1.5563882904634665e-06,
+      "loss": 0.6771,
+      "step": 11158
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.1305676976726942,
+      "learning_rate": 1.5563080141879628e-06,
+      "loss": 0.6628,
+      "step": 11159
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7389754162693332,
+      "learning_rate": 1.556227732720387e-06,
+      "loss": 0.6758,
+      "step": 11160
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7448466918886252,
+      "learning_rate": 1.5561474460614886e-06,
+      "loss": 0.6875,
+      "step": 11161
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.9028245922372903,
+      "learning_rate": 1.5560671542120173e-06,
+      "loss": 0.6647,
+      "step": 11162
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.1372294506350304,
+      "learning_rate": 1.555986857172722e-06,
+      "loss": 0.6823,
+      "step": 11163
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.949842386813308,
+      "learning_rate": 1.555906554944352e-06,
+      "loss": 0.6927,
+      "step": 11164
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.756256500394645,
+      "learning_rate": 1.555826247527657e-06,
+      "loss": 0.6953,
+      "step": 11165
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.5817706477439955,
+      "learning_rate": 1.5557459349233866e-06,
+      "loss": 0.6745,
+      "step": 11166
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7544729902383263,
+      "learning_rate": 1.5556656171322901e-06,
+      "loss": 0.6738,
+      "step": 11167
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.106231058219863,
+      "learning_rate": 1.5555852941551174e-06,
+      "loss": 0.6966,
+      "step": 11168
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.389920412067286,
+      "learning_rate": 1.5555049659926179e-06,
+      "loss": 0.6465,
+      "step": 11169
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.4299782045115614,
+      "learning_rate": 1.5554246326455416e-06,
+      "loss": 0.6686,
+      "step": 11170
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.4409682833028752,
+      "learning_rate": 1.5553442941146378e-06,
+      "loss": 0.6777,
+      "step": 11171
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.6825286652887876,
+      "learning_rate": 1.5552639504006568e-06,
+      "loss": 0.7031,
+      "step": 11172
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.777780014755392,
+      "learning_rate": 1.555183601504348e-06,
+      "loss": 0.6921,
+      "step": 11173
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.1906987549718724,
+      "learning_rate": 1.5551032474264618e-06,
+      "loss": 0.7233,
+      "step": 11174
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.2706835592684076,
+      "learning_rate": 1.5550228881677478e-06,
+      "loss": 0.6647,
+      "step": 11175
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3454484119802808,
+      "learning_rate": 1.5549425237289563e-06,
+      "loss": 0.6719,
+      "step": 11176
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.362782424238008,
+      "learning_rate": 1.5548621541108369e-06,
+      "loss": 0.7025,
+      "step": 11177
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.00587245419975,
+      "learning_rate": 1.55478177931414e-06,
+      "loss": 0.6602,
+      "step": 11178
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.723942776585052,
+      "learning_rate": 1.5547013993396155e-06,
+      "loss": 0.6816,
+      "step": 11179
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.73219934688831,
+      "learning_rate": 1.5546210141880143e-06,
+      "loss": 0.7012,
+      "step": 11180
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.2882157101118974,
+      "learning_rate": 1.554540623860086e-06,
+      "loss": 0.6745,
+      "step": 11181
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9918882762294894,
+      "learning_rate": 1.5544602283565806e-06,
+      "loss": 0.6966,
+      "step": 11182
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3017436079918494,
+      "learning_rate": 1.5543798276782492e-06,
+      "loss": 0.6647,
+      "step": 11183
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7594423413902487,
+      "learning_rate": 1.5542994218258417e-06,
+      "loss": 0.6868,
+      "step": 11184
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6346779758366456,
+      "learning_rate": 1.5542190108001088e-06,
+      "loss": 0.6895,
+      "step": 11185
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.1158435064111574,
+      "learning_rate": 1.5541385946018007e-06,
+      "loss": 0.668,
+      "step": 11186
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6937310033872162,
+      "learning_rate": 1.554058173231668e-06,
+      "loss": 0.6458,
+      "step": 11187
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8162248047886786,
+      "learning_rate": 1.5539777466904615e-06,
+      "loss": 0.6615,
+      "step": 11188
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.703276509718176,
+      "learning_rate": 1.553897314978932e-06,
+      "loss": 0.681,
+      "step": 11189
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.14217425879038,
+      "learning_rate": 1.5538168780978294e-06,
+      "loss": 0.6654,
+      "step": 11190
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0104738564360718,
+      "learning_rate": 1.5537364360479053e-06,
+      "loss": 0.666,
+      "step": 11191
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3280552532481134,
+      "learning_rate": 1.5536559888299098e-06,
+      "loss": 0.6829,
+      "step": 11192
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.746705666943857,
+      "learning_rate": 1.5535755364445939e-06,
+      "loss": 0.6882,
+      "step": 11193
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.783906904236911,
+      "learning_rate": 1.5534950788927087e-06,
+      "loss": 0.6589,
+      "step": 11194
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8240558385780643,
+      "learning_rate": 1.5534146161750052e-06,
+      "loss": 0.6602,
+      "step": 11195
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.37758349891066,
+      "learning_rate": 1.553334148292234e-06,
+      "loss": 0.6764,
+      "step": 11196
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.1454286089043038,
+      "learning_rate": 1.5532536752451462e-06,
+      "loss": 0.6816,
+      "step": 11197
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.498592492754459,
+      "learning_rate": 1.553173197034493e-06,
+      "loss": 0.6771,
+      "step": 11198
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7825582758505125,
+      "learning_rate": 1.5530927136610253e-06,
+      "loss": 0.6576,
+      "step": 11199
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.701342383112746,
+      "learning_rate": 1.5530122251254945e-06,
+      "loss": 0.6595,
+      "step": 11200
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3304031332093138,
+      "learning_rate": 1.5529317314286518e-06,
+      "loss": 0.6549,
+      "step": 11201
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3615885620381958,
+      "learning_rate": 1.5528512325712482e-06,
+      "loss": 0.6582,
+      "step": 11202
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.5262101662590513,
+      "learning_rate": 1.552770728554035e-06,
+      "loss": 0.6589,
+      "step": 11203
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.662180176417295,
+      "learning_rate": 1.552690219377764e-06,
+      "loss": 0.6699,
+      "step": 11204
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.0303152162735767,
+      "learning_rate": 1.5526097050431863e-06,
+      "loss": 0.6745,
+      "step": 11205
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.9759255556346815,
+      "learning_rate": 1.552529185551053e-06,
+      "loss": 0.666,
+      "step": 11206
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9499883212624459,
+      "learning_rate": 1.5524486609021162e-06,
+      "loss": 0.6875,
+      "step": 11207
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3732557426224352,
+      "learning_rate": 1.5523681310971272e-06,
+      "loss": 0.6517,
+      "step": 11208
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.448777726036567,
+      "learning_rate": 1.5522875961368378e-06,
+      "loss": 0.6615,
+      "step": 11209
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.635503730021206,
+      "learning_rate": 1.552207056021999e-06,
+      "loss": 0.6484,
+      "step": 11210
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5501810754757017,
+      "learning_rate": 1.5521265107533631e-06,
+      "loss": 0.7005,
+      "step": 11211
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6524542531999384,
+      "learning_rate": 1.5520459603316818e-06,
+      "loss": 0.6654,
+      "step": 11212
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.020323785288818,
+      "learning_rate": 1.5519654047577063e-06,
+      "loss": 0.6836,
+      "step": 11213
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 4.642929297029692,
+      "learning_rate": 1.5518848440321892e-06,
+      "loss": 0.707,
+      "step": 11214
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.9316985462653227,
+      "learning_rate": 1.551804278155882e-06,
+      "loss": 0.6589,
+      "step": 11215
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9713724280667181,
+      "learning_rate": 1.5517237071295366e-06,
+      "loss": 0.6829,
+      "step": 11216
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.212491612108119,
+      "learning_rate": 1.5516431309539053e-06,
+      "loss": 0.6803,
+      "step": 11217
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.03688897214889,
+      "learning_rate": 1.5515625496297394e-06,
+      "loss": 0.6387,
+      "step": 11218
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.937279063129707,
+      "learning_rate": 1.551481963157792e-06,
+      "loss": 0.6803,
+      "step": 11219
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.4043822217593456,
+      "learning_rate": 1.5514013715388143e-06,
+      "loss": 0.6849,
+      "step": 11220
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.8440437585067654,
+      "learning_rate": 1.551320774773559e-06,
+      "loss": 0.6562,
+      "step": 11221
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.5157679037488527,
+      "learning_rate": 1.5512401728627781e-06,
+      "loss": 0.6654,
+      "step": 11222
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.6129741990851256,
+      "learning_rate": 1.551159565807224e-06,
+      "loss": 0.6953,
+      "step": 11223
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.788784880054805,
+      "learning_rate": 1.5510789536076488e-06,
+      "loss": 0.6908,
+      "step": 11224
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8622546749617819,
+      "learning_rate": 1.5509983362648051e-06,
+      "loss": 0.6634,
+      "step": 11225
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.9265252550635537,
+      "learning_rate": 1.5509177137794448e-06,
+      "loss": 0.6595,
+      "step": 11226
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.3029899052927767,
+      "learning_rate": 1.5508370861523212e-06,
+      "loss": 0.6562,
+      "step": 11227
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8110218676719474,
+      "learning_rate": 1.5507564533841863e-06,
+      "loss": 0.6452,
+      "step": 11228
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1633586318780214,
+      "learning_rate": 1.5506758154757927e-06,
+      "loss": 0.6895,
+      "step": 11229
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.626131361225816,
+      "learning_rate": 1.5505951724278928e-06,
+      "loss": 0.6895,
+      "step": 11230
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.2987824628603146,
+      "learning_rate": 1.5505145242412398e-06,
+      "loss": 0.6712,
+      "step": 11231
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.8406325353072934,
+      "learning_rate": 1.5504338709165858e-06,
+      "loss": 0.6549,
+      "step": 11232
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.1239900192931511,
+      "learning_rate": 1.5503532124546838e-06,
+      "loss": 0.6927,
+      "step": 11233
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.464375720220149,
+      "learning_rate": 1.550272548856287e-06,
+      "loss": 0.6797,
+      "step": 11234
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.891191271639314,
+      "learning_rate": 1.5501918801221474e-06,
+      "loss": 0.6986,
+      "step": 11235
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8546637103594538,
+      "learning_rate": 1.5501112062530185e-06,
+      "loss": 0.6849,
+      "step": 11236
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8654736632383196,
+      "learning_rate": 1.5500305272496529e-06,
+      "loss": 0.6751,
+      "step": 11237
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.7104699710423326,
+      "learning_rate": 1.5499498431128037e-06,
+      "loss": 0.6634,
+      "step": 11238
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8786368048872732,
+      "learning_rate": 1.5498691538432243e-06,
+      "loss": 0.7018,
+      "step": 11239
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.112706948718815,
+      "learning_rate": 1.5497884594416673e-06,
+      "loss": 0.6458,
+      "step": 11240
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0653847210376266,
+      "learning_rate": 1.549707759908886e-06,
+      "loss": 0.6602,
+      "step": 11241
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.137347084059323,
+      "learning_rate": 1.5496270552456336e-06,
+      "loss": 0.6536,
+      "step": 11242
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4124486886880492,
+      "learning_rate": 1.5495463454526631e-06,
+      "loss": 0.6693,
+      "step": 11243
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4009008931408218,
+      "learning_rate": 1.5494656305307284e-06,
+      "loss": 0.6862,
+      "step": 11244
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.835543916638712,
+      "learning_rate": 1.5493849104805818e-06,
+      "loss": 0.681,
+      "step": 11245
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8796240089325402,
+      "learning_rate": 1.5493041853029775e-06,
+      "loss": 0.6699,
+      "step": 11246
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0021590954882085,
+      "learning_rate": 1.5492234549986689e-06,
+      "loss": 0.6335,
+      "step": 11247
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.0454330948967248,
+      "learning_rate": 1.549142719568409e-06,
+      "loss": 0.6842,
+      "step": 11248
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.194481220383077,
+      "learning_rate": 1.5490619790129515e-06,
+      "loss": 0.6628,
+      "step": 11249
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.316169570737417,
+      "learning_rate": 1.54898123333305e-06,
+      "loss": 0.6784,
+      "step": 11250
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.653225764014253,
+      "learning_rate": 1.5489004825294582e-06,
+      "loss": 0.6908,
+      "step": 11251
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4713555062178114,
+      "learning_rate": 1.5488197266029298e-06,
+      "loss": 0.6862,
+      "step": 11252
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9123517240286612,
+      "learning_rate": 1.548738965554218e-06,
+      "loss": 0.6628,
+      "step": 11253
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.090250049411171,
+      "learning_rate": 1.5486581993840771e-06,
+      "loss": 0.6719,
+      "step": 11254
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.030720473826189,
+      "learning_rate": 1.5485774280932606e-06,
+      "loss": 0.7018,
+      "step": 11255
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.9374253206031111,
+      "learning_rate": 1.5484966516825223e-06,
+      "loss": 0.6654,
+      "step": 11256
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.8237772383640707,
+      "learning_rate": 1.5484158701526167e-06,
+      "loss": 0.653,
+      "step": 11257
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.064423066856631,
+      "learning_rate": 1.548335083504297e-06,
+      "loss": 0.6686,
+      "step": 11258
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 5.419814444524726,
+      "learning_rate": 1.5482542917383175e-06,
+      "loss": 0.6816,
+      "step": 11259
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.2705233544584664,
+      "learning_rate": 1.548173494855432e-06,
+      "loss": 0.6712,
+      "step": 11260
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.812481355520745,
+      "learning_rate": 1.5480926928563954e-06,
+      "loss": 0.6582,
+      "step": 11261
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.383451544551159,
+      "learning_rate": 1.5480118857419607e-06,
+      "loss": 0.6634,
+      "step": 11262
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.660456058432206,
+      "learning_rate": 1.5479310735128825e-06,
+      "loss": 0.6934,
+      "step": 11263
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.1764403639587622,
+      "learning_rate": 1.5478502561699157e-06,
+      "loss": 0.6686,
+      "step": 11264
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.134016113491482,
+      "learning_rate": 1.5477694337138135e-06,
+      "loss": 0.6667,
+      "step": 11265
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.6609323690856286,
+      "learning_rate": 1.547688606145331e-06,
+      "loss": 0.6882,
+      "step": 11266
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4684936803424862,
+      "learning_rate": 1.5476077734652222e-06,
+      "loss": 0.6667,
+      "step": 11267
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.405350591833407,
+      "learning_rate": 1.5475269356742416e-06,
+      "loss": 0.6803,
+      "step": 11268
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.5469746117383931,
+      "learning_rate": 1.5474460927731436e-06,
+      "loss": 0.6842,
+      "step": 11269
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0408798723061385,
+      "learning_rate": 1.547365244762683e-06,
+      "loss": 0.6445,
+      "step": 11270
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.715647558182814,
+      "learning_rate": 1.5472843916436142e-06,
+      "loss": 0.6543,
+      "step": 11271
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.0091174632463797,
+      "learning_rate": 1.5472035334166917e-06,
+      "loss": 0.6771,
+      "step": 11272
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.155576590543522,
+      "learning_rate": 1.54712267008267e-06,
+      "loss": 0.6243,
+      "step": 11273
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4293189664559534,
+      "learning_rate": 1.5470418016423045e-06,
+      "loss": 0.6615,
+      "step": 11274
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.186449272996043,
+      "learning_rate": 1.5469609280963495e-06,
+      "loss": 0.6621,
+      "step": 11275
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.837489970902029,
+      "learning_rate": 1.5468800494455591e-06,
+      "loss": 0.6432,
+      "step": 11276
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.9032470181912642,
+      "learning_rate": 1.5467991656906895e-06,
+      "loss": 0.6491,
+      "step": 11277
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0418803347339793,
+      "learning_rate": 1.5467182768324946e-06,
+      "loss": 0.6914,
+      "step": 11278
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.6548792851534657,
+      "learning_rate": 1.54663738287173e-06,
+      "loss": 0.6882,
+      "step": 11279
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0452618571509031,
+      "learning_rate": 1.5465564838091498e-06,
+      "loss": 0.6719,
+      "step": 11280
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.951539210697477,
+      "learning_rate": 1.54647557964551e-06,
+      "loss": 0.6621,
+      "step": 11281
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.704095187061839,
+      "learning_rate": 1.5463946703815651e-06,
+      "loss": 0.6471,
+      "step": 11282
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.0746511354889523,
+      "learning_rate": 1.5463137560180709e-06,
+      "loss": 0.7201,
+      "step": 11283
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.5797007847621907,
+      "learning_rate": 1.5462328365557815e-06,
+      "loss": 0.6953,
+      "step": 11284
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.2300206725101535,
+      "learning_rate": 1.5461519119954529e-06,
+      "loss": 0.653,
+      "step": 11285
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.771209644424251,
+      "learning_rate": 1.5460709823378402e-06,
+      "loss": 0.6641,
+      "step": 11286
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.579381017045527,
+      "learning_rate": 1.545990047583699e-06,
+      "loss": 0.6556,
+      "step": 11287
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.9760420434121126,
+      "learning_rate": 1.545909107733784e-06,
+      "loss": 0.6725,
+      "step": 11288
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.3684206125210907,
+      "learning_rate": 1.5458281627888515e-06,
+      "loss": 0.6686,
+      "step": 11289
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9310820835639162,
+      "learning_rate": 1.545747212749656e-06,
+      "loss": 0.6374,
+      "step": 11290
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.352642451559995,
+      "learning_rate": 1.5456662576169538e-06,
+      "loss": 0.6556,
+      "step": 11291
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.08857080724838,
+      "learning_rate": 1.5455852973915001e-06,
+      "loss": 0.7103,
+      "step": 11292
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 4.582657216868768,
+      "learning_rate": 1.5455043320740505e-06,
+      "loss": 0.6706,
+      "step": 11293
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 8.034694436120494,
+      "learning_rate": 1.5454233616653607e-06,
+      "loss": 0.6803,
+      "step": 11294
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.2327009067842445,
+      "learning_rate": 1.5453423861661867e-06,
+      "loss": 0.6777,
+      "step": 11295
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.705485474650443,
+      "learning_rate": 1.5452614055772838e-06,
+      "loss": 0.6777,
+      "step": 11296
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.0794755691117914,
+      "learning_rate": 1.5451804198994078e-06,
+      "loss": 0.6947,
+      "step": 11297
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8915000800003741,
+      "learning_rate": 1.5450994291333151e-06,
+      "loss": 0.6667,
+      "step": 11298
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.983015185724108,
+      "learning_rate": 1.5450184332797607e-06,
+      "loss": 0.6777,
+      "step": 11299
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4982249777581058,
+      "learning_rate": 1.5449374323395015e-06,
+      "loss": 0.653,
+      "step": 11300
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.963645389433203,
+      "learning_rate": 1.5448564263132929e-06,
+      "loss": 0.6882,
+      "step": 11301
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8765478923855543,
+      "learning_rate": 1.544775415201891e-06,
+      "loss": 0.679,
+      "step": 11302
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.471109141613477,
+      "learning_rate": 1.5446943990060522e-06,
+      "loss": 0.6667,
+      "step": 11303
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8044445212580242,
+      "learning_rate": 1.5446133777265323e-06,
+      "loss": 0.666,
+      "step": 11304
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5037424880235595,
+      "learning_rate": 1.5445323513640875e-06,
+      "loss": 0.6654,
+      "step": 11305
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.7794801218487497,
+      "learning_rate": 1.5444513199194743e-06,
+      "loss": 0.6654,
+      "step": 11306
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.997109070315956,
+      "learning_rate": 1.5443702833934486e-06,
+      "loss": 0.694,
+      "step": 11307
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.1509118118828296,
+      "learning_rate": 1.544289241786767e-06,
+      "loss": 0.6673,
+      "step": 11308
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.459004880230745,
+      "learning_rate": 1.5442081951001855e-06,
+      "loss": 0.6465,
+      "step": 11309
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.327210729484986,
+      "learning_rate": 1.5441271433344612e-06,
+      "loss": 0.6576,
+      "step": 11310
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9584109291822471,
+      "learning_rate": 1.5440460864903502e-06,
+      "loss": 0.6686,
+      "step": 11311
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.9037343653377965,
+      "learning_rate": 1.5439650245686085e-06,
+      "loss": 0.679,
+      "step": 11312
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7785853177309929,
+      "learning_rate": 1.5438839575699933e-06,
+      "loss": 0.666,
+      "step": 11313
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9987285827825527,
+      "learning_rate": 1.5438028854952614e-06,
+      "loss": 0.681,
+      "step": 11314
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8963111520126065,
+      "learning_rate": 1.5437218083451687e-06,
+      "loss": 0.7083,
+      "step": 11315
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.407704659950976,
+      "learning_rate": 1.543640726120472e-06,
+      "loss": 0.6706,
+      "step": 11316
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7155554033062366,
+      "learning_rate": 1.5435596388219288e-06,
+      "loss": 0.6699,
+      "step": 11317
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 8.501391891220823,
+      "learning_rate": 1.5434785464502952e-06,
+      "loss": 0.7012,
+      "step": 11318
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.7899643588268295,
+      "learning_rate": 1.5433974490063287e-06,
+      "loss": 0.6602,
+      "step": 11319
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9800574858201209,
+      "learning_rate": 1.5433163464907852e-06,
+      "loss": 0.6367,
+      "step": 11320
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.365003059808672,
+      "learning_rate": 1.5432352389044226e-06,
+      "loss": 0.6328,
+      "step": 11321
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7933114555221286,
+      "learning_rate": 1.543154126247997e-06,
+      "loss": 0.6504,
+      "step": 11322
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8509903964906169,
+      "learning_rate": 1.5430730085222661e-06,
+      "loss": 0.651,
+      "step": 11323
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.423108666081131,
+      "learning_rate": 1.5429918857279871e-06,
+      "loss": 0.6829,
+      "step": 11324
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.991520536736925,
+      "learning_rate": 1.5429107578659163e-06,
+      "loss": 0.6699,
+      "step": 11325
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.27064474980368,
+      "learning_rate": 1.5428296249368118e-06,
+      "loss": 0.6777,
+      "step": 11326
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 5.692779716752442,
+      "learning_rate": 1.5427484869414303e-06,
+      "loss": 0.6322,
+      "step": 11327
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.9598256756506396,
+      "learning_rate": 1.5426673438805292e-06,
+      "loss": 0.6621,
+      "step": 11328
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.341847383437105,
+      "learning_rate": 1.5425861957548657e-06,
+      "loss": 0.6738,
+      "step": 11329
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.053845612643833,
+      "learning_rate": 1.5425050425651973e-06,
+      "loss": 0.6914,
+      "step": 11330
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.6859556606721897,
+      "learning_rate": 1.5424238843122814e-06,
+      "loss": 0.696,
+      "step": 11331
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.062907077288447,
+      "learning_rate": 1.5423427209968752e-06,
+      "loss": 0.6647,
+      "step": 11332
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.0310133684033636,
+      "learning_rate": 1.5422615526197366e-06,
+      "loss": 0.6947,
+      "step": 11333
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.8450975845588578,
+      "learning_rate": 1.542180379181623e-06,
+      "loss": 0.7233,
+      "step": 11334
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3725480455418841,
+      "learning_rate": 1.5420992006832918e-06,
+      "loss": 0.6953,
+      "step": 11335
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.0808836374354696,
+      "learning_rate": 1.5420180171255011e-06,
+      "loss": 0.6491,
+      "step": 11336
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.2749635370840353,
+      "learning_rate": 1.5419368285090081e-06,
+      "loss": 0.6868,
+      "step": 11337
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.627223133676176,
+      "learning_rate": 1.5418556348345706e-06,
+      "loss": 0.6354,
+      "step": 11338
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.0373001990058945,
+      "learning_rate": 1.5417744361029467e-06,
+      "loss": 0.6999,
+      "step": 11339
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4505359282290686,
+      "learning_rate": 1.5416932323148942e-06,
+      "loss": 0.6751,
+      "step": 11340
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9851505041718704,
+      "learning_rate": 1.5416120234711708e-06,
+      "loss": 0.6803,
+      "step": 11341
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6432473005771593,
+      "learning_rate": 1.5415308095725346e-06,
+      "loss": 0.6641,
+      "step": 11342
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 6.012054917768955,
+      "learning_rate": 1.5414495906197434e-06,
+      "loss": 0.6634,
+      "step": 11343
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.100511809320942,
+      "learning_rate": 1.5413683666135553e-06,
+      "loss": 0.7266,
+      "step": 11344
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9018241901249615,
+      "learning_rate": 1.541287137554728e-06,
+      "loss": 0.6569,
+      "step": 11345
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.743705175797526,
+      "learning_rate": 1.5412059034440203e-06,
+      "loss": 0.6927,
+      "step": 11346
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.6704043730612945,
+      "learning_rate": 1.54112466428219e-06,
+      "loss": 0.6595,
+      "step": 11347
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.167721030653473,
+      "learning_rate": 1.5410434200699953e-06,
+      "loss": 0.681,
+      "step": 11348
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4847142391894248,
+      "learning_rate": 1.5409621708081945e-06,
+      "loss": 0.6732,
+      "step": 11349
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.275888891366519,
+      "learning_rate": 1.5408809164975463e-06,
+      "loss": 0.7031,
+      "step": 11350
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.2470156651361965,
+      "learning_rate": 1.5407996571388082e-06,
+      "loss": 0.6947,
+      "step": 11351
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 3.027309122590078,
+      "learning_rate": 1.5407183927327391e-06,
+      "loss": 0.6849,
+      "step": 11352
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.0230874292019843,
+      "learning_rate": 1.5406371232800977e-06,
+      "loss": 0.6895,
+      "step": 11353
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 4.053723298801417,
+      "learning_rate": 1.540555848781642e-06,
+      "loss": 0.6549,
+      "step": 11354
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.0390140678852415,
+      "learning_rate": 1.540474569238131e-06,
+      "loss": 0.6523,
+      "step": 11355
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.458888974358896,
+      "learning_rate": 1.5403932846503228e-06,
+      "loss": 0.6458,
+      "step": 11356
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.9803766979722295,
+      "learning_rate": 1.5403119950189764e-06,
+      "loss": 0.6667,
+      "step": 11357
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.1218547660589415,
+      "learning_rate": 1.5402307003448503e-06,
+      "loss": 0.6882,
+      "step": 11358
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9706750253407144,
+      "learning_rate": 1.5401494006287031e-06,
+      "loss": 0.651,
+      "step": 11359
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8447784819645549,
+      "learning_rate": 1.5400680958712942e-06,
+      "loss": 0.7083,
+      "step": 11360
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.1199126640918595,
+      "learning_rate": 1.5399867860733816e-06,
+      "loss": 0.7018,
+      "step": 11361
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.608948963219059,
+      "learning_rate": 1.539905471235725e-06,
+      "loss": 0.6615,
+      "step": 11362
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8491034409152949,
+      "learning_rate": 1.5398241513590826e-06,
+      "loss": 0.6497,
+      "step": 11363
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.7816283651743097,
+      "learning_rate": 1.5397428264442137e-06,
+      "loss": 0.6576,
+      "step": 11364
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0451897200132,
+      "learning_rate": 1.5396614964918773e-06,
+      "loss": 0.6797,
+      "step": 11365
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.777576196186719,
+      "learning_rate": 1.5395801615028324e-06,
+      "loss": 0.6979,
+      "step": 11366
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0677974404762325,
+      "learning_rate": 1.539498821477838e-06,
+      "loss": 0.681,
+      "step": 11367
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9618170680778946,
+      "learning_rate": 1.5394174764176534e-06,
+      "loss": 0.6504,
+      "step": 11368
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.045711365334792,
+      "learning_rate": 1.5393361263230382e-06,
+      "loss": 0.6797,
+      "step": 11369
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9837123093844516,
+      "learning_rate": 1.5392547711947513e-06,
+      "loss": 0.6784,
+      "step": 11370
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.3680770992880302,
+      "learning_rate": 1.5391734110335512e-06,
+      "loss": 0.6673,
+      "step": 11371
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.1151117104447144,
+      "learning_rate": 1.5390920458401983e-06,
+      "loss": 0.6849,
+      "step": 11372
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8590358070470037,
+      "learning_rate": 1.5390106756154519e-06,
+      "loss": 0.6621,
+      "step": 11373
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 7.816570726211608,
+      "learning_rate": 1.538929300360071e-06,
+      "loss": 0.681,
+      "step": 11374
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0850259394114408,
+      "learning_rate": 1.5388479200748154e-06,
+      "loss": 0.6758,
+      "step": 11375
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.859913015470543,
+      "learning_rate": 1.5387665347604443e-06,
+      "loss": 0.694,
+      "step": 11376
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1495377174470636,
+      "learning_rate": 1.5386851444177176e-06,
+      "loss": 0.6628,
+      "step": 11377
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.475884730737466,
+      "learning_rate": 1.5386037490473948e-06,
+      "loss": 0.6895,
+      "step": 11378
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.375312653559208,
+      "learning_rate": 1.5385223486502354e-06,
+      "loss": 0.6829,
+      "step": 11379
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7994531102219278,
+      "learning_rate": 1.5384409432269999e-06,
+      "loss": 0.6842,
+      "step": 11380
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.101460006577832,
+      "learning_rate": 1.538359532778447e-06,
+      "loss": 0.6829,
+      "step": 11381
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.714637833671125,
+      "learning_rate": 1.5382781173053368e-06,
+      "loss": 0.6908,
+      "step": 11382
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8691100593281973,
+      "learning_rate": 1.5381966968084297e-06,
+      "loss": 0.6523,
+      "step": 11383
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6583209246782091,
+      "learning_rate": 1.5381152712884848e-06,
+      "loss": 0.6921,
+      "step": 11384
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.3845344866483744,
+      "learning_rate": 1.538033840746263e-06,
+      "loss": 0.6693,
+      "step": 11385
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.2188796210627664,
+      "learning_rate": 1.5379524051825237e-06,
+      "loss": 0.681,
+      "step": 11386
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.6247618764041345,
+      "learning_rate": 1.5378709645980265e-06,
+      "loss": 0.6836,
+      "step": 11387
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.4276062640750147,
+      "learning_rate": 1.5377895189935326e-06,
+      "loss": 0.6855,
+      "step": 11388
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.9485673693746617,
+      "learning_rate": 1.5377080683698012e-06,
+      "loss": 0.6556,
+      "step": 11389
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.4578961681074065,
+      "learning_rate": 1.537626612727593e-06,
+      "loss": 0.6784,
+      "step": 11390
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1018381711886223,
+      "learning_rate": 1.5375451520676684e-06,
+      "loss": 0.6751,
+      "step": 11391
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.046423694722265,
+      "learning_rate": 1.5374636863907866e-06,
+      "loss": 0.6712,
+      "step": 11392
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.3650593066452625,
+      "learning_rate": 1.5373822156977091e-06,
+      "loss": 0.6562,
+      "step": 11393
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.584154407424906,
+      "learning_rate": 1.5373007399891961e-06,
+      "loss": 0.6758,
+      "step": 11394
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6148091661615163,
+      "learning_rate": 1.5372192592660075e-06,
+      "loss": 0.6478,
+      "step": 11395
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.1684253503152457,
+      "learning_rate": 1.5371377735289043e-06,
+      "loss": 0.679,
+      "step": 11396
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 6.398820365292819,
+      "learning_rate": 1.5370562827786464e-06,
+      "loss": 0.6895,
+      "step": 11397
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.629063397751374,
+      "learning_rate": 1.5369747870159947e-06,
+      "loss": 0.6777,
+      "step": 11398
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.1961799893775464,
+      "learning_rate": 1.5368932862417098e-06,
+      "loss": 0.6725,
+      "step": 11399
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.3610088717152777,
+      "learning_rate": 1.5368117804565529e-06,
+      "loss": 0.64,
+      "step": 11400
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.845246557267634,
+      "learning_rate": 1.5367302696612838e-06,
+      "loss": 0.7005,
+      "step": 11401
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.3114434862107434,
+      "learning_rate": 1.5366487538566637e-06,
+      "loss": 0.6699,
+      "step": 11402
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.282611504453395,
+      "learning_rate": 1.536567233043453e-06,
+      "loss": 0.6784,
+      "step": 11403
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9623115339863134,
+      "learning_rate": 1.5364857072224137e-06,
+      "loss": 0.6445,
+      "step": 11404
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7684628828455622,
+      "learning_rate": 1.536404176394305e-06,
+      "loss": 0.6504,
+      "step": 11405
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1958553625007986,
+      "learning_rate": 1.536322640559889e-06,
+      "loss": 0.6973,
+      "step": 11406
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.0117156692592495,
+      "learning_rate": 1.5362410997199265e-06,
+      "loss": 0.6784,
+      "step": 11407
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.027850448747667,
+      "learning_rate": 1.536159553875178e-06,
+      "loss": 0.6497,
+      "step": 11408
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.1466018135579863,
+      "learning_rate": 1.5360780030264055e-06,
+      "loss": 0.7096,
+      "step": 11409
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.3859707061338744,
+      "learning_rate": 1.5359964471743695e-06,
+      "loss": 0.6712,
+      "step": 11410
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9900256676547536,
+      "learning_rate": 1.5359148863198307e-06,
+      "loss": 0.6758,
+      "step": 11411
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.964517994342586,
+      "learning_rate": 1.5358333204635514e-06,
+      "loss": 0.6667,
+      "step": 11412
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 4.104979377334326,
+      "learning_rate": 1.5357517496062922e-06,
+      "loss": 0.6764,
+      "step": 11413
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9854846979464653,
+      "learning_rate": 1.5356701737488145e-06,
+      "loss": 0.6641,
+      "step": 11414
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.8268068699629156,
+      "learning_rate": 1.5355885928918796e-06,
+      "loss": 0.651,
+      "step": 11415
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9398670033601974,
+      "learning_rate": 1.5355070070362493e-06,
+      "loss": 0.6803,
+      "step": 11416
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1154322753364858,
+      "learning_rate": 1.5354254161826845e-06,
+      "loss": 0.6706,
+      "step": 11417
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.387410120027395,
+      "learning_rate": 1.5353438203319468e-06,
+      "loss": 0.6738,
+      "step": 11418
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0118329552413687,
+      "learning_rate": 1.5352622194847982e-06,
+      "loss": 0.7116,
+      "step": 11419
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0187445241798279,
+      "learning_rate": 1.5351806136419997e-06,
+      "loss": 0.6647,
+      "step": 11420
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.081193874472834,
+      "learning_rate": 1.5350990028043134e-06,
+      "loss": 0.6868,
+      "step": 11421
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.873703436376526,
+      "learning_rate": 1.5350173869725009e-06,
+      "loss": 0.6641,
+      "step": 11422
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5410507606135453,
+      "learning_rate": 1.5349357661473236e-06,
+      "loss": 0.6641,
+      "step": 11423
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9048672490073183,
+      "learning_rate": 1.5348541403295437e-06,
+      "loss": 0.653,
+      "step": 11424
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.547601303010298,
+      "learning_rate": 1.5347725095199224e-06,
+      "loss": 0.6654,
+      "step": 11425
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.147600974256876,
+      "learning_rate": 1.5346908737192224e-06,
+      "loss": 0.6543,
+      "step": 11426
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9398433574644719,
+      "learning_rate": 1.5346092329282047e-06,
+      "loss": 0.6699,
+      "step": 11427
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 3.318675853424503,
+      "learning_rate": 1.5345275871476324e-06,
+      "loss": 0.7057,
+      "step": 11428
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1124358153428304,
+      "learning_rate": 1.5344459363782666e-06,
+      "loss": 0.7044,
+      "step": 11429
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9602810504173785,
+      "learning_rate": 1.5343642806208691e-06,
+      "loss": 0.6654,
+      "step": 11430
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.471980470201261,
+      "learning_rate": 1.5342826198762033e-06,
+      "loss": 0.6667,
+      "step": 11431
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 5.579599053598115,
+      "learning_rate": 1.5342009541450298e-06,
+      "loss": 0.7116,
+      "step": 11432
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3407061896450891,
+      "learning_rate": 1.5341192834281125e-06,
+      "loss": 0.6517,
+      "step": 11433
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1369377868853032,
+      "learning_rate": 1.534037607726212e-06,
+      "loss": 0.6745,
+      "step": 11434
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1504474686840287,
+      "learning_rate": 1.533955927040091e-06,
+      "loss": 0.6504,
+      "step": 11435
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8387564375102324,
+      "learning_rate": 1.533874241370513e-06,
+      "loss": 0.6523,
+      "step": 11436
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.6561067959161653,
+      "learning_rate": 1.533792550718239e-06,
+      "loss": 0.6882,
+      "step": 11437
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.6609975220956306,
+      "learning_rate": 1.533710855084032e-06,
+      "loss": 0.6719,
+      "step": 11438
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3840268381794052,
+      "learning_rate": 1.5336291544686541e-06,
+      "loss": 0.6549,
+      "step": 11439
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.311509529553836,
+      "learning_rate": 1.5335474488728685e-06,
+      "loss": 0.6953,
+      "step": 11440
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.680169732680658,
+      "learning_rate": 1.5334657382974371e-06,
+      "loss": 0.6667,
+      "step": 11441
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.064664933553885,
+      "learning_rate": 1.5333840227431228e-06,
+      "loss": 0.6836,
+      "step": 11442
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.185613249524422,
+      "learning_rate": 1.5333023022106883e-06,
+      "loss": 0.6882,
+      "step": 11443
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.331456772144496,
+      "learning_rate": 1.5332205767008966e-06,
+      "loss": 0.6484,
+      "step": 11444
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.66111508651182,
+      "learning_rate": 1.5331388462145097e-06,
+      "loss": 0.6934,
+      "step": 11445
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.474631327227673,
+      "learning_rate": 1.533057110752291e-06,
+      "loss": 0.6823,
+      "step": 11446
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.863875252857407,
+      "learning_rate": 1.532975370315003e-06,
+      "loss": 0.6686,
+      "step": 11447
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.3024271047808993,
+      "learning_rate": 1.5328936249034086e-06,
+      "loss": 0.6615,
+      "step": 11448
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7123612633406033,
+      "learning_rate": 1.5328118745182713e-06,
+      "loss": 0.6901,
+      "step": 11449
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.0424551542609777,
+      "learning_rate": 1.5327301191603533e-06,
+      "loss": 0.7038,
+      "step": 11450
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.047243753424626,
+      "learning_rate": 1.5326483588304182e-06,
+      "loss": 0.6543,
+      "step": 11451
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.975807550474671,
+      "learning_rate": 1.532566593529229e-06,
+      "loss": 0.6686,
+      "step": 11452
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.002418416130879,
+      "learning_rate": 1.5324848232575482e-06,
+      "loss": 0.6602,
+      "step": 11453
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.767661190035643,
+      "learning_rate": 1.5324030480161398e-06,
+      "loss": 0.6654,
+      "step": 11454
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.366206635343584,
+      "learning_rate": 1.5323212678057664e-06,
+      "loss": 0.6556,
+      "step": 11455
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9974021481801986,
+      "learning_rate": 1.5322394826271919e-06,
+      "loss": 0.6868,
+      "step": 11456
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8884514117753523,
+      "learning_rate": 1.532157692481179e-06,
+      "loss": 0.6465,
+      "step": 11457
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.6687308908623497,
+      "learning_rate": 1.5320758973684913e-06,
+      "loss": 0.6738,
+      "step": 11458
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.1997620189593705,
+      "learning_rate": 1.5319940972898921e-06,
+      "loss": 0.6458,
+      "step": 11459
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.1964648417588277,
+      "learning_rate": 1.5319122922461452e-06,
+      "loss": 0.6582,
+      "step": 11460
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8669426614462568,
+      "learning_rate": 1.5318304822380136e-06,
+      "loss": 0.6862,
+      "step": 11461
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.3256166192769836,
+      "learning_rate": 1.5317486672662615e-06,
+      "loss": 0.6803,
+      "step": 11462
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.88182124372709,
+      "learning_rate": 1.5316668473316517e-06,
+      "loss": 0.7207,
+      "step": 11463
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.334631295017903,
+      "learning_rate": 1.531585022434948e-06,
+      "loss": 0.7018,
+      "step": 11464
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7270526357154794,
+      "learning_rate": 1.5315031925769146e-06,
+      "loss": 0.6953,
+      "step": 11465
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3488509783022633,
+      "learning_rate": 1.5314213577583149e-06,
+      "loss": 0.6543,
+      "step": 11466
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.4768040178398525,
+      "learning_rate": 1.5313395179799127e-06,
+      "loss": 0.6543,
+      "step": 11467
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.776314009714664,
+      "learning_rate": 1.5312576732424716e-06,
+      "loss": 0.6667,
+      "step": 11468
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.5206484193664638,
+      "learning_rate": 1.5311758235467559e-06,
+      "loss": 0.6966,
+      "step": 11469
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.2287876009149676,
+      "learning_rate": 1.5310939688935289e-06,
+      "loss": 0.651,
+      "step": 11470
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8771289508932841,
+      "learning_rate": 1.531012109283555e-06,
+      "loss": 0.6536,
+      "step": 11471
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.0468882364446648,
+      "learning_rate": 1.5309302447175983e-06,
+      "loss": 0.681,
+      "step": 11472
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2621108883406331,
+      "learning_rate": 1.5308483751964228e-06,
+      "loss": 0.6589,
+      "step": 11473
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8433338204304525,
+      "learning_rate": 1.5307665007207923e-06,
+      "loss": 0.6621,
+      "step": 11474
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.9047871971496346,
+      "learning_rate": 1.530684621291471e-06,
+      "loss": 0.6764,
+      "step": 11475
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.720679134319679,
+      "learning_rate": 1.5306027369092234e-06,
+      "loss": 0.6849,
+      "step": 11476
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9221373226732609,
+      "learning_rate": 1.5305208475748132e-06,
+      "loss": 0.6732,
+      "step": 11477
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9602985082524109,
+      "learning_rate": 1.5304389532890055e-06,
+      "loss": 0.6419,
+      "step": 11478
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.4092210094630935,
+      "learning_rate": 1.5303570540525637e-06,
+      "loss": 0.666,
+      "step": 11479
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.327330519181849,
+      "learning_rate": 1.5302751498662527e-06,
+      "loss": 0.696,
+      "step": 11480
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2026757597840179,
+      "learning_rate": 1.5301932407308372e-06,
+      "loss": 0.6966,
+      "step": 11481
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.3098703438694783,
+      "learning_rate": 1.530111326647081e-06,
+      "loss": 0.6667,
+      "step": 11482
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.58466421352664,
+      "learning_rate": 1.530029407615749e-06,
+      "loss": 0.6719,
+      "step": 11483
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8999739990618183,
+      "learning_rate": 1.5299474836376055e-06,
+      "loss": 0.6595,
+      "step": 11484
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.173676847950758,
+      "learning_rate": 1.5298655547134151e-06,
+      "loss": 0.6647,
+      "step": 11485
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 5.012409366197969,
+      "learning_rate": 1.5297836208439432e-06,
+      "loss": 0.6543,
+      "step": 11486
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.4202006969302725,
+      "learning_rate": 1.5297016820299535e-06,
+      "loss": 0.668,
+      "step": 11487
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.4555468700866956,
+      "learning_rate": 1.5296197382722111e-06,
+      "loss": 0.6784,
+      "step": 11488
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8455076334877527,
+      "learning_rate": 1.5295377895714812e-06,
+      "loss": 0.6634,
+      "step": 11489
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.297455136615023,
+      "learning_rate": 1.529455835928528e-06,
+      "loss": 0.6914,
+      "step": 11490
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.8148464926304158,
+      "learning_rate": 1.5293738773441168e-06,
+      "loss": 0.6628,
+      "step": 11491
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2359938150402179,
+      "learning_rate": 1.5292919138190124e-06,
+      "loss": 0.6706,
+      "step": 11492
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.9154114390176344,
+      "learning_rate": 1.5292099453539796e-06,
+      "loss": 0.6758,
+      "step": 11493
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.258893337458056,
+      "learning_rate": 1.5291279719497836e-06,
+      "loss": 0.6992,
+      "step": 11494
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 6.559363678137722,
+      "learning_rate": 1.5290459936071898e-06,
+      "loss": 0.707,
+      "step": 11495
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 4.643891220486227,
+      "learning_rate": 1.5289640103269623e-06,
+      "loss": 0.6562,
+      "step": 11496
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.8736400985051713,
+      "learning_rate": 1.5288820221098675e-06,
+      "loss": 0.679,
+      "step": 11497
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.1140829373384817,
+      "learning_rate": 1.5288000289566697e-06,
+      "loss": 0.7012,
+      "step": 11498
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2873171559170815,
+      "learning_rate": 1.5287180308681347e-06,
+      "loss": 0.6719,
+      "step": 11499
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.225033352239221,
+      "learning_rate": 1.5286360278450272e-06,
+      "loss": 0.6764,
+      "step": 11500
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.787640481425196,
+      "learning_rate": 1.528554019888113e-06,
+      "loss": 0.6966,
+      "step": 11501
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.6840679425058362,
+      "learning_rate": 1.5284720069981578e-06,
+      "loss": 0.6641,
+      "step": 11502
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.843423212809562,
+      "learning_rate": 1.5283899891759263e-06,
+      "loss": 0.6836,
+      "step": 11503
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9723516461622862,
+      "learning_rate": 1.5283079664221842e-06,
+      "loss": 0.6953,
+      "step": 11504
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.1809509352633007,
+      "learning_rate": 1.5282259387376974e-06,
+      "loss": 0.6934,
+      "step": 11505
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.4997349985195285,
+      "learning_rate": 1.528143906123231e-06,
+      "loss": 0.6628,
+      "step": 11506
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.8413255828883,
+      "learning_rate": 1.528061868579551e-06,
+      "loss": 0.6478,
+      "step": 11507
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.2035092397153395,
+      "learning_rate": 1.5279798261074225e-06,
+      "loss": 0.6803,
+      "step": 11508
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 5.721390392327863,
+      "learning_rate": 1.527897778707612e-06,
+      "loss": 0.6829,
+      "step": 11509
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.159561171716866,
+      "learning_rate": 1.5278157263808846e-06,
+      "loss": 0.6927,
+      "step": 11510
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8184733457015813,
+      "learning_rate": 1.5277336691280064e-06,
+      "loss": 0.6751,
+      "step": 11511
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.264140928457679,
+      "learning_rate": 1.5276516069497435e-06,
+      "loss": 0.6647,
+      "step": 11512
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6940433394726021,
+      "learning_rate": 1.5275695398468611e-06,
+      "loss": 0.6862,
+      "step": 11513
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7766680109416464,
+      "learning_rate": 1.5274874678201254e-06,
+      "loss": 0.6725,
+      "step": 11514
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.3594121081005044,
+      "learning_rate": 1.5274053908703033e-06,
+      "loss": 0.6647,
+      "step": 11515
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.0532721629855162,
+      "learning_rate": 1.5273233089981593e-06,
+      "loss": 0.6602,
+      "step": 11516
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.5887161415626203,
+      "learning_rate": 1.5272412222044606e-06,
+      "loss": 0.6758,
+      "step": 11517
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.0288319090190476,
+      "learning_rate": 1.5271591304899727e-06,
+      "loss": 0.668,
+      "step": 11518
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.8389809756854498,
+      "learning_rate": 1.5270770338554622e-06,
+      "loss": 0.6725,
+      "step": 11519
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.993579845984833,
+      "learning_rate": 1.5269949323016951e-06,
+      "loss": 0.6908,
+      "step": 11520
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.6534945297771086,
+      "learning_rate": 1.5269128258294375e-06,
+      "loss": 0.6673,
+      "step": 11521
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8851601505627456,
+      "learning_rate": 1.526830714439456e-06,
+      "loss": 0.6634,
+      "step": 11522
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.6165733959729487,
+      "learning_rate": 1.5267485981325169e-06,
+      "loss": 0.6712,
+      "step": 11523
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.2407716553316825,
+      "learning_rate": 1.5266664769093865e-06,
+      "loss": 0.6836,
+      "step": 11524
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1359782295825271,
+      "learning_rate": 1.5265843507708314e-06,
+      "loss": 0.6562,
+      "step": 11525
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.823461378386351,
+      "learning_rate": 1.5265022197176178e-06,
+      "loss": 0.6647,
+      "step": 11526
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.795326432441512,
+      "learning_rate": 1.5264200837505125e-06,
+      "loss": 0.6595,
+      "step": 11527
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.7243942192831616,
+      "learning_rate": 1.5263379428702822e-06,
+      "loss": 0.6908,
+      "step": 11528
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.7178726440038345,
+      "learning_rate": 1.526255797077693e-06,
+      "loss": 0.6654,
+      "step": 11529
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.640359819172765,
+      "learning_rate": 1.5261736463735119e-06,
+      "loss": 0.6901,
+      "step": 11530
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.46441937152636,
+      "learning_rate": 1.526091490758506e-06,
+      "loss": 0.6426,
+      "step": 11531
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9625234293609433,
+      "learning_rate": 1.5260093302334415e-06,
+      "loss": 0.6953,
+      "step": 11532
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1522603940980256,
+      "learning_rate": 1.5259271647990857e-06,
+      "loss": 0.6712,
+      "step": 11533
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.825528634606967,
+      "learning_rate": 1.5258449944562046e-06,
+      "loss": 0.6771,
+      "step": 11534
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.0755662892487736,
+      "learning_rate": 1.5257628192055658e-06,
+      "loss": 0.6556,
+      "step": 11535
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9721265607543156,
+      "learning_rate": 1.5256806390479364e-06,
+      "loss": 0.6576,
+      "step": 11536
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.234082042987946,
+      "learning_rate": 1.5255984539840827e-06,
+      "loss": 0.6751,
+      "step": 11537
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.405182931164534,
+      "learning_rate": 1.5255162640147726e-06,
+      "loss": 0.6823,
+      "step": 11538
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.55610003390821,
+      "learning_rate": 1.5254340691407725e-06,
+      "loss": 0.6855,
+      "step": 11539
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.0863282999185495,
+      "learning_rate": 1.5253518693628497e-06,
+      "loss": 0.7025,
+      "step": 11540
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.225132543648698,
+      "learning_rate": 1.5252696646817716e-06,
+      "loss": 0.6569,
+      "step": 11541
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.0596821418595606,
+      "learning_rate": 1.5251874550983053e-06,
+      "loss": 0.6895,
+      "step": 11542
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7173364952028267,
+      "learning_rate": 1.5251052406132177e-06,
+      "loss": 0.6758,
+      "step": 11543
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.0601139788425298,
+      "learning_rate": 1.5250230212272767e-06,
+      "loss": 0.679,
+      "step": 11544
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.578498645671937,
+      "learning_rate": 1.5249407969412499e-06,
+      "loss": 0.6732,
+      "step": 11545
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.5884692739635544,
+      "learning_rate": 1.5248585677559032e-06,
+      "loss": 0.6875,
+      "step": 11546
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9211365974546788,
+      "learning_rate": 1.5247763336720057e-06,
+      "loss": 0.6836,
+      "step": 11547
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8705697555393813,
+      "learning_rate": 1.5246940946903242e-06,
+      "loss": 0.6725,
+      "step": 11548
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.001256513189764,
+      "learning_rate": 1.5246118508116263e-06,
+      "loss": 0.6562,
+      "step": 11549
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.2517210659199414,
+      "learning_rate": 1.5245296020366795e-06,
+      "loss": 0.6921,
+      "step": 11550
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.1652624977340456,
+      "learning_rate": 1.5244473483662517e-06,
+      "loss": 0.6686,
+      "step": 11551
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.0617956631762056,
+      "learning_rate": 1.5243650898011105e-06,
+      "loss": 0.6647,
+      "step": 11552
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.5138729086952383,
+      "learning_rate": 1.5242828263420231e-06,
+      "loss": 0.6738,
+      "step": 11553
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9763973219088,
+      "learning_rate": 1.5242005579897582e-06,
+      "loss": 0.6667,
+      "step": 11554
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7950439098109021,
+      "learning_rate": 1.5241182847450828e-06,
+      "loss": 0.6751,
+      "step": 11555
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.5545326066505156,
+      "learning_rate": 1.5240360066087652e-06,
+      "loss": 0.6543,
+      "step": 11556
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.712131246247518,
+      "learning_rate": 1.5239537235815733e-06,
+      "loss": 0.6823,
+      "step": 11557
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8436324167312812,
+      "learning_rate": 1.523871435664275e-06,
+      "loss": 0.679,
+      "step": 11558
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.547067308639291,
+      "learning_rate": 1.523789142857638e-06,
+      "loss": 0.6927,
+      "step": 11559
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 6.282559267879227,
+      "learning_rate": 1.523706845162431e-06,
+      "loss": 0.6953,
+      "step": 11560
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.8268317708739295,
+      "learning_rate": 1.5236245425794211e-06,
+      "loss": 0.6888,
+      "step": 11561
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.834019569154629,
+      "learning_rate": 1.5235422351093775e-06,
+      "loss": 0.666,
+      "step": 11562
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.9898444084273643,
+      "learning_rate": 1.5234599227530678e-06,
+      "loss": 0.6576,
+      "step": 11563
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.029967106432987,
+      "learning_rate": 1.52337760551126e-06,
+      "loss": 0.6595,
+      "step": 11564
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.648024925383263,
+      "learning_rate": 1.5232952833847237e-06,
+      "loss": 0.6751,
+      "step": 11565
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9316147656609647,
+      "learning_rate": 1.5232129563742256e-06,
+      "loss": 0.6803,
+      "step": 11566
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.0215615920315675,
+      "learning_rate": 1.5231306244805345e-06,
+      "loss": 0.6641,
+      "step": 11567
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.5871326680364233,
+      "learning_rate": 1.5230482877044195e-06,
+      "loss": 0.6901,
+      "step": 11568
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.3178526910816928,
+      "learning_rate": 1.5229659460466482e-06,
+      "loss": 0.6608,
+      "step": 11569
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8308826039558111,
+      "learning_rate": 1.5228835995079896e-06,
+      "loss": 0.6621,
+      "step": 11570
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.397895244788803,
+      "learning_rate": 1.522801248089212e-06,
+      "loss": 0.6914,
+      "step": 11571
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.0882705767391025,
+      "learning_rate": 1.5227188917910842e-06,
+      "loss": 0.6823,
+      "step": 11572
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.5058803095394655,
+      "learning_rate": 1.522636530614375e-06,
+      "loss": 0.6465,
+      "step": 11573
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.1363946447048785,
+      "learning_rate": 1.5225541645598524e-06,
+      "loss": 0.6895,
+      "step": 11574
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.1944721711300355,
+      "learning_rate": 1.522471793628286e-06,
+      "loss": 0.6953,
+      "step": 11575
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8110336339702339,
+      "learning_rate": 1.522389417820444e-06,
+      "loss": 0.6517,
+      "step": 11576
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.588500399779806,
+      "learning_rate": 1.5223070371370953e-06,
+      "loss": 0.6719,
+      "step": 11577
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.3921723111493836,
+      "learning_rate": 1.5222246515790088e-06,
+      "loss": 0.7142,
+      "step": 11578
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 5.1171214123031845,
+      "learning_rate": 1.5221422611469533e-06,
+      "loss": 0.6745,
+      "step": 11579
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.3396402673710743,
+      "learning_rate": 1.5220598658416983e-06,
+      "loss": 0.6771,
+      "step": 11580
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.697104120544942,
+      "learning_rate": 1.5219774656640123e-06,
+      "loss": 0.696,
+      "step": 11581
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7781985169403002,
+      "learning_rate": 1.5218950606146644e-06,
+      "loss": 0.6608,
+      "step": 11582
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7566417626600026,
+      "learning_rate": 1.5218126506944238e-06,
+      "loss": 0.6914,
+      "step": 11583
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.2830568251260455,
+      "learning_rate": 1.521730235904059e-06,
+      "loss": 0.6803,
+      "step": 11584
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.4539331009798455,
+      "learning_rate": 1.5216478162443406e-06,
+      "loss": 0.668,
+      "step": 11585
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 6.9534531607532,
+      "learning_rate": 1.5215653917160367e-06,
+      "loss": 0.6758,
+      "step": 11586
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.8038603171777474,
+      "learning_rate": 1.5214829623199167e-06,
+      "loss": 0.6634,
+      "step": 11587
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.691953235362084,
+      "learning_rate": 1.5214005280567501e-06,
+      "loss": 0.668,
+      "step": 11588
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7498415002097164,
+      "learning_rate": 1.5213180889273068e-06,
+      "loss": 0.6608,
+      "step": 11589
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.71542207741776,
+      "learning_rate": 1.5212356449323549e-06,
+      "loss": 0.6732,
+      "step": 11590
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9196561802623485,
+      "learning_rate": 1.521153196072665e-06,
+      "loss": 0.6732,
+      "step": 11591
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7710991431737699,
+      "learning_rate": 1.5210707423490063e-06,
+      "loss": 0.6934,
+      "step": 11592
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.962186235562875,
+      "learning_rate": 1.5209882837621483e-06,
+      "loss": 0.6816,
+      "step": 11593
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.678815193929617,
+      "learning_rate": 1.5209058203128604e-06,
+      "loss": 0.6829,
+      "step": 11594
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.3330679312002494,
+      "learning_rate": 1.5208233520019125e-06,
+      "loss": 0.6602,
+      "step": 11595
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8984089826381332,
+      "learning_rate": 1.520740878830074e-06,
+      "loss": 0.6719,
+      "step": 11596
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.8343623229536266,
+      "learning_rate": 1.520658400798115e-06,
+      "loss": 0.6712,
+      "step": 11597
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.3834998225452395,
+      "learning_rate": 1.5205759179068052e-06,
+      "loss": 0.6934,
+      "step": 11598
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8105577233361673,
+      "learning_rate": 1.5204934301569145e-06,
+      "loss": 0.6439,
+      "step": 11599
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.048531568203376,
+      "learning_rate": 1.520410937549212e-06,
+      "loss": 0.6758,
+      "step": 11600
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.9296005252046927,
+      "learning_rate": 1.5203284400844688e-06,
+      "loss": 0.6953,
+      "step": 11601
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7920750781823325,
+      "learning_rate": 1.520245937763454e-06,
+      "loss": 0.6562,
+      "step": 11602
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.755575377984757,
+      "learning_rate": 1.5201634305869377e-06,
+      "loss": 0.6868,
+      "step": 11603
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.293168813708451,
+      "learning_rate": 1.5200809185556901e-06,
+      "loss": 0.6823,
+      "step": 11604
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.152477707174796,
+      "learning_rate": 1.5199984016704815e-06,
+      "loss": 0.7103,
+      "step": 11605
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.812932331061227,
+      "learning_rate": 1.5199158799320818e-06,
+      "loss": 0.6829,
+      "step": 11606
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.689979610829265,
+      "learning_rate": 1.5198333533412611e-06,
+      "loss": 0.6667,
+      "step": 11607
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.947381732256308,
+      "learning_rate": 1.51975082189879e-06,
+      "loss": 0.6725,
+      "step": 11608
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8493489373623441,
+      "learning_rate": 1.519668285605438e-06,
+      "loss": 0.6693,
+      "step": 11609
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.923026684726481,
+      "learning_rate": 1.5195857444619764e-06,
+      "loss": 0.6706,
+      "step": 11610
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.9706312979209764,
+      "learning_rate": 1.5195031984691751e-06,
+      "loss": 0.6842,
+      "step": 11611
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 4.116667782073485,
+      "learning_rate": 1.519420647627804e-06,
+      "loss": 0.6393,
+      "step": 11612
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7841469498733964,
+      "learning_rate": 1.5193380919386347e-06,
+      "loss": 0.6855,
+      "step": 11613
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.4075788508962361,
+      "learning_rate": 1.5192555314024367e-06,
+      "loss": 0.6947,
+      "step": 11614
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8347634397416688,
+      "learning_rate": 1.519172966019981e-06,
+      "loss": 0.6797,
+      "step": 11615
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.1362814766622633,
+      "learning_rate": 1.519090395792038e-06,
+      "loss": 0.6341,
+      "step": 11616
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.596588025691175,
+      "learning_rate": 1.5190078207193783e-06,
+      "loss": 0.6862,
+      "step": 11617
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.5529075220931947,
+      "learning_rate": 1.5189252408027732e-06,
+      "loss": 0.6374,
+      "step": 11618
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8573236431435471,
+      "learning_rate": 1.5188426560429925e-06,
+      "loss": 0.6549,
+      "step": 11619
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7438893202869956,
+      "learning_rate": 1.5187600664408078e-06,
+      "loss": 0.6673,
+      "step": 11620
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.3847075786487735,
+      "learning_rate": 1.5186774719969892e-06,
+      "loss": 0.6784,
+      "step": 11621
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.8787568012608378,
+      "learning_rate": 1.518594872712308e-06,
+      "loss": 0.6777,
+      "step": 11622
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.583205494389302,
+      "learning_rate": 1.5185122685875349e-06,
+      "loss": 0.6576,
+      "step": 11623
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.8319387431716527,
+      "learning_rate": 1.5184296596234409e-06,
+      "loss": 0.6458,
+      "step": 11624
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.0310397087664336,
+      "learning_rate": 1.5183470458207972e-06,
+      "loss": 0.6738,
+      "step": 11625
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.048240224710574,
+      "learning_rate": 1.5182644271803746e-06,
+      "loss": 0.6934,
+      "step": 11626
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.3226785533415217,
+      "learning_rate": 1.5181818037029443e-06,
+      "loss": 0.6732,
+      "step": 11627
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.1248389000887298,
+      "learning_rate": 1.5180991753892773e-06,
+      "loss": 0.6374,
+      "step": 11628
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.9169367184289778,
+      "learning_rate": 1.518016542240145e-06,
+      "loss": 0.6582,
+      "step": 11629
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6834015868563286,
+      "learning_rate": 1.5179339042563184e-06,
+      "loss": 0.6784,
+      "step": 11630
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.980353420324588,
+      "learning_rate": 1.5178512614385693e-06,
+      "loss": 0.668,
+      "step": 11631
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.0836534779786946,
+      "learning_rate": 1.5177686137876679e-06,
+      "loss": 0.696,
+      "step": 11632
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 3.1520682866616996,
+      "learning_rate": 1.5176859613043865e-06,
+      "loss": 0.7109,
+      "step": 11633
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.578042247536175,
+      "learning_rate": 1.5176033039894964e-06,
+      "loss": 0.6667,
+      "step": 11634
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.1648769501707026,
+      "learning_rate": 1.517520641843769e-06,
+      "loss": 0.6953,
+      "step": 11635
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.0690594805974056,
+      "learning_rate": 1.5174379748679754e-06,
+      "loss": 0.6647,
+      "step": 11636
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.734042493306766,
+      "learning_rate": 1.5173553030628874e-06,
+      "loss": 0.6777,
+      "step": 11637
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.674909294002542,
+      "learning_rate": 1.5172726264292766e-06,
+      "loss": 0.625,
+      "step": 11638
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5597498693590062,
+      "learning_rate": 1.517189944967915e-06,
+      "loss": 0.6602,
+      "step": 11639
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.534674558937592,
+      "learning_rate": 1.5171072586795734e-06,
+      "loss": 0.6875,
+      "step": 11640
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.3925490427918277,
+      "learning_rate": 1.5170245675650245e-06,
+      "loss": 0.6725,
+      "step": 11641
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7830981449797252,
+      "learning_rate": 1.5169418716250394e-06,
+      "loss": 0.6875,
+      "step": 11642
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6718027732936311,
+      "learning_rate": 1.5168591708603898e-06,
+      "loss": 0.7174,
+      "step": 11643
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.3655796330956846,
+      "learning_rate": 1.5167764652718481e-06,
+      "loss": 0.6426,
+      "step": 11644
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.366375865099568,
+      "learning_rate": 1.516693754860186e-06,
+      "loss": 0.6654,
+      "step": 11645
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6845297689640262,
+      "learning_rate": 1.5166110396261752e-06,
+      "loss": 0.6673,
+      "step": 11646
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.6131129807422697,
+      "learning_rate": 1.5165283195705882e-06,
+      "loss": 0.6497,
+      "step": 11647
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5480615405526605,
+      "learning_rate": 1.5164455946941964e-06,
+      "loss": 0.6914,
+      "step": 11648
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.290725911115285,
+      "learning_rate": 1.5163628649977724e-06,
+      "loss": 0.6686,
+      "step": 11649
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.7056990708665234,
+      "learning_rate": 1.5162801304820879e-06,
+      "loss": 0.6829,
+      "step": 11650
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.086941574720548,
+      "learning_rate": 1.5161973911479154e-06,
+      "loss": 0.668,
+      "step": 11651
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.5073850971692573,
+      "learning_rate": 1.516114646996027e-06,
+      "loss": 0.6621,
+      "step": 11652
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.408000172865694,
+      "learning_rate": 1.516031898027195e-06,
+      "loss": 0.6855,
+      "step": 11653
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8671338941878278,
+      "learning_rate": 1.5159491442421916e-06,
+      "loss": 0.6549,
+      "step": 11654
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.324143360298283,
+      "learning_rate": 1.5158663856417896e-06,
+      "loss": 0.694,
+      "step": 11655
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 5.1036436938391665,
+      "learning_rate": 1.51578362222676e-06,
+      "loss": 0.6888,
+      "step": 11656
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0384618015227467,
+      "learning_rate": 1.515700853997877e-06,
+      "loss": 0.6751,
+      "step": 11657
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.6521469710078587,
+      "learning_rate": 1.5156180809559126e-06,
+      "loss": 0.6908,
+      "step": 11658
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9122986460173937,
+      "learning_rate": 1.5155353031016382e-06,
+      "loss": 0.6745,
+      "step": 11659
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8215996448005556,
+      "learning_rate": 1.515452520435828e-06,
+      "loss": 0.6934,
+      "step": 11660
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.8730755115475497,
+      "learning_rate": 1.5153697329592534e-06,
+      "loss": 0.6641,
+      "step": 11661
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7957477053966993,
+      "learning_rate": 1.5152869406726875e-06,
+      "loss": 0.6973,
+      "step": 11662
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.705606701711637,
+      "learning_rate": 1.5152041435769033e-06,
+      "loss": 0.6777,
+      "step": 11663
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.3599023865150701,
+      "learning_rate": 1.515121341672673e-06,
+      "loss": 0.6797,
+      "step": 11664
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.1937504044767993,
+      "learning_rate": 1.51503853496077e-06,
+      "loss": 0.6536,
+      "step": 11665
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.2271377330301583,
+      "learning_rate": 1.5149557234419662e-06,
+      "loss": 0.6934,
+      "step": 11666
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5345637101273377,
+      "learning_rate": 1.5148729071170356e-06,
+      "loss": 0.7038,
+      "step": 11667
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.56054385704243,
+      "learning_rate": 1.5147900859867506e-06,
+      "loss": 0.6797,
+      "step": 11668
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.9796679633809237,
+      "learning_rate": 1.5147072600518839e-06,
+      "loss": 0.6452,
+      "step": 11669
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 6.189061867063472,
+      "learning_rate": 1.5146244293132094e-06,
+      "loss": 0.6608,
+      "step": 11670
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0175976025365658,
+      "learning_rate": 1.514541593771499e-06,
+      "loss": 0.6621,
+      "step": 11671
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.3684597700371743,
+      "learning_rate": 1.5144587534275266e-06,
+      "loss": 0.6953,
+      "step": 11672
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8238939219503716,
+      "learning_rate": 1.5143759082820653e-06,
+      "loss": 0.6576,
+      "step": 11673
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.128895102397412,
+      "learning_rate": 1.514293058335888e-06,
+      "loss": 0.6719,
+      "step": 11674
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.7589103092609037,
+      "learning_rate": 1.514210203589768e-06,
+      "loss": 0.6784,
+      "step": 11675
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7630656788660078,
+      "learning_rate": 1.5141273440444789e-06,
+      "loss": 0.6634,
+      "step": 11676
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.8370886725229035,
+      "learning_rate": 1.5140444797007939e-06,
+      "loss": 0.6706,
+      "step": 11677
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.1732053176297605,
+      "learning_rate": 1.5139616105594861e-06,
+      "loss": 0.6686,
+      "step": 11678
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.066842656356591,
+      "learning_rate": 1.5138787366213293e-06,
+      "loss": 0.6654,
+      "step": 11679
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.958821708229822,
+      "learning_rate": 1.5137958578870967e-06,
+      "loss": 0.6771,
+      "step": 11680
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.7209718174774036,
+      "learning_rate": 1.513712974357562e-06,
+      "loss": 0.6732,
+      "step": 11681
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.1888983458149633,
+      "learning_rate": 1.5136300860334985e-06,
+      "loss": 0.6621,
+      "step": 11682
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4225821767614661,
+      "learning_rate": 1.5135471929156798e-06,
+      "loss": 0.651,
+      "step": 11683
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.0175234328720022,
+      "learning_rate": 1.5134642950048804e-06,
+      "loss": 0.6615,
+      "step": 11684
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.639854662788865,
+      "learning_rate": 1.5133813923018728e-06,
+      "loss": 0.6673,
+      "step": 11685
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.820807585151161,
+      "learning_rate": 1.5132984848074316e-06,
+      "loss": 0.696,
+      "step": 11686
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.5068967910038906,
+      "learning_rate": 1.5132155725223302e-06,
+      "loss": 0.6836,
+      "step": 11687
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.442580002273388,
+      "learning_rate": 1.513132655447342e-06,
+      "loss": 0.7057,
+      "step": 11688
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.099993433891396,
+      "learning_rate": 1.5130497335832417e-06,
+      "loss": 0.6719,
+      "step": 11689
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.888585341331257,
+      "learning_rate": 1.512966806930803e-06,
+      "loss": 0.7077,
+      "step": 11690
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.2196715299738263,
+      "learning_rate": 1.5128838754907993e-06,
+      "loss": 0.6562,
+      "step": 11691
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.1530351908870633,
+      "learning_rate": 1.5128009392640056e-06,
+      "loss": 0.6764,
+      "step": 11692
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.6675962397131356,
+      "learning_rate": 1.5127179982511948e-06,
+      "loss": 0.6797,
+      "step": 11693
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.9382717209708455,
+      "learning_rate": 1.5126350524531421e-06,
+      "loss": 0.6458,
+      "step": 11694
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.9544699931989893,
+      "learning_rate": 1.5125521018706208e-06,
+      "loss": 0.6693,
+      "step": 11695
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.8733528750059127,
+      "learning_rate": 1.5124691465044056e-06,
+      "loss": 0.6608,
+      "step": 11696
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.3477874018045912,
+      "learning_rate": 1.5123861863552706e-06,
+      "loss": 0.6751,
+      "step": 11697
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 4.522340410197998,
+      "learning_rate": 1.5123032214239898e-06,
+      "loss": 0.6719,
+      "step": 11698
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.7754910642119817,
+      "learning_rate": 1.5122202517113376e-06,
+      "loss": 0.6895,
+      "step": 11699
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.668492754257447,
+      "learning_rate": 1.5121372772180889e-06,
+      "loss": 0.6901,
+      "step": 11700
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9808868201252721,
+      "learning_rate": 1.5120542979450173e-06,
+      "loss": 0.6719,
+      "step": 11701
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.1852698068562173,
+      "learning_rate": 1.5119713138928978e-06,
+      "loss": 0.6868,
+      "step": 11702
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.985251422585913,
+      "learning_rate": 1.5118883250625047e-06,
+      "loss": 0.6699,
+      "step": 11703
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.845808520644887,
+      "learning_rate": 1.5118053314546128e-06,
+      "loss": 0.6686,
+      "step": 11704
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.670887622681206,
+      "learning_rate": 1.5117223330699963e-06,
+      "loss": 0.6882,
+      "step": 11705
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.663667101670586,
+      "learning_rate": 1.5116393299094302e-06,
+      "loss": 0.7148,
+      "step": 11706
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8474500720777891,
+      "learning_rate": 1.5115563219736886e-06,
+      "loss": 0.6608,
+      "step": 11707
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.9685927150975844,
+      "learning_rate": 1.511473309263547e-06,
+      "loss": 0.6888,
+      "step": 11708
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9589983700045894,
+      "learning_rate": 1.5113902917797795e-06,
+      "loss": 0.6647,
+      "step": 11709
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.8002578771345643,
+      "learning_rate": 1.5113072695231615e-06,
+      "loss": 0.666,
+      "step": 11710
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.000152634315775,
+      "learning_rate": 1.5112242424944674e-06,
+      "loss": 0.6797,
+      "step": 11711
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.9322798172933306,
+      "learning_rate": 1.511141210694472e-06,
+      "loss": 0.6777,
+      "step": 11712
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5522351026181698,
+      "learning_rate": 1.511058174123951e-06,
+      "loss": 0.6771,
+      "step": 11713
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.4376500280099753,
+      "learning_rate": 1.5109751327836783e-06,
+      "loss": 0.6823,
+      "step": 11714
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.133744393444653,
+      "learning_rate": 1.5108920866744298e-06,
+      "loss": 0.6608,
+      "step": 11715
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.546706879089998,
+      "learning_rate": 1.5108090357969801e-06,
+      "loss": 0.6706,
+      "step": 11716
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.731104980194241,
+      "learning_rate": 1.5107259801521047e-06,
+      "loss": 0.7064,
+      "step": 11717
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7583058862422003,
+      "learning_rate": 1.5106429197405784e-06,
+      "loss": 0.6667,
+      "step": 11718
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8482528283940698,
+      "learning_rate": 1.5105598545631764e-06,
+      "loss": 0.6745,
+      "step": 11719
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0380172378742518,
+      "learning_rate": 1.5104767846206742e-06,
+      "loss": 0.6712,
+      "step": 11720
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7738137506707194,
+      "learning_rate": 1.5103937099138475e-06,
+      "loss": 0.6628,
+      "step": 11721
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.3619134645527178,
+      "learning_rate": 1.5103106304434706e-06,
+      "loss": 0.6829,
+      "step": 11722
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.8115747149917643,
+      "learning_rate": 1.51022754621032e-06,
+      "loss": 0.6576,
+      "step": 11723
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8911371882306691,
+      "learning_rate": 1.51014445721517e-06,
+      "loss": 0.6836,
+      "step": 11724
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9441731241047443,
+      "learning_rate": 1.5100613634587972e-06,
+      "loss": 0.6732,
+      "step": 11725
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.124621832842882,
+      "learning_rate": 1.509978264941976e-06,
+      "loss": 0.6921,
+      "step": 11726
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9042321596353384,
+      "learning_rate": 1.5098951616654831e-06,
+      "loss": 0.6348,
+      "step": 11727
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8568073939113832,
+      "learning_rate": 1.5098120536300933e-06,
+      "loss": 0.6725,
+      "step": 11728
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.007098443783929,
+      "learning_rate": 1.5097289408365828e-06,
+      "loss": 0.6719,
+      "step": 11729
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.742681432105714,
+      "learning_rate": 1.5096458232857265e-06,
+      "loss": 0.6654,
+      "step": 11730
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7390802220852914,
+      "learning_rate": 1.5095627009783012e-06,
+      "loss": 0.6784,
+      "step": 11731
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.5044016492412275,
+      "learning_rate": 1.509479573915082e-06,
+      "loss": 0.6797,
+      "step": 11732
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.420161221106136,
+      "learning_rate": 1.5093964420968446e-06,
+      "loss": 0.6842,
+      "step": 11733
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8716573670366264,
+      "learning_rate": 1.5093133055243655e-06,
+      "loss": 0.668,
+      "step": 11734
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0926197160540283,
+      "learning_rate": 1.5092301641984198e-06,
+      "loss": 0.6764,
+      "step": 11735
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.124886753499349,
+      "learning_rate": 1.5091470181197842e-06,
+      "loss": 0.681,
+      "step": 11736
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.6996541516420773,
+      "learning_rate": 1.5090638672892348e-06,
+      "loss": 0.6576,
+      "step": 11737
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.079867183281646,
+      "learning_rate": 1.5089807117075468e-06,
+      "loss": 0.6569,
+      "step": 11738
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8626901102552127,
+      "learning_rate": 1.508897551375497e-06,
+      "loss": 0.6602,
+      "step": 11739
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.668982829161387,
+      "learning_rate": 1.5088143862938615e-06,
+      "loss": 0.681,
+      "step": 11740
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8485745049430399,
+      "learning_rate": 1.508731216463416e-06,
+      "loss": 0.6738,
+      "step": 11741
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.2750799942495576,
+      "learning_rate": 1.5086480418849374e-06,
+      "loss": 0.6602,
+      "step": 11742
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 5.518797548600508,
+      "learning_rate": 1.5085648625592015e-06,
+      "loss": 0.6615,
+      "step": 11743
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.246773864064693,
+      "learning_rate": 1.508481678486985e-06,
+      "loss": 0.6634,
+      "step": 11744
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.2517240488636565,
+      "learning_rate": 1.5083984896690637e-06,
+      "loss": 0.6615,
+      "step": 11745
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.912652900589948,
+      "learning_rate": 1.5083152961062144e-06,
+      "loss": 0.6914,
+      "step": 11746
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5736414098559168,
+      "learning_rate": 1.5082320977992136e-06,
+      "loss": 0.6478,
+      "step": 11747
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.9195811523248505,
+      "learning_rate": 1.5081488947488374e-06,
+      "loss": 0.6836,
+      "step": 11748
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8881323627140221,
+      "learning_rate": 1.508065686955863e-06,
+      "loss": 0.6888,
+      "step": 11749
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.55157200683781,
+      "learning_rate": 1.5079824744210666e-06,
+      "loss": 0.7201,
+      "step": 11750
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.78725460763632,
+      "learning_rate": 1.5078992571452248e-06,
+      "loss": 0.7129,
+      "step": 11751
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.097324154866867,
+      "learning_rate": 1.5078160351291141e-06,
+      "loss": 0.6589,
+      "step": 11752
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.3551762123864088,
+      "learning_rate": 1.5077328083735114e-06,
+      "loss": 0.6608,
+      "step": 11753
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9158405835821134,
+      "learning_rate": 1.5076495768791938e-06,
+      "loss": 0.6725,
+      "step": 11754
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.854191454704687,
+      "learning_rate": 1.5075663406469374e-06,
+      "loss": 0.6784,
+      "step": 11755
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8042457062349715,
+      "learning_rate": 1.50748309967752e-06,
+      "loss": 0.6927,
+      "step": 11756
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.216094859454101,
+      "learning_rate": 1.5073998539717174e-06,
+      "loss": 0.679,
+      "step": 11757
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 3.445695995665239,
+      "learning_rate": 1.5073166035303074e-06,
+      "loss": 0.6647,
+      "step": 11758
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8292577800199691,
+      "learning_rate": 1.5072333483540668e-06,
+      "loss": 0.6849,
+      "step": 11759
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.35317902254148,
+      "learning_rate": 1.507150088443772e-06,
+      "loss": 0.6712,
+      "step": 11760
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.729325958170044,
+      "learning_rate": 1.507066823800201e-06,
+      "loss": 0.6908,
+      "step": 11761
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.2129496355750091,
+      "learning_rate": 1.5069835544241306e-06,
+      "loss": 0.6504,
+      "step": 11762
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8359255765640746,
+      "learning_rate": 1.5069002803163375e-06,
+      "loss": 0.6719,
+      "step": 11763
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.4293475387923937,
+      "learning_rate": 1.5068170014775996e-06,
+      "loss": 0.6934,
+      "step": 11764
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.6840015164287596,
+      "learning_rate": 1.5067337179086935e-06,
+      "loss": 0.6914,
+      "step": 11765
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.0304237853686806,
+      "learning_rate": 1.506650429610397e-06,
+      "loss": 0.6764,
+      "step": 11766
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9625838750111442,
+      "learning_rate": 1.506567136583487e-06,
+      "loss": 0.6562,
+      "step": 11767
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.2260194575504286,
+      "learning_rate": 1.5064838388287414e-06,
+      "loss": 0.6719,
+      "step": 11768
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.457817352460932,
+      "learning_rate": 1.506400536346937e-06,
+      "loss": 0.7122,
+      "step": 11769
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8617622502390823,
+      "learning_rate": 1.506317229138852e-06,
+      "loss": 0.6803,
+      "step": 11770
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8030583791832338,
+      "learning_rate": 1.5062339172052632e-06,
+      "loss": 0.6738,
+      "step": 11771
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5428288515733437,
+      "learning_rate": 1.5061506005469485e-06,
+      "loss": 0.666,
+      "step": 11772
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7538789451176319,
+      "learning_rate": 1.5060672791646857e-06,
+      "loss": 0.6888,
+      "step": 11773
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.8512013183999585,
+      "learning_rate": 1.5059839530592523e-06,
+      "loss": 0.6816,
+      "step": 11774
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8577341025300944,
+      "learning_rate": 1.5059006222314255e-06,
+      "loss": 0.6589,
+      "step": 11775
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.3893111473171877,
+      "learning_rate": 1.5058172866819838e-06,
+      "loss": 0.681,
+      "step": 11776
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0051430494885982,
+      "learning_rate": 1.5057339464117045e-06,
+      "loss": 0.679,
+      "step": 11777
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.370607030106942,
+      "learning_rate": 1.5056506014213655e-06,
+      "loss": 0.6947,
+      "step": 11778
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.2944088098218303,
+      "learning_rate": 1.5055672517117448e-06,
+      "loss": 0.6517,
+      "step": 11779
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5889702407908473,
+      "learning_rate": 1.5054838972836203e-06,
+      "loss": 0.6602,
+      "step": 11780
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.06425270626814,
+      "learning_rate": 1.50540053813777e-06,
+      "loss": 0.6797,
+      "step": 11781
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.513488673548674,
+      "learning_rate": 1.5053171742749713e-06,
+      "loss": 0.6836,
+      "step": 11782
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9612609255488433,
+      "learning_rate": 1.5052338056960033e-06,
+      "loss": 0.6777,
+      "step": 11783
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.2982710414217042,
+      "learning_rate": 1.505150432401643e-06,
+      "loss": 0.6452,
+      "step": 11784
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.001752747085555,
+      "learning_rate": 1.5050670543926694e-06,
+      "loss": 0.6673,
+      "step": 11785
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8386356885927152,
+      "learning_rate": 1.50498367166986e-06,
+      "loss": 0.6569,
+      "step": 11786
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.216404248177488,
+      "learning_rate": 1.5049002842339937e-06,
+      "loss": 0.6901,
+      "step": 11787
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8746050647518806,
+      "learning_rate": 1.5048168920858481e-06,
+      "loss": 0.6641,
+      "step": 11788
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.007936045132978,
+      "learning_rate": 1.504733495226202e-06,
+      "loss": 0.6732,
+      "step": 11789
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9940780530170762,
+      "learning_rate": 1.5046500936558336e-06,
+      "loss": 0.6621,
+      "step": 11790
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.286668490600123,
+      "learning_rate": 1.504566687375521e-06,
+      "loss": 0.6615,
+      "step": 11791
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.113884433397335,
+      "learning_rate": 1.504483276386043e-06,
+      "loss": 0.694,
+      "step": 11792
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.130701131838655,
+      "learning_rate": 1.5043998606881777e-06,
+      "loss": 0.6927,
+      "step": 11793
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.9707057589661299,
+      "learning_rate": 1.5043164402827043e-06,
+      "loss": 0.707,
+      "step": 11794
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1333850275071067,
+      "learning_rate": 1.504233015170401e-06,
+      "loss": 0.6458,
+      "step": 11795
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.967345358930288,
+      "learning_rate": 1.5041495853520463e-06,
+      "loss": 0.6751,
+      "step": 11796
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.8766178748133813,
+      "learning_rate": 1.504066150828419e-06,
+      "loss": 0.6628,
+      "step": 11797
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0868168440137946,
+      "learning_rate": 1.5039827116002973e-06,
+      "loss": 0.6738,
+      "step": 11798
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.4361660013999082,
+      "learning_rate": 1.5038992676684607e-06,
+      "loss": 0.6992,
+      "step": 11799
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.874610968906433,
+      "learning_rate": 1.5038158190336876e-06,
+      "loss": 0.6517,
+      "step": 11800
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.2347193810875197,
+      "learning_rate": 1.503732365696757e-06,
+      "loss": 0.6543,
+      "step": 11801
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.7807097968519403,
+      "learning_rate": 1.5036489076584476e-06,
+      "loss": 0.6634,
+      "step": 11802
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.2052829909748612,
+      "learning_rate": 1.5035654449195384e-06,
+      "loss": 0.6966,
+      "step": 11803
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0407520402530044,
+      "learning_rate": 1.5034819774808083e-06,
+      "loss": 0.6654,
+      "step": 11804
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9516562565618926,
+      "learning_rate": 1.5033985053430365e-06,
+      "loss": 0.6842,
+      "step": 11805
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.303113726662779,
+      "learning_rate": 1.503315028507002e-06,
+      "loss": 0.696,
+      "step": 11806
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.5516482023384734,
+      "learning_rate": 1.5032315469734835e-06,
+      "loss": 0.6693,
+      "step": 11807
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.24024794211164,
+      "learning_rate": 1.503148060743261e-06,
+      "loss": 0.6921,
+      "step": 11808
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.2816281408427155,
+      "learning_rate": 1.5030645698171127e-06,
+      "loss": 0.6406,
+      "step": 11809
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.382884878161706,
+      "learning_rate": 1.5029810741958184e-06,
+      "loss": 0.6751,
+      "step": 11810
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.5239743386000164,
+      "learning_rate": 1.5028975738801575e-06,
+      "loss": 0.6556,
+      "step": 11811
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 7.2634624871023075,
+      "learning_rate": 1.5028140688709085e-06,
+      "loss": 0.6471,
+      "step": 11812
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9986330481248755,
+      "learning_rate": 1.502730559168852e-06,
+      "loss": 0.6699,
+      "step": 11813
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.667338882905726,
+      "learning_rate": 1.5026470447747662e-06,
+      "loss": 0.6862,
+      "step": 11814
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9467785987362792,
+      "learning_rate": 1.5025635256894314e-06,
+      "loss": 0.681,
+      "step": 11815
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 6.003128370304303,
+      "learning_rate": 1.5024800019136269e-06,
+      "loss": 0.653,
+      "step": 11816
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.753263194373789,
+      "learning_rate": 1.5023964734481315e-06,
+      "loss": 0.6706,
+      "step": 11817
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8972675126174536,
+      "learning_rate": 1.5023129402937258e-06,
+      "loss": 0.6589,
+      "step": 11818
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0723374399305439,
+      "learning_rate": 1.502229402451189e-06,
+      "loss": 0.6803,
+      "step": 11819
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0091499648867628,
+      "learning_rate": 1.5021458599213008e-06,
+      "loss": 0.7142,
+      "step": 11820
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.6102205784224637,
+      "learning_rate": 1.5020623127048407e-06,
+      "loss": 0.6908,
+      "step": 11821
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 4.819287759029297,
+      "learning_rate": 1.501978760802589e-06,
+      "loss": 0.6549,
+      "step": 11822
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0554777611044828,
+      "learning_rate": 1.5018952042153248e-06,
+      "loss": 0.6842,
+      "step": 11823
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8227552776926362,
+      "learning_rate": 1.5018116429438283e-06,
+      "loss": 0.6641,
+      "step": 11824
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.05556486331212,
+      "learning_rate": 1.5017280769888791e-06,
+      "loss": 0.6621,
+      "step": 11825
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.8905808026755597,
+      "learning_rate": 1.501644506351258e-06,
+      "loss": 0.6855,
+      "step": 11826
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.960532632624404,
+      "learning_rate": 1.501560931031744e-06,
+      "loss": 0.6803,
+      "step": 11827
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.9311620083048266,
+      "learning_rate": 1.5014773510311173e-06,
+      "loss": 0.6628,
+      "step": 11828
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.2672474016022472,
+      "learning_rate": 1.5013937663501587e-06,
+      "loss": 0.6296,
+      "step": 11829
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.861423708593997,
+      "learning_rate": 1.5013101769896475e-06,
+      "loss": 0.6816,
+      "step": 11830
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 5.031430030185488,
+      "learning_rate": 1.5012265829503638e-06,
+      "loss": 0.679,
+      "step": 11831
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8450543670503132,
+      "learning_rate": 1.5011429842330884e-06,
+      "loss": 0.6589,
+      "step": 11832
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.916012516038645,
+      "learning_rate": 1.501059380838601e-06,
+      "loss": 0.638,
+      "step": 11833
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.761877612027408,
+      "learning_rate": 1.5009757727676822e-06,
+      "loss": 0.6504,
+      "step": 11834
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8950531470791189,
+      "learning_rate": 1.5008921600211121e-06,
+      "loss": 0.6836,
+      "step": 11835
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 7.92493753689475,
+      "learning_rate": 1.5008085425996715e-06,
+      "loss": 0.6855,
+      "step": 11836
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.796187927828091,
+      "learning_rate": 1.50072492050414e-06,
+      "loss": 0.6647,
+      "step": 11837
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.4514569406666022,
+      "learning_rate": 1.5006412937352988e-06,
+      "loss": 0.7051,
+      "step": 11838
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.4705795237423653,
+      "learning_rate": 1.5005576622939283e-06,
+      "loss": 0.6413,
+      "step": 11839
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9116800195713789,
+      "learning_rate": 1.5004740261808086e-06,
+      "loss": 0.6693,
+      "step": 11840
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9496101375052376,
+      "learning_rate": 1.5003903853967206e-06,
+      "loss": 0.6712,
+      "step": 11841
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.92020781511227,
+      "learning_rate": 1.5003067399424451e-06,
+      "loss": 0.7129,
+      "step": 11842
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.188144097008329,
+      "learning_rate": 1.5002230898187622e-06,
+      "loss": 0.6621,
+      "step": 11843
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.4918613041038322,
+      "learning_rate": 1.5001394350264529e-06,
+      "loss": 0.6979,
+      "step": 11844
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8203228877493811,
+      "learning_rate": 1.5000557755662983e-06,
+      "loss": 0.6706,
+      "step": 11845
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.041312645270769,
+      "learning_rate": 1.4999721114390788e-06,
+      "loss": 0.7077,
+      "step": 11846
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.964176428886507,
+      "learning_rate": 1.4998884426455755e-06,
+      "loss": 0.7122,
+      "step": 11847
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5640747797333203,
+      "learning_rate": 1.4998047691865692e-06,
+      "loss": 0.6634,
+      "step": 11848
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9422494964405078,
+      "learning_rate": 1.4997210910628407e-06,
+      "loss": 0.6536,
+      "step": 11849
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.0585500664188214,
+      "learning_rate": 1.4996374082751711e-06,
+      "loss": 0.6751,
+      "step": 11850
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.6681037504993805,
+      "learning_rate": 1.4995537208243412e-06,
+      "loss": 0.6348,
+      "step": 11851
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8274509530317121,
+      "learning_rate": 1.4994700287111323e-06,
+      "loss": 0.6686,
+      "step": 11852
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.4504090668719085,
+      "learning_rate": 1.4993863319363257e-06,
+      "loss": 0.6686,
+      "step": 11853
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8132152108299262,
+      "learning_rate": 1.499302630500702e-06,
+      "loss": 0.6816,
+      "step": 11854
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 6.301834957961106,
+      "learning_rate": 1.499218924405043e-06,
+      "loss": 0.679,
+      "step": 11855
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.309230422224279,
+      "learning_rate": 1.4991352136501295e-06,
+      "loss": 0.6576,
+      "step": 11856
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9867877826442824,
+      "learning_rate": 1.4990514982367425e-06,
+      "loss": 0.6602,
+      "step": 11857
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8019476117946285,
+      "learning_rate": 1.4989677781656642e-06,
+      "loss": 0.6634,
+      "step": 11858
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.0308371064191941,
+      "learning_rate": 1.4988840534376755e-06,
+      "loss": 0.6908,
+      "step": 11859
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.357422136554526,
+      "learning_rate": 1.4988003240535575e-06,
+      "loss": 0.6673,
+      "step": 11860
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5799936965226196,
+      "learning_rate": 1.4987165900140923e-06,
+      "loss": 0.6875,
+      "step": 11861
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.222815955533011,
+      "learning_rate": 1.498632851320061e-06,
+      "loss": 0.6803,
+      "step": 11862
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.1426809405156746,
+      "learning_rate": 1.4985491079722452e-06,
+      "loss": 0.6771,
+      "step": 11863
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8467305029649089,
+      "learning_rate": 1.4984653599714262e-06,
+      "loss": 0.6628,
+      "step": 11864
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.2409581988026055,
+      "learning_rate": 1.4983816073183864e-06,
+      "loss": 0.6641,
+      "step": 11865
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.1139425088957298,
+      "learning_rate": 1.4982978500139067e-06,
+      "loss": 0.6823,
+      "step": 11866
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.7658589013112707,
+      "learning_rate": 1.498214088058769e-06,
+      "loss": 0.6868,
+      "step": 11867
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.110973695374311,
+      "learning_rate": 1.4981303214537552e-06,
+      "loss": 0.6829,
+      "step": 11868
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5515386446411894,
+      "learning_rate": 1.4980465501996471e-06,
+      "loss": 0.6901,
+      "step": 11869
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 5.634978959427809,
+      "learning_rate": 1.4979627742972266e-06,
+      "loss": 0.7214,
+      "step": 11870
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.3836400575376067,
+      "learning_rate": 1.4978789937472754e-06,
+      "loss": 0.6556,
+      "step": 11871
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.0147205892281166,
+      "learning_rate": 1.4977952085505757e-06,
+      "loss": 0.6849,
+      "step": 11872
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.344861692173813,
+      "learning_rate": 1.497711418707909e-06,
+      "loss": 0.6621,
+      "step": 11873
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.0205742766844867,
+      "learning_rate": 1.4976276242200578e-06,
+      "loss": 0.6816,
+      "step": 11874
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.338425760108114,
+      "learning_rate": 1.4975438250878038e-06,
+      "loss": 0.7044,
+      "step": 11875
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.082614463799419,
+      "learning_rate": 1.4974600213119292e-06,
+      "loss": 0.6615,
+      "step": 11876
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.616517977386176,
+      "learning_rate": 1.4973762128932167e-06,
+      "loss": 0.7031,
+      "step": 11877
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9376554526609723,
+      "learning_rate": 1.4972923998324476e-06,
+      "loss": 0.6784,
+      "step": 11878
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 4.008562640359669,
+      "learning_rate": 1.497208582130405e-06,
+      "loss": 0.6862,
+      "step": 11879
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8244749610448849,
+      "learning_rate": 1.4971247597878702e-06,
+      "loss": 0.6823,
+      "step": 11880
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8341742137596365,
+      "learning_rate": 1.4970409328056262e-06,
+      "loss": 0.6803,
+      "step": 11881
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.9826178979596656,
+      "learning_rate": 1.4969571011844556e-06,
+      "loss": 0.6491,
+      "step": 11882
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.1533546988486982,
+      "learning_rate": 1.4968732649251398e-06,
+      "loss": 0.679,
+      "step": 11883
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 3.048673033854098,
+      "learning_rate": 1.4967894240284624e-06,
+      "loss": 0.6719,
+      "step": 11884
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7110782278830865,
+      "learning_rate": 1.4967055784952054e-06,
+      "loss": 0.6888,
+      "step": 11885
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7204755348090529,
+      "learning_rate": 1.496621728326151e-06,
+      "loss": 0.6673,
+      "step": 11886
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7252211892485284,
+      "learning_rate": 1.4965378735220821e-06,
+      "loss": 0.679,
+      "step": 11887
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.3396140316993785,
+      "learning_rate": 1.4964540140837811e-06,
+      "loss": 0.651,
+      "step": 11888
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7460820223769195,
+      "learning_rate": 1.4963701500120314e-06,
+      "loss": 0.6862,
+      "step": 11889
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.3870264996725012,
+      "learning_rate": 1.496286281307615e-06,
+      "loss": 0.679,
+      "step": 11890
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.7052917494123963,
+      "learning_rate": 1.4962024079713146e-06,
+      "loss": 0.6732,
+      "step": 11891
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8996906448173868,
+      "learning_rate": 1.4961185300039136e-06,
+      "loss": 0.6868,
+      "step": 11892
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7861706217275792,
+      "learning_rate": 1.4960346474061944e-06,
+      "loss": 0.6576,
+      "step": 11893
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.9507765812620215,
+      "learning_rate": 1.4959507601789398e-06,
+      "loss": 0.6816,
+      "step": 11894
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8651622062190325,
+      "learning_rate": 1.4958668683229335e-06,
+      "loss": 0.6667,
+      "step": 11895
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7214303148764029,
+      "learning_rate": 1.495782971838957e-06,
+      "loss": 0.6686,
+      "step": 11896
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8333290010918981,
+      "learning_rate": 1.4956990707277946e-06,
+      "loss": 0.6699,
+      "step": 11897
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.6835061481677327,
+      "learning_rate": 1.495615164990229e-06,
+      "loss": 0.6556,
+      "step": 11898
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.121039729635472,
+      "learning_rate": 1.495531254627043e-06,
+      "loss": 0.6549,
+      "step": 11899
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9036911073309608,
+      "learning_rate": 1.4954473396390203e-06,
+      "loss": 0.696,
+      "step": 11900
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9561516795920127,
+      "learning_rate": 1.4953634200269435e-06,
+      "loss": 0.6947,
+      "step": 11901
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.144152904754944,
+      "learning_rate": 1.4952794957915963e-06,
+      "loss": 0.6719,
+      "step": 11902
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.097225228244218,
+      "learning_rate": 1.4951955669337614e-06,
+      "loss": 0.638,
+      "step": 11903
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2829292091950437,
+      "learning_rate": 1.4951116334542227e-06,
+      "loss": 0.6602,
+      "step": 11904
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1816851366846484,
+      "learning_rate": 1.4950276953537632e-06,
+      "loss": 0.6868,
+      "step": 11905
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.29633935943123,
+      "learning_rate": 1.494943752633167e-06,
+      "loss": 0.6523,
+      "step": 11906
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1111959131688873,
+      "learning_rate": 1.4948598052932164e-06,
+      "loss": 0.6667,
+      "step": 11907
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.3737695267751824,
+      "learning_rate": 1.494775853334696e-06,
+      "loss": 0.6556,
+      "step": 11908
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2762668783396212,
+      "learning_rate": 1.494691896758388e-06,
+      "loss": 0.6432,
+      "step": 11909
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.18497274254999,
+      "learning_rate": 1.4946079355650772e-06,
+      "loss": 0.7077,
+      "step": 11910
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.737302863331133,
+      "learning_rate": 1.4945239697555469e-06,
+      "loss": 0.6628,
+      "step": 11911
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1368534735248739,
+      "learning_rate": 1.4944399993305806e-06,
+      "loss": 0.6556,
+      "step": 11912
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.4329242832253484,
+      "learning_rate": 1.4943560242909624e-06,
+      "loss": 0.6823,
+      "step": 11913
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2850862382346653,
+      "learning_rate": 1.494272044637475e-06,
+      "loss": 0.6296,
+      "step": 11914
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7403928642933437,
+      "learning_rate": 1.4941880603709034e-06,
+      "loss": 0.6497,
+      "step": 11915
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.9612363915610491,
+      "learning_rate": 1.4941040714920307e-06,
+      "loss": 0.7142,
+      "step": 11916
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1317689795769958,
+      "learning_rate": 1.494020078001641e-06,
+      "loss": 0.6621,
+      "step": 11917
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.8427468115715566,
+      "learning_rate": 1.4939360799005183e-06,
+      "loss": 0.6771,
+      "step": 11918
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2947167097242702,
+      "learning_rate": 1.4938520771894467e-06,
+      "loss": 0.6771,
+      "step": 11919
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.327916624652714,
+      "learning_rate": 1.4937680698692096e-06,
+      "loss": 0.6823,
+      "step": 11920
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7230629017586712,
+      "learning_rate": 1.4936840579405918e-06,
+      "loss": 0.6589,
+      "step": 11921
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.689463641058938,
+      "learning_rate": 1.4936000414043767e-06,
+      "loss": 0.6569,
+      "step": 11922
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9120818416043547,
+      "learning_rate": 1.493516020261349e-06,
+      "loss": 0.6699,
+      "step": 11923
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.064874820586603,
+      "learning_rate": 1.4934319945122927e-06,
+      "loss": 0.6777,
+      "step": 11924
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.685247495060427,
+      "learning_rate": 1.4933479641579916e-06,
+      "loss": 0.6211,
+      "step": 11925
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.876406389124897,
+      "learning_rate": 1.4932639291992305e-06,
+      "loss": 0.6816,
+      "step": 11926
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1102815904510832,
+      "learning_rate": 1.4931798896367936e-06,
+      "loss": 0.6562,
+      "step": 11927
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.6394869033668757,
+      "learning_rate": 1.4930958454714651e-06,
+      "loss": 0.6576,
+      "step": 11928
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.780731970463914,
+      "learning_rate": 1.4930117967040296e-06,
+      "loss": 0.6758,
+      "step": 11929
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.4974105950965768,
+      "learning_rate": 1.4929277433352712e-06,
+      "loss": 0.6706,
+      "step": 11930
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.049732351195819,
+      "learning_rate": 1.4928436853659744e-06,
+      "loss": 0.6849,
+      "step": 11931
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.4108486554292097,
+      "learning_rate": 1.4927596227969246e-06,
+      "loss": 0.6895,
+      "step": 11932
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.240513847990413,
+      "learning_rate": 1.4926755556289053e-06,
+      "loss": 0.653,
+      "step": 11933
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.1511509515329594,
+      "learning_rate": 1.4925914838627014e-06,
+      "loss": 0.6602,
+      "step": 11934
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.822153785667129,
+      "learning_rate": 1.4925074074990976e-06,
+      "loss": 0.6719,
+      "step": 11935
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.7276137127640174,
+      "learning_rate": 1.4924233265388786e-06,
+      "loss": 0.6667,
+      "step": 11936
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.462806184688508,
+      "learning_rate": 1.492339240982829e-06,
+      "loss": 0.6719,
+      "step": 11937
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.8226597773353523,
+      "learning_rate": 1.492255150831734e-06,
+      "loss": 0.6367,
+      "step": 11938
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.171563141680207,
+      "learning_rate": 1.492171056086378e-06,
+      "loss": 0.6725,
+      "step": 11939
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.0817821219180772,
+      "learning_rate": 1.492086956747546e-06,
+      "loss": 0.6484,
+      "step": 11940
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.585320023523208,
+      "learning_rate": 1.4920028528160231e-06,
+      "loss": 0.6882,
+      "step": 11941
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6672328469142215,
+      "learning_rate": 1.4919187442925939e-06,
+      "loss": 0.6562,
+      "step": 11942
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.111536722724045,
+      "learning_rate": 1.4918346311780436e-06,
+      "loss": 0.6673,
+      "step": 11943
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.0750881211127,
+      "learning_rate": 1.4917505134731573e-06,
+      "loss": 0.7148,
+      "step": 11944
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 7.902852149065372,
+      "learning_rate": 1.49166639117872e-06,
+      "loss": 0.7012,
+      "step": 11945
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2506703394202714,
+      "learning_rate": 1.4915822642955165e-06,
+      "loss": 0.6914,
+      "step": 11946
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.363325083034512,
+      "learning_rate": 1.4914981328243323e-06,
+      "loss": 0.6901,
+      "step": 11947
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.5135204240044855,
+      "learning_rate": 1.491413996765953e-06,
+      "loss": 0.6595,
+      "step": 11948
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.205101993979222,
+      "learning_rate": 1.4913298561211627e-06,
+      "loss": 0.6693,
+      "step": 11949
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2298943758375105,
+      "learning_rate": 1.491245710890748e-06,
+      "loss": 0.6934,
+      "step": 11950
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.093379173846382,
+      "learning_rate": 1.4911615610754935e-06,
+      "loss": 0.6882,
+      "step": 11951
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9823267525787449,
+      "learning_rate": 1.4910774066761845e-06,
+      "loss": 0.6387,
+      "step": 11952
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.559858576314931,
+      "learning_rate": 1.4909932476936067e-06,
+      "loss": 0.7012,
+      "step": 11953
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.44828583728488,
+      "learning_rate": 1.4909090841285454e-06,
+      "loss": 0.6784,
+      "step": 11954
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.4025181082084,
+      "learning_rate": 1.4908249159817864e-06,
+      "loss": 0.6823,
+      "step": 11955
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.441849385399167,
+      "learning_rate": 1.4907407432541152e-06,
+      "loss": 0.6497,
+      "step": 11956
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.471638948273353,
+      "learning_rate": 1.490656565946317e-06,
+      "loss": 0.6387,
+      "step": 11957
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 4.2725654746421124,
+      "learning_rate": 1.4905723840591777e-06,
+      "loss": 0.6823,
+      "step": 11958
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.495426642285769,
+      "learning_rate": 1.4904881975934832e-06,
+      "loss": 0.6608,
+      "step": 11959
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9534433377114144,
+      "learning_rate": 1.4904040065500186e-06,
+      "loss": 0.6595,
+      "step": 11960
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.8248139529308183,
+      "learning_rate": 1.4903198109295702e-06,
+      "loss": 0.6562,
+      "step": 11961
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 5.138833185792895,
+      "learning_rate": 1.4902356107329238e-06,
+      "loss": 0.6732,
+      "step": 11962
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7561814256683792,
+      "learning_rate": 1.4901514059608646e-06,
+      "loss": 0.6589,
+      "step": 11963
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.191512519726874,
+      "learning_rate": 1.4900671966141793e-06,
+      "loss": 0.6732,
+      "step": 11964
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5824661279980774,
+      "learning_rate": 1.489982982693654e-06,
+      "loss": 0.668,
+      "step": 11965
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.429381395038681,
+      "learning_rate": 1.4898987642000735e-06,
+      "loss": 0.6986,
+      "step": 11966
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.0805610330141326,
+      "learning_rate": 1.4898145411342247e-06,
+      "loss": 0.6562,
+      "step": 11967
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.86143670949728,
+      "learning_rate": 1.4897303134968935e-06,
+      "loss": 0.6862,
+      "step": 11968
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3826229267521497,
+      "learning_rate": 1.489646081288866e-06,
+      "loss": 0.6934,
+      "step": 11969
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8718093169285492,
+      "learning_rate": 1.4895618445109283e-06,
+      "loss": 0.6725,
+      "step": 11970
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.9806540532126193,
+      "learning_rate": 1.4894776031638666e-06,
+      "loss": 0.6589,
+      "step": 11971
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.874363518201063,
+      "learning_rate": 1.4893933572484674e-06,
+      "loss": 0.6615,
+      "step": 11972
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.9758600283997585,
+      "learning_rate": 1.489309106765516e-06,
+      "loss": 0.6693,
+      "step": 11973
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.327691358223645,
+      "learning_rate": 1.4892248517158003e-06,
+      "loss": 0.6914,
+      "step": 11974
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8747140248117506,
+      "learning_rate": 1.4891405921001052e-06,
+      "loss": 0.6595,
+      "step": 11975
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9919055912523779,
+      "learning_rate": 1.4890563279192179e-06,
+      "loss": 0.6803,
+      "step": 11976
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.0072851791503985,
+      "learning_rate": 1.4889720591739247e-06,
+      "loss": 0.6823,
+      "step": 11977
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.275412461617239,
+      "learning_rate": 1.488887785865012e-06,
+      "loss": 0.6934,
+      "step": 11978
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.568234150153362,
+      "learning_rate": 1.4888035079932664e-06,
+      "loss": 0.6478,
+      "step": 11979
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.936525567173268,
+      "learning_rate": 1.4887192255594744e-06,
+      "loss": 0.6823,
+      "step": 11980
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4300932612329162,
+      "learning_rate": 1.4886349385644227e-06,
+      "loss": 0.6719,
+      "step": 11981
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.1260829054688455,
+      "learning_rate": 1.4885506470088977e-06,
+      "loss": 0.6621,
+      "step": 11982
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9508405700665794,
+      "learning_rate": 1.4884663508936864e-06,
+      "loss": 0.6745,
+      "step": 11983
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8519669406682906,
+      "learning_rate": 1.4883820502195754e-06,
+      "loss": 0.6654,
+      "step": 11984
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4938402470553358,
+      "learning_rate": 1.4882977449873517e-06,
+      "loss": 0.6608,
+      "step": 11985
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.8335153563970676,
+      "learning_rate": 1.4882134351978018e-06,
+      "loss": 0.6589,
+      "step": 11986
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.912249496261774,
+      "learning_rate": 1.4881291208517128e-06,
+      "loss": 0.6953,
+      "step": 11987
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.2411814978501865,
+      "learning_rate": 1.4880448019498716e-06,
+      "loss": 0.6699,
+      "step": 11988
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9393804363021436,
+      "learning_rate": 1.4879604784930648e-06,
+      "loss": 0.6589,
+      "step": 11989
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.33063837628838,
+      "learning_rate": 1.48787615048208e-06,
+      "loss": 0.6322,
+      "step": 11990
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.6698813749368258,
+      "learning_rate": 1.4877918179177043e-06,
+      "loss": 0.6608,
+      "step": 11991
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.007833272815319,
+      "learning_rate": 1.4877074808007237e-06,
+      "loss": 0.6725,
+      "step": 11992
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.079409185802692,
+      "learning_rate": 1.4876231391319266e-06,
+      "loss": 0.6862,
+      "step": 11993
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.522861984935972,
+      "learning_rate": 1.4875387929120993e-06,
+      "loss": 0.6966,
+      "step": 11994
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.665493787381403,
+      "learning_rate": 1.4874544421420295e-06,
+      "loss": 0.6758,
+      "step": 11995
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9832037110236159,
+      "learning_rate": 1.487370086822504e-06,
+      "loss": 0.6719,
+      "step": 11996
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.745175891071176,
+      "learning_rate": 1.4872857269543103e-06,
+      "loss": 0.6504,
+      "step": 11997
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.9569971894953369,
+      "learning_rate": 1.4872013625382365e-06,
+      "loss": 0.6536,
+      "step": 11998
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.091938850330173,
+      "learning_rate": 1.4871169935750687e-06,
+      "loss": 0.6738,
+      "step": 11999
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.373537300571694,
+      "learning_rate": 1.4870326200655954e-06,
+      "loss": 0.653,
+      "step": 12000
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.240364764317561,
+      "learning_rate": 1.4869482420106031e-06,
+      "loss": 0.6497,
+      "step": 12001
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.4210250605384394,
+      "learning_rate": 1.48686385941088e-06,
+      "loss": 0.6589,
+      "step": 12002
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.674576143264329,
+      "learning_rate": 1.4867794722672137e-06,
+      "loss": 0.6211,
+      "step": 12003
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 5.028672537297378,
+      "learning_rate": 1.4866950805803912e-06,
+      "loss": 0.6927,
+      "step": 12004
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.240410854456695,
+      "learning_rate": 1.4866106843512003e-06,
+      "loss": 0.6491,
+      "step": 12005
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.129549207176684,
+      "learning_rate": 1.4865262835804294e-06,
+      "loss": 0.6439,
+      "step": 12006
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.201149173830624,
+      "learning_rate": 1.4864418782688651e-06,
+      "loss": 0.6875,
+      "step": 12007
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.234432726050111,
+      "learning_rate": 1.4863574684172962e-06,
+      "loss": 0.6777,
+      "step": 12008
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.5295374231769885,
+      "learning_rate": 1.4862730540265097e-06,
+      "loss": 0.6934,
+      "step": 12009
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.136019167186263,
+      "learning_rate": 1.4861886350972938e-06,
+      "loss": 0.6458,
+      "step": 12010
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.494628698553357,
+      "learning_rate": 1.4861042116304369e-06,
+      "loss": 0.7057,
+      "step": 12011
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.2393610969109363,
+      "learning_rate": 1.486019783626726e-06,
+      "loss": 0.6543,
+      "step": 12012
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.4196667334079556,
+      "learning_rate": 1.4859353510869493e-06,
+      "loss": 0.6497,
+      "step": 12013
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.367693736681806,
+      "learning_rate": 1.4858509140118954e-06,
+      "loss": 0.668,
+      "step": 12014
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.160136743689239,
+      "learning_rate": 1.4857664724023516e-06,
+      "loss": 0.6816,
+      "step": 12015
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.6613262618526314,
+      "learning_rate": 1.4856820262591064e-06,
+      "loss": 0.6556,
+      "step": 12016
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.185994143700659,
+      "learning_rate": 1.485597575582948e-06,
+      "loss": 0.6764,
+      "step": 12017
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.0800600741311772,
+      "learning_rate": 1.4855131203746644e-06,
+      "loss": 0.6725,
+      "step": 12018
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.0278047482699126,
+      "learning_rate": 1.485428660635044e-06,
+      "loss": 0.6745,
+      "step": 12019
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.3849303469875056,
+      "learning_rate": 1.4853441963648747e-06,
+      "loss": 0.6699,
+      "step": 12020
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.387428588705207,
+      "learning_rate": 1.4852597275649455e-06,
+      "loss": 0.6576,
+      "step": 12021
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.9308607650350296,
+      "learning_rate": 1.4851752542360444e-06,
+      "loss": 0.6966,
+      "step": 12022
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 6.225661596601007,
+      "learning_rate": 1.4850907763789593e-06,
+      "loss": 0.6491,
+      "step": 12023
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.2976335390351637,
+      "learning_rate": 1.4850062939944794e-06,
+      "loss": 0.6654,
+      "step": 12024
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.4377949146033235,
+      "learning_rate": 1.484921807083393e-06,
+      "loss": 0.6582,
+      "step": 12025
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7870910314415351,
+      "learning_rate": 1.484837315646488e-06,
+      "loss": 0.6875,
+      "step": 12026
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.4966881962542473,
+      "learning_rate": 1.484752819684554e-06,
+      "loss": 0.6719,
+      "step": 12027
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3562010912854268,
+      "learning_rate": 1.4846683191983786e-06,
+      "loss": 0.6823,
+      "step": 12028
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.2469826365851087,
+      "learning_rate": 1.484583814188751e-06,
+      "loss": 0.6953,
+      "step": 12029
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.057389720129214,
+      "learning_rate": 1.48449930465646e-06,
+      "loss": 0.6465,
+      "step": 12030
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.3605248765154725,
+      "learning_rate": 1.4844147906022942e-06,
+      "loss": 0.6504,
+      "step": 12031
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.1283479630993243,
+      "learning_rate": 1.484330272027042e-06,
+      "loss": 0.7005,
+      "step": 12032
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 4.198378825586257,
+      "learning_rate": 1.4842457489314928e-06,
+      "loss": 0.6641,
+      "step": 12033
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.535630108056997,
+      "learning_rate": 1.4841612213164352e-06,
+      "loss": 0.6771,
+      "step": 12034
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.1918674633595243,
+      "learning_rate": 1.484076689182658e-06,
+      "loss": 0.6882,
+      "step": 12035
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6923248449950044,
+      "learning_rate": 1.4839921525309501e-06,
+      "loss": 0.6569,
+      "step": 12036
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7263510436763516,
+      "learning_rate": 1.483907611362101e-06,
+      "loss": 0.6647,
+      "step": 12037
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.527591517269369,
+      "learning_rate": 1.483823065676899e-06,
+      "loss": 0.6901,
+      "step": 12038
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1748680736652997,
+      "learning_rate": 1.4837385154761339e-06,
+      "loss": 0.6309,
+      "step": 12039
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.211742313624715,
+      "learning_rate": 1.4836539607605942e-06,
+      "loss": 0.7161,
+      "step": 12040
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.2586307455049432,
+      "learning_rate": 1.4835694015310695e-06,
+      "loss": 0.6654,
+      "step": 12041
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1560704310785046,
+      "learning_rate": 1.4834848377883486e-06,
+      "loss": 0.6862,
+      "step": 12042
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.090365496487219,
+      "learning_rate": 1.483400269533221e-06,
+      "loss": 0.6927,
+      "step": 12043
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.769488322022434,
+      "learning_rate": 1.4833156967664761e-06,
+      "loss": 0.7031,
+      "step": 12044
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.4768180923360186,
+      "learning_rate": 1.4832311194889031e-06,
+      "loss": 0.6901,
+      "step": 12045
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.344699181391958,
+      "learning_rate": 1.483146537701291e-06,
+      "loss": 0.6471,
+      "step": 12046
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9860854267011161,
+      "learning_rate": 1.4830619514044298e-06,
+      "loss": 0.6751,
+      "step": 12047
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7822320815597124,
+      "learning_rate": 1.4829773605991088e-06,
+      "loss": 0.6816,
+      "step": 12048
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.862195790069949,
+      "learning_rate": 1.4828927652861169e-06,
+      "loss": 0.6523,
+      "step": 12049
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.3209114882052007,
+      "learning_rate": 1.4828081654662445e-06,
+      "loss": 0.6797,
+      "step": 12050
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6449368236549404,
+      "learning_rate": 1.4827235611402808e-06,
+      "loss": 0.6497,
+      "step": 12051
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.21767672022918,
+      "learning_rate": 1.4826389523090153e-06,
+      "loss": 0.6602,
+      "step": 12052
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.2069514686835694,
+      "learning_rate": 1.4825543389732378e-06,
+      "loss": 0.6764,
+      "step": 12053
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.3727050722865604,
+      "learning_rate": 1.482469721133738e-06,
+      "loss": 0.6283,
+      "step": 12054
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.5999451010325103,
+      "learning_rate": 1.4823850987913053e-06,
+      "loss": 0.6836,
+      "step": 12055
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.21899868413518,
+      "learning_rate": 1.4823004719467303e-06,
+      "loss": 0.6471,
+      "step": 12056
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4413765907761096,
+      "learning_rate": 1.482215840600802e-06,
+      "loss": 0.6556,
+      "step": 12057
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.0481492200495865,
+      "learning_rate": 1.4821312047543107e-06,
+      "loss": 0.6745,
+      "step": 12058
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.0714746383232523,
+      "learning_rate": 1.4820465644080464e-06,
+      "loss": 0.6882,
+      "step": 12059
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.379094100126444,
+      "learning_rate": 1.4819619195627987e-06,
+      "loss": 0.7298,
+      "step": 12060
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.975051649888666,
+      "learning_rate": 1.4818772702193581e-06,
+      "loss": 0.6393,
+      "step": 12061
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.147403876144402,
+      "learning_rate": 1.4817926163785139e-06,
+      "loss": 0.7005,
+      "step": 12062
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.624780684416257,
+      "learning_rate": 1.4817079580410567e-06,
+      "loss": 0.6413,
+      "step": 12063
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.2053974305261383,
+      "learning_rate": 1.4816232952077768e-06,
+      "loss": 0.6868,
+      "step": 12064
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.3298791901292235,
+      "learning_rate": 1.4815386278794639e-06,
+      "loss": 0.6641,
+      "step": 12065
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0322979137397708,
+      "learning_rate": 1.4814539560569083e-06,
+      "loss": 0.6927,
+      "step": 12066
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.1750476045358127,
+      "learning_rate": 1.4813692797409005e-06,
+      "loss": 0.6654,
+      "step": 12067
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.4525021185984284,
+      "learning_rate": 1.4812845989322304e-06,
+      "loss": 0.6738,
+      "step": 12068
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.079207544803609,
+      "learning_rate": 1.4811999136316891e-06,
+      "loss": 0.6901,
+      "step": 12069
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.74949297989608,
+      "learning_rate": 1.4811152238400658e-06,
+      "loss": 0.696,
+      "step": 12070
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6383091072598408,
+      "learning_rate": 1.481030529558152e-06,
+      "loss": 0.6947,
+      "step": 12071
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0153805356595655,
+      "learning_rate": 1.4809458307867376e-06,
+      "loss": 0.6849,
+      "step": 12072
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0240033174352305,
+      "learning_rate": 1.480861127526613e-06,
+      "loss": 0.6751,
+      "step": 12073
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7785613836231444,
+      "learning_rate": 1.4807764197785696e-06,
+      "loss": 0.6647,
+      "step": 12074
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4210120584879078,
+      "learning_rate": 1.480691707543397e-06,
+      "loss": 0.6641,
+      "step": 12075
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.254986962675533,
+      "learning_rate": 1.4806069908218858e-06,
+      "loss": 0.6465,
+      "step": 12076
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.734806698445326,
+      "learning_rate": 1.4805222696148278e-06,
+      "loss": 0.6413,
+      "step": 12077
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 7.128206295877615,
+      "learning_rate": 1.4804375439230126e-06,
+      "loss": 0.709,
+      "step": 12078
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8618183135244816,
+      "learning_rate": 1.480352813747231e-06,
+      "loss": 0.6589,
+      "step": 12079
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.073158137652573,
+      "learning_rate": 1.4802680790882746e-06,
+      "loss": 0.6914,
+      "step": 12080
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.782532061526765,
+      "learning_rate": 1.4801833399469335e-06,
+      "loss": 0.6556,
+      "step": 12081
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.2069460998011066,
+      "learning_rate": 1.480098596323999e-06,
+      "loss": 0.6829,
+      "step": 12082
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 8.135164883892065,
+      "learning_rate": 1.4800138482202615e-06,
+      "loss": 0.6693,
+      "step": 12083
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.7114235366014676,
+      "learning_rate": 1.4799290956365126e-06,
+      "loss": 0.6712,
+      "step": 12084
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6565517245006345,
+      "learning_rate": 1.4798443385735429e-06,
+      "loss": 0.6699,
+      "step": 12085
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.8630527915489836,
+      "learning_rate": 1.4797595770321434e-06,
+      "loss": 0.666,
+      "step": 12086
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0392027668948636,
+      "learning_rate": 1.4796748110131057e-06,
+      "loss": 0.6699,
+      "step": 12087
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.841498083050827,
+      "learning_rate": 1.4795900405172204e-06,
+      "loss": 0.6491,
+      "step": 12088
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0568916610905608,
+      "learning_rate": 1.4795052655452786e-06,
+      "loss": 0.6589,
+      "step": 12089
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0238291206712873,
+      "learning_rate": 1.4794204860980722e-06,
+      "loss": 0.6771,
+      "step": 12090
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0105076454482964,
+      "learning_rate": 1.4793357021763917e-06,
+      "loss": 0.6517,
+      "step": 12091
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9866690989813218,
+      "learning_rate": 1.4792509137810286e-06,
+      "loss": 0.6602,
+      "step": 12092
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.671781150564738,
+      "learning_rate": 1.4791661209127749e-06,
+      "loss": 0.6712,
+      "step": 12093
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 4.357963706075764,
+      "learning_rate": 1.479081323572421e-06,
+      "loss": 0.6829,
+      "step": 12094
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.73719391401932,
+      "learning_rate": 1.4789965217607585e-06,
+      "loss": 0.6953,
+      "step": 12095
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.786371495110824,
+      "learning_rate": 1.4789117154785795e-06,
+      "loss": 0.6589,
+      "step": 12096
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 3.514759137973322,
+      "learning_rate": 1.478826904726675e-06,
+      "loss": 0.6549,
+      "step": 12097
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.5282746260113123,
+      "learning_rate": 1.4787420895058367e-06,
+      "loss": 0.6589,
+      "step": 12098
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.3984319703028918,
+      "learning_rate": 1.478657269816856e-06,
+      "loss": 0.6413,
+      "step": 12099
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.4458211801538416,
+      "learning_rate": 1.4785724456605246e-06,
+      "loss": 0.6647,
+      "step": 12100
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.9183751410198273,
+      "learning_rate": 1.4784876170376345e-06,
+      "loss": 0.6699,
+      "step": 12101
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.0312018201784707,
+      "learning_rate": 1.478402783948977e-06,
+      "loss": 0.6634,
+      "step": 12102
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.0991990357266115,
+      "learning_rate": 1.4783179463953441e-06,
+      "loss": 0.6647,
+      "step": 12103
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 7.253145165706257,
+      "learning_rate": 1.4782331043775276e-06,
+      "loss": 0.7005,
+      "step": 12104
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.20128048698092,
+      "learning_rate": 1.478148257896319e-06,
+      "loss": 0.6849,
+      "step": 12105
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.7260406005823885,
+      "learning_rate": 1.4780634069525105e-06,
+      "loss": 0.6465,
+      "step": 12106
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.436791249687259,
+      "learning_rate": 1.4779785515468941e-06,
+      "loss": 0.707,
+      "step": 12107
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.454326363683667,
+      "learning_rate": 1.4778936916802614e-06,
+      "loss": 0.6934,
+      "step": 12108
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.255444693185455,
+      "learning_rate": 1.4778088273534046e-06,
+      "loss": 0.694,
+      "step": 12109
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.2973159525892775,
+      "learning_rate": 1.477723958567116e-06,
+      "loss": 0.709,
+      "step": 12110
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.921759707065492,
+      "learning_rate": 1.4776390853221875e-06,
+      "loss": 0.6569,
+      "step": 12111
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5795433265886647,
+      "learning_rate": 1.4775542076194107e-06,
+      "loss": 0.7077,
+      "step": 12112
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4418085001371916,
+      "learning_rate": 1.4774693254595785e-06,
+      "loss": 0.6901,
+      "step": 12113
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.863198017597598,
+      "learning_rate": 1.4773844388434832e-06,
+      "loss": 0.6777,
+      "step": 12114
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.945930710722776,
+      "learning_rate": 1.4772995477719161e-06,
+      "loss": 0.6569,
+      "step": 12115
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.006512964537014,
+      "learning_rate": 1.4772146522456705e-06,
+      "loss": 0.7096,
+      "step": 12116
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.0270345653627104,
+      "learning_rate": 1.4771297522655384e-06,
+      "loss": 0.6315,
+      "step": 12117
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.7565337862653394,
+      "learning_rate": 1.4770448478323117e-06,
+      "loss": 0.6855,
+      "step": 12118
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.1650199410588633,
+      "learning_rate": 1.4769599389467837e-06,
+      "loss": 0.6836,
+      "step": 12119
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.389694949563129,
+      "learning_rate": 1.4768750256097461e-06,
+      "loss": 0.6908,
+      "step": 12120
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.9768923282600275,
+      "learning_rate": 1.476790107821992e-06,
+      "loss": 0.696,
+      "step": 12121
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.365769173749654,
+      "learning_rate": 1.4767051855843137e-06,
+      "loss": 0.6921,
+      "step": 12122
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9609206914394944,
+      "learning_rate": 1.4766202588975034e-06,
+      "loss": 0.6797,
+      "step": 12123
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.9211501134061997,
+      "learning_rate": 1.4765353277623541e-06,
+      "loss": 0.6764,
+      "step": 12124
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9073307480461267,
+      "learning_rate": 1.4764503921796587e-06,
+      "loss": 0.666,
+      "step": 12125
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.3909902388208037,
+      "learning_rate": 1.4763654521502096e-06,
+      "loss": 0.6882,
+      "step": 12126
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5797220300510944,
+      "learning_rate": 1.4762805076747995e-06,
+      "loss": 0.6953,
+      "step": 12127
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.417801973574752,
+      "learning_rate": 1.4761955587542211e-06,
+      "loss": 0.6595,
+      "step": 12128
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.799895791340954,
+      "learning_rate": 1.4761106053892675e-06,
+      "loss": 0.6764,
+      "step": 12129
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.697266161024543,
+      "learning_rate": 1.476025647580732e-06,
+      "loss": 0.6647,
+      "step": 12130
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.78821732819301,
+      "learning_rate": 1.4759406853294067e-06,
+      "loss": 0.6556,
+      "step": 12131
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.397227737310146,
+      "learning_rate": 1.475855718636085e-06,
+      "loss": 0.6784,
+      "step": 12132
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.390438664213853,
+      "learning_rate": 1.4757707475015598e-06,
+      "loss": 0.6758,
+      "step": 12133
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.904876288192505,
+      "learning_rate": 1.4756857719266238e-06,
+      "loss": 0.6725,
+      "step": 12134
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7405178273408124,
+      "learning_rate": 1.4756007919120708e-06,
+      "loss": 0.6628,
+      "step": 12135
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.718702543480057,
+      "learning_rate": 1.4755158074586935e-06,
+      "loss": 0.6641,
+      "step": 12136
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.260811776941801,
+      "learning_rate": 1.4754308185672847e-06,
+      "loss": 0.6914,
+      "step": 12137
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.3877057513181206,
+      "learning_rate": 1.4753458252386386e-06,
+      "loss": 0.6771,
+      "step": 12138
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2036072382111562,
+      "learning_rate": 1.4752608274735475e-06,
+      "loss": 0.6543,
+      "step": 12139
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1774586495621202,
+      "learning_rate": 1.4751758252728053e-06,
+      "loss": 0.666,
+      "step": 12140
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7635173276552129,
+      "learning_rate": 1.4750908186372051e-06,
+      "loss": 0.6732,
+      "step": 12141
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.044479602441556,
+      "learning_rate": 1.47500580756754e-06,
+      "loss": 0.6751,
+      "step": 12142
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.871094779639037,
+      "learning_rate": 1.4749207920646041e-06,
+      "loss": 0.6914,
+      "step": 12143
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.125580134644548,
+      "learning_rate": 1.47483577212919e-06,
+      "loss": 0.6602,
+      "step": 12144
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.839178627595194,
+      "learning_rate": 1.474750747762092e-06,
+      "loss": 0.666,
+      "step": 12145
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.549270324803005,
+      "learning_rate": 1.4746657189641033e-06,
+      "loss": 0.6849,
+      "step": 12146
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.348687802285327,
+      "learning_rate": 1.4745806857360172e-06,
+      "loss": 0.6947,
+      "step": 12147
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.9772660671053742,
+      "learning_rate": 1.4744956480786277e-06,
+      "loss": 0.6556,
+      "step": 12148
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.860417906176917,
+      "learning_rate": 1.4744106059927284e-06,
+      "loss": 0.6745,
+      "step": 12149
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.9137155883275994,
+      "learning_rate": 1.4743255594791128e-06,
+      "loss": 0.6732,
+      "step": 12150
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.3580256870255814,
+      "learning_rate": 1.474240508538575e-06,
+      "loss": 0.6426,
+      "step": 12151
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.912440459094422,
+      "learning_rate": 1.4741554531719084e-06,
+      "loss": 0.6914,
+      "step": 12152
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.2345847423108234,
+      "learning_rate": 1.4740703933799073e-06,
+      "loss": 0.6452,
+      "step": 12153
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.3879155747501093,
+      "learning_rate": 1.4739853291633654e-06,
+      "loss": 0.6556,
+      "step": 12154
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.7484793168821167,
+      "learning_rate": 1.473900260523076e-06,
+      "loss": 0.6816,
+      "step": 12155
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4095066919642023,
+      "learning_rate": 1.473815187459834e-06,
+      "loss": 0.6738,
+      "step": 12156
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.059152299539841,
+      "learning_rate": 1.4737301099744327e-06,
+      "loss": 0.6569,
+      "step": 12157
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6900554464715765,
+      "learning_rate": 1.4736450280676663e-06,
+      "loss": 0.6628,
+      "step": 12158
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1089321977601845,
+      "learning_rate": 1.473559941740329e-06,
+      "loss": 0.6777,
+      "step": 12159
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.8425282223862016,
+      "learning_rate": 1.4734748509932154e-06,
+      "loss": 0.6771,
+      "step": 12160
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.5144857769478457,
+      "learning_rate": 1.4733897558271185e-06,
+      "loss": 0.6615,
+      "step": 12161
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.789872250728702,
+      "learning_rate": 1.4733046562428334e-06,
+      "loss": 0.6895,
+      "step": 12162
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 5.801981314157978,
+      "learning_rate": 1.4732195522411542e-06,
+      "loss": 0.6953,
+      "step": 12163
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.081616219948454,
+      "learning_rate": 1.473134443822875e-06,
+      "loss": 0.6725,
+      "step": 12164
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.0539212426823057,
+      "learning_rate": 1.47304933098879e-06,
+      "loss": 0.6452,
+      "step": 12165
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1309284702340994,
+      "learning_rate": 1.472964213739694e-06,
+      "loss": 0.6738,
+      "step": 12166
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.8918748861272716,
+      "learning_rate": 1.4728790920763809e-06,
+      "loss": 0.6738,
+      "step": 12167
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.027236049848507,
+      "learning_rate": 1.4727939659996456e-06,
+      "loss": 0.6419,
+      "step": 12168
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.6322579716937002,
+      "learning_rate": 1.4727088355102824e-06,
+      "loss": 0.6712,
+      "step": 12169
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.172044023820262,
+      "learning_rate": 1.4726237006090858e-06,
+      "loss": 0.6602,
+      "step": 12170
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1857720999610357,
+      "learning_rate": 1.4725385612968504e-06,
+      "loss": 0.6387,
+      "step": 12171
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.910893105954159,
+      "learning_rate": 1.4724534175743707e-06,
+      "loss": 0.6973,
+      "step": 12172
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0302188312419693,
+      "learning_rate": 1.472368269442442e-06,
+      "loss": 0.6602,
+      "step": 12173
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.3499746861111688,
+      "learning_rate": 1.4722831169018578e-06,
+      "loss": 0.6771,
+      "step": 12174
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.228667406531583,
+      "learning_rate": 1.472197959953414e-06,
+      "loss": 0.651,
+      "step": 12175
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.2494666653183666,
+      "learning_rate": 1.4721127985979048e-06,
+      "loss": 0.6855,
+      "step": 12176
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0256695609052624,
+      "learning_rate": 1.472027632836125e-06,
+      "loss": 0.6751,
+      "step": 12177
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.2686818902262575,
+      "learning_rate": 1.4719424626688692e-06,
+      "loss": 0.6901,
+      "step": 12178
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7638420662760224,
+      "learning_rate": 1.471857288096933e-06,
+      "loss": 0.6484,
+      "step": 12179
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.8902161731991245,
+      "learning_rate": 1.471772109121111e-06,
+      "loss": 0.6517,
+      "step": 12180
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0971918533759857,
+      "learning_rate": 1.4716869257421981e-06,
+      "loss": 0.6771,
+      "step": 12181
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2641619467238285,
+      "learning_rate": 1.4716017379609897e-06,
+      "loss": 0.6589,
+      "step": 12182
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.617203794901538,
+      "learning_rate": 1.4715165457782803e-06,
+      "loss": 0.625,
+      "step": 12183
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.855183586402882,
+      "learning_rate": 1.4714313491948653e-06,
+      "loss": 0.6849,
+      "step": 12184
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.176997201596278,
+      "learning_rate": 1.4713461482115398e-06,
+      "loss": 0.6582,
+      "step": 12185
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8390437711344654,
+      "learning_rate": 1.4712609428290991e-06,
+      "loss": 0.6654,
+      "step": 12186
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.4989655468828778,
+      "learning_rate": 1.4711757330483382e-06,
+      "loss": 0.6549,
+      "step": 12187
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.4311979515035345,
+      "learning_rate": 1.4710905188700529e-06,
+      "loss": 0.6745,
+      "step": 12188
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.853219259148362,
+      "learning_rate": 1.4710053002950378e-06,
+      "loss": 0.6283,
+      "step": 12189
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.9794625230196337,
+      "learning_rate": 1.4709200773240884e-06,
+      "loss": 0.6836,
+      "step": 12190
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.159720281121457,
+      "learning_rate": 1.4708348499580007e-06,
+      "loss": 0.696,
+      "step": 12191
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.5940219423748354,
+      "learning_rate": 1.4707496181975693e-06,
+      "loss": 0.668,
+      "step": 12192
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.130476090101031,
+      "learning_rate": 1.4706643820435906e-06,
+      "loss": 0.6921,
+      "step": 12193
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.3911841794676711,
+      "learning_rate": 1.4705791414968591e-06,
+      "loss": 0.7044,
+      "step": 12194
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8973876494332793,
+      "learning_rate": 1.470493896558171e-06,
+      "loss": 0.6309,
+      "step": 12195
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.0362471854955593,
+      "learning_rate": 1.4704086472283218e-06,
+      "loss": 0.6712,
+      "step": 12196
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.3840373410360005,
+      "learning_rate": 1.470323393508107e-06,
+      "loss": 0.709,
+      "step": 12197
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9443160364712853,
+      "learning_rate": 1.4702381353983228e-06,
+      "loss": 0.6549,
+      "step": 12198
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1367359528477212,
+      "learning_rate": 1.470152872899764e-06,
+      "loss": 0.6986,
+      "step": 12199
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9084199961328143,
+      "learning_rate": 1.4700676060132272e-06,
+      "loss": 0.7044,
+      "step": 12200
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.66145692107494,
+      "learning_rate": 1.4699823347395078e-06,
+      "loss": 0.6582,
+      "step": 12201
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9148642627878733,
+      "learning_rate": 1.4698970590794016e-06,
+      "loss": 0.679,
+      "step": 12202
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6634795325880574,
+      "learning_rate": 1.4698117790337044e-06,
+      "loss": 0.6536,
+      "step": 12203
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9173252246704615,
+      "learning_rate": 1.4697264946032127e-06,
+      "loss": 0.6771,
+      "step": 12204
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.3895436310578657,
+      "learning_rate": 1.4696412057887217e-06,
+      "loss": 0.6771,
+      "step": 12205
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1138592939688303,
+      "learning_rate": 1.4695559125910279e-06,
+      "loss": 0.6855,
+      "step": 12206
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.334869881853277,
+      "learning_rate": 1.4694706150109275e-06,
+      "loss": 0.6979,
+      "step": 12207
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.907297350920222,
+      "learning_rate": 1.469385313049216e-06,
+      "loss": 0.653,
+      "step": 12208
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 6.221501268253955,
+      "learning_rate": 1.4693000067066903e-06,
+      "loss": 0.679,
+      "step": 12209
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8822565425947434,
+      "learning_rate": 1.4692146959841459e-06,
+      "loss": 0.6595,
+      "step": 12210
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.3576131880279723,
+      "learning_rate": 1.469129380882379e-06,
+      "loss": 0.6738,
+      "step": 12211
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.525304043654753,
+      "learning_rate": 1.4690440614021864e-06,
+      "loss": 0.6445,
+      "step": 12212
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.941410628340744,
+      "learning_rate": 1.468958737544364e-06,
+      "loss": 0.6504,
+      "step": 12213
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9968948514294721,
+      "learning_rate": 1.468873409309708e-06,
+      "loss": 0.6829,
+      "step": 12214
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1332532367814068,
+      "learning_rate": 1.4687880766990152e-06,
+      "loss": 0.6816,
+      "step": 12215
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.407609600299468,
+      "learning_rate": 1.468702739713082e-06,
+      "loss": 0.6634,
+      "step": 12216
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.3662197913481284,
+      "learning_rate": 1.4686173983527045e-06,
+      "loss": 0.6953,
+      "step": 12217
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 5.959049000203143,
+      "learning_rate": 1.4685320526186794e-06,
+      "loss": 0.6706,
+      "step": 12218
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2558622971623978,
+      "learning_rate": 1.4684467025118036e-06,
+      "loss": 0.6784,
+      "step": 12219
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.314623053860325,
+      "learning_rate": 1.4683613480328729e-06,
+      "loss": 0.6816,
+      "step": 12220
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1621680219816615,
+      "learning_rate": 1.468275989182684e-06,
+      "loss": 0.6751,
+      "step": 12221
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.909326765735582,
+      "learning_rate": 1.4681906259620346e-06,
+      "loss": 0.6667,
+      "step": 12222
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.829617225154419,
+      "learning_rate": 1.4681052583717202e-06,
+      "loss": 0.6888,
+      "step": 12223
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.9059988359683264,
+      "learning_rate": 1.468019886412538e-06,
+      "loss": 0.6816,
+      "step": 12224
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.7397905442518535,
+      "learning_rate": 1.4679345100852852e-06,
+      "loss": 0.6882,
+      "step": 12225
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2366698028507426,
+      "learning_rate": 1.4678491293907579e-06,
+      "loss": 0.6517,
+      "step": 12226
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8534758527083415,
+      "learning_rate": 1.4677637443297532e-06,
+      "loss": 0.6693,
+      "step": 12227
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.9668255120387312,
+      "learning_rate": 1.4676783549030684e-06,
+      "loss": 0.653,
+      "step": 12228
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.071400541038119,
+      "learning_rate": 1.4675929611115002e-06,
+      "loss": 0.6693,
+      "step": 12229
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.11342802259629,
+      "learning_rate": 1.4675075629558454e-06,
+      "loss": 0.6712,
+      "step": 12230
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.1589209291258762,
+      "learning_rate": 1.4674221604369011e-06,
+      "loss": 0.6784,
+      "step": 12231
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.009006944092179,
+      "learning_rate": 1.4673367535554647e-06,
+      "loss": 0.6491,
+      "step": 12232
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.633927288012724,
+      "learning_rate": 1.4672513423123328e-06,
+      "loss": 0.6641,
+      "step": 12233
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.254568011408006,
+      "learning_rate": 1.4671659267083027e-06,
+      "loss": 0.6771,
+      "step": 12234
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.2954071766926245,
+      "learning_rate": 1.467080506744172e-06,
+      "loss": 0.6908,
+      "step": 12235
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.9167464565078394,
+      "learning_rate": 1.4669950824207376e-06,
+      "loss": 0.6758,
+      "step": 12236
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.6500327236021017,
+      "learning_rate": 1.4669096537387964e-06,
+      "loss": 0.6706,
+      "step": 12237
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.198619315240072,
+      "learning_rate": 1.4668242206991464e-06,
+      "loss": 0.6283,
+      "step": 12238
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.9789931771083307,
+      "learning_rate": 1.4667387833025848e-06,
+      "loss": 0.653,
+      "step": 12239
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.0864853020216017,
+      "learning_rate": 1.4666533415499085e-06,
+      "loss": 0.6667,
+      "step": 12240
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0736077071093415,
+      "learning_rate": 1.4665678954419157e-06,
+      "loss": 0.6751,
+      "step": 12241
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.831118674920704,
+      "learning_rate": 1.4664824449794031e-06,
+      "loss": 0.7103,
+      "step": 12242
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.2753920768910434,
+      "learning_rate": 1.466396990163169e-06,
+      "loss": 0.6855,
+      "step": 12243
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.8362627683924595,
+      "learning_rate": 1.4663115309940103e-06,
+      "loss": 0.6582,
+      "step": 12244
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.2584900440195037,
+      "learning_rate": 1.4662260674727247e-06,
+      "loss": 0.6654,
+      "step": 12245
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.2374137859759051,
+      "learning_rate": 1.4661405996001102e-06,
+      "loss": 0.6771,
+      "step": 12246
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.1004868496447404,
+      "learning_rate": 1.4660551273769642e-06,
+      "loss": 0.6751,
+      "step": 12247
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.675647795311902,
+      "learning_rate": 1.4659696508040844e-06,
+      "loss": 0.6484,
+      "step": 12248
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.925490344192749,
+      "learning_rate": 1.4658841698822688e-06,
+      "loss": 0.6686,
+      "step": 12249
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.4638509049356916,
+      "learning_rate": 1.4657986846123147e-06,
+      "loss": 0.7116,
+      "step": 12250
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.4608492902062735,
+      "learning_rate": 1.4657131949950208e-06,
+      "loss": 0.668,
+      "step": 12251
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 5.236589086925721,
+      "learning_rate": 1.4656277010311842e-06,
+      "loss": 0.7005,
+      "step": 12252
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.9295723547627053,
+      "learning_rate": 1.4655422027216028e-06,
+      "loss": 0.6719,
+      "step": 12253
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.548901952606786,
+      "learning_rate": 1.4654567000670755e-06,
+      "loss": 0.6758,
+      "step": 12254
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.126201609893276,
+      "learning_rate": 1.4653711930683992e-06,
+      "loss": 0.6823,
+      "step": 12255
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.7035623455242157,
+      "learning_rate": 1.4652856817263723e-06,
+      "loss": 0.6745,
+      "step": 12256
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9090749107489317,
+      "learning_rate": 1.4652001660417934e-06,
+      "loss": 0.6901,
+      "step": 12257
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.8784216109637932,
+      "learning_rate": 1.4651146460154598e-06,
+      "loss": 0.64,
+      "step": 12258
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.6829338435562287,
+      "learning_rate": 1.4650291216481706e-06,
+      "loss": 0.6699,
+      "step": 12259
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.1140652024051234,
+      "learning_rate": 1.464943592940723e-06,
+      "loss": 0.6484,
+      "step": 12260
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0845389910890069,
+      "learning_rate": 1.4648580598939157e-06,
+      "loss": 0.6855,
+      "step": 12261
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.9531714655348753,
+      "learning_rate": 1.4647725225085474e-06,
+      "loss": 0.6777,
+      "step": 12262
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.839548262876379,
+      "learning_rate": 1.4646869807854159e-06,
+      "loss": 0.6816,
+      "step": 12263
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.607495674852962,
+      "learning_rate": 1.4646014347253194e-06,
+      "loss": 0.6621,
+      "step": 12264
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9102826487911456,
+      "learning_rate": 1.464515884329057e-06,
+      "loss": 0.6901,
+      "step": 12265
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.718181247025702,
+      "learning_rate": 1.4644303295974267e-06,
+      "loss": 0.6895,
+      "step": 12266
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.783370641702589,
+      "learning_rate": 1.464344770531227e-06,
+      "loss": 0.6719,
+      "step": 12267
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1337175361217111,
+      "learning_rate": 1.4642592071312563e-06,
+      "loss": 0.6484,
+      "step": 12268
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6965784994479907,
+      "learning_rate": 1.4641736393983137e-06,
+      "loss": 0.6549,
+      "step": 12269
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.5307486254807428,
+      "learning_rate": 1.4640880673331975e-06,
+      "loss": 0.6608,
+      "step": 12270
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.385572256152192,
+      "learning_rate": 1.464002490936706e-06,
+      "loss": 0.6608,
+      "step": 12271
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.5345397072655897,
+      "learning_rate": 1.4639169102096385e-06,
+      "loss": 0.7142,
+      "step": 12272
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.4013148321292728,
+      "learning_rate": 1.4638313251527934e-06,
+      "loss": 0.6686,
+      "step": 12273
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.9540194161266564,
+      "learning_rate": 1.4637457357669693e-06,
+      "loss": 0.694,
+      "step": 12274
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.5538915953240795,
+      "learning_rate": 1.4636601420529655e-06,
+      "loss": 0.6719,
+      "step": 12275
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.030981407556664,
+      "learning_rate": 1.4635745440115805e-06,
+      "loss": 0.6543,
+      "step": 12276
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.0259675044491035,
+      "learning_rate": 1.4634889416436132e-06,
+      "loss": 0.6595,
+      "step": 12277
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.447418166667284,
+      "learning_rate": 1.4634033349498628e-06,
+      "loss": 0.6673,
+      "step": 12278
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.58422669431328,
+      "learning_rate": 1.463317723931128e-06,
+      "loss": 0.6875,
+      "step": 12279
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.656233789357628,
+      "learning_rate": 1.4632321085882079e-06,
+      "loss": 0.6725,
+      "step": 12280
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.818181431534818,
+      "learning_rate": 1.4631464889219014e-06,
+      "loss": 0.6693,
+      "step": 12281
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0617684316645033,
+      "learning_rate": 1.4630608649330078e-06,
+      "loss": 0.6862,
+      "step": 12282
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.2336850506361867,
+      "learning_rate": 1.4629752366223265e-06,
+      "loss": 0.6523,
+      "step": 12283
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.88554347829679,
+      "learning_rate": 1.462889603990656e-06,
+      "loss": 0.6764,
+      "step": 12284
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.9712316489654642,
+      "learning_rate": 1.4628039670387961e-06,
+      "loss": 0.6862,
+      "step": 12285
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.035024065188161,
+      "learning_rate": 1.462718325767546e-06,
+      "loss": 0.6458,
+      "step": 12286
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1532072740614168,
+      "learning_rate": 1.4626326801777045e-06,
+      "loss": 0.6562,
+      "step": 12287
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.4141536231436191,
+      "learning_rate": 1.4625470302700715e-06,
+      "loss": 0.6634,
+      "step": 12288
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1188219970536133,
+      "learning_rate": 1.4624613760454463e-06,
+      "loss": 0.6484,
+      "step": 12289
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0425376396936408,
+      "learning_rate": 1.4623757175046278e-06,
+      "loss": 0.6602,
+      "step": 12290
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.6391487122239554,
+      "learning_rate": 1.462290054648416e-06,
+      "loss": 0.6836,
+      "step": 12291
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.5901841904649814,
+      "learning_rate": 1.4622043874776104e-06,
+      "loss": 0.6576,
+      "step": 12292
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0072619709212776,
+      "learning_rate": 1.46211871599301e-06,
+      "loss": 0.6673,
+      "step": 12293
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.3541586463947233,
+      "learning_rate": 1.4620330401954153e-06,
+      "loss": 0.6641,
+      "step": 12294
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.8596208830471894,
+      "learning_rate": 1.4619473600856249e-06,
+      "loss": 0.6816,
+      "step": 12295
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.1982281591773456,
+      "learning_rate": 1.4618616756644392e-06,
+      "loss": 0.6621,
+      "step": 12296
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.428965359617639,
+      "learning_rate": 1.4617759869326577e-06,
+      "loss": 0.6452,
+      "step": 12297
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.3058418820214253,
+      "learning_rate": 1.4616902938910797e-06,
+      "loss": 0.6842,
+      "step": 12298
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.763162300228691,
+      "learning_rate": 1.4616045965405058e-06,
+      "loss": 0.7018,
+      "step": 12299
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.2950347824529973,
+      "learning_rate": 1.4615188948817349e-06,
+      "loss": 0.6589,
+      "step": 12300
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.312429170685652,
+      "learning_rate": 1.4614331889155679e-06,
+      "loss": 0.679,
+      "step": 12301
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.4277439289045653,
+      "learning_rate": 1.4613474786428037e-06,
+      "loss": 0.6986,
+      "step": 12302
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.201563274647204,
+      "learning_rate": 1.4612617640642427e-06,
+      "loss": 0.6562,
+      "step": 12303
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.697960303165918,
+      "learning_rate": 1.461176045180685e-06,
+      "loss": 0.6406,
+      "step": 12304
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1254844161233222,
+      "learning_rate": 1.4610903219929306e-06,
+      "loss": 0.6732,
+      "step": 12305
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.218461842239054,
+      "learning_rate": 1.4610045945017794e-06,
+      "loss": 0.651,
+      "step": 12306
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.362126564908107,
+      "learning_rate": 1.4609188627080313e-06,
+      "loss": 0.6738,
+      "step": 12307
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.647285563179238,
+      "learning_rate": 1.460833126612487e-06,
+      "loss": 0.6628,
+      "step": 12308
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1296657324083497,
+      "learning_rate": 1.4607473862159464e-06,
+      "loss": 0.6934,
+      "step": 12309
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.3908690375270725,
+      "learning_rate": 1.4606616415192093e-06,
+      "loss": 0.6719,
+      "step": 12310
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.547802478657936,
+      "learning_rate": 1.4605758925230768e-06,
+      "loss": 0.6569,
+      "step": 12311
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6995316640214142,
+      "learning_rate": 1.4604901392283484e-06,
+      "loss": 0.666,
+      "step": 12312
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9177561426223245,
+      "learning_rate": 1.460404381635825e-06,
+      "loss": 0.6966,
+      "step": 12313
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0456373101499186,
+      "learning_rate": 1.4603186197463067e-06,
+      "loss": 0.6628,
+      "step": 12314
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0682570371122386,
+      "learning_rate": 1.4602328535605942e-06,
+      "loss": 0.651,
+      "step": 12315
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.4323205212131374,
+      "learning_rate": 1.4601470830794873e-06,
+      "loss": 0.6797,
+      "step": 12316
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.5614638203942555,
+      "learning_rate": 1.4600613083037874e-06,
+      "loss": 0.6243,
+      "step": 12317
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.251124243359059,
+      "learning_rate": 1.4599755292342949e-06,
+      "loss": 0.6576,
+      "step": 12318
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0475711028883206,
+      "learning_rate": 1.4598897458718097e-06,
+      "loss": 0.6797,
+      "step": 12319
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.9275563789284362,
+      "learning_rate": 1.4598039582171326e-06,
+      "loss": 0.6432,
+      "step": 12320
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.898369043340807,
+      "learning_rate": 1.459718166271065e-06,
+      "loss": 0.6543,
+      "step": 12321
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.2272741175726067,
+      "learning_rate": 1.4596323700344066e-06,
+      "loss": 0.6543,
+      "step": 12322
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.0126871980693988,
+      "learning_rate": 1.4595465695079589e-06,
+      "loss": 0.6979,
+      "step": 12323
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.2011333407352227,
+      "learning_rate": 1.4594607646925226e-06,
+      "loss": 0.6862,
+      "step": 12324
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6132565971470834,
+      "learning_rate": 1.4593749555888984e-06,
+      "loss": 0.6699,
+      "step": 12325
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.086127426322782,
+      "learning_rate": 1.4592891421978867e-06,
+      "loss": 0.6634,
+      "step": 12326
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.2200696642400461,
+      "learning_rate": 1.4592033245202889e-06,
+      "loss": 0.6536,
+      "step": 12327
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0066207730990318,
+      "learning_rate": 1.4591175025569063e-06,
+      "loss": 0.666,
+      "step": 12328
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.1365620199172,
+      "learning_rate": 1.459031676308539e-06,
+      "loss": 0.6842,
+      "step": 12329
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.3832639755199048,
+      "learning_rate": 1.4589458457759886e-06,
+      "loss": 0.6895,
+      "step": 12330
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.393335813800698,
+      "learning_rate": 1.4588600109600563e-06,
+      "loss": 0.6693,
+      "step": 12331
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.606156635189357,
+      "learning_rate": 1.4587741718615425e-06,
+      "loss": 0.6966,
+      "step": 12332
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 6.645119196469157,
+      "learning_rate": 1.458688328481249e-06,
+      "loss": 0.6725,
+      "step": 12333
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.3340969656665214,
+      "learning_rate": 1.4586024808199766e-06,
+      "loss": 0.6667,
+      "step": 12334
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.091680068802187,
+      "learning_rate": 1.458516628878527e-06,
+      "loss": 0.6797,
+      "step": 12335
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5535338376449659,
+      "learning_rate": 1.458430772657701e-06,
+      "loss": 0.6836,
+      "step": 12336
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.5767259251974135,
+      "learning_rate": 1.4583449121582998e-06,
+      "loss": 0.6634,
+      "step": 12337
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.386220919017554,
+      "learning_rate": 1.4582590473811253e-06,
+      "loss": 0.6712,
+      "step": 12338
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.1494769227814907,
+      "learning_rate": 1.4581731783269786e-06,
+      "loss": 0.6289,
+      "step": 12339
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.5678335719004501,
+      "learning_rate": 1.4580873049966609e-06,
+      "loss": 0.6777,
+      "step": 12340
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.613936727181965,
+      "learning_rate": 1.4580014273909744e-06,
+      "loss": 0.7012,
+      "step": 12341
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 5.355254274831479,
+      "learning_rate": 1.4579155455107196e-06,
+      "loss": 0.6888,
+      "step": 12342
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.8729346309110053,
+      "learning_rate": 1.4578296593566982e-06,
+      "loss": 0.6497,
+      "step": 12343
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.0725620151061532,
+      "learning_rate": 1.4577437689297127e-06,
+      "loss": 0.6771,
+      "step": 12344
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.0023685514120753,
+      "learning_rate": 1.4576578742305642e-06,
+      "loss": 0.6823,
+      "step": 12345
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.8318585517327284,
+      "learning_rate": 1.4575719752600536e-06,
+      "loss": 0.6667,
+      "step": 12346
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8746218086921653,
+      "learning_rate": 1.4574860720189838e-06,
+      "loss": 0.696,
+      "step": 12347
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9178024385810218,
+      "learning_rate": 1.457400164508156e-06,
+      "loss": 0.653,
+      "step": 12348
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9681020906389491,
+      "learning_rate": 1.457314252728372e-06,
+      "loss": 0.666,
+      "step": 12349
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7536934820314978,
+      "learning_rate": 1.4572283366804335e-06,
+      "loss": 0.6829,
+      "step": 12350
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.892422773944393,
+      "learning_rate": 1.4571424163651425e-06,
+      "loss": 0.679,
+      "step": 12351
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.303243052175317,
+      "learning_rate": 1.457056491783301e-06,
+      "loss": 0.6569,
+      "step": 12352
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.616104963120338,
+      "learning_rate": 1.4569705629357107e-06,
+      "loss": 0.6751,
+      "step": 12353
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1742226927596253,
+      "learning_rate": 1.4568846298231742e-06,
+      "loss": 0.6882,
+      "step": 12354
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.4168245859600526,
+      "learning_rate": 1.4567986924464925e-06,
+      "loss": 0.6934,
+      "step": 12355
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.2655810833351862,
+      "learning_rate": 1.4567127508064684e-06,
+      "loss": 0.6712,
+      "step": 12356
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.3664419155898795,
+      "learning_rate": 1.4566268049039038e-06,
+      "loss": 0.6667,
+      "step": 12357
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9648417689894088,
+      "learning_rate": 1.4565408547396007e-06,
+      "loss": 0.6641,
+      "step": 12358
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.369490038360759,
+      "learning_rate": 1.4564549003143617e-06,
+      "loss": 0.6777,
+      "step": 12359
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.8145846877002754,
+      "learning_rate": 1.4563689416289887e-06,
+      "loss": 0.6862,
+      "step": 12360
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.2775625350252207,
+      "learning_rate": 1.4562829786842838e-06,
+      "loss": 0.6784,
+      "step": 12361
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.6239690637634294,
+      "learning_rate": 1.4561970114810497e-06,
+      "loss": 0.6693,
+      "step": 12362
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.1593929295021512,
+      "learning_rate": 1.4561110400200884e-06,
+      "loss": 0.6797,
+      "step": 12363
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.1491249250045785,
+      "learning_rate": 1.4560250643022027e-06,
+      "loss": 0.6497,
+      "step": 12364
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.6183071740515564,
+      "learning_rate": 1.4559390843281944e-06,
+      "loss": 0.6686,
+      "step": 12365
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.6534247920776286,
+      "learning_rate": 1.4558531000988666e-06,
+      "loss": 0.6647,
+      "step": 12366
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.510353228967995,
+      "learning_rate": 1.4557671116150214e-06,
+      "loss": 0.6699,
+      "step": 12367
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9284576944822343,
+      "learning_rate": 1.4556811188774613e-06,
+      "loss": 0.6478,
+      "step": 12368
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9473501568395873,
+      "learning_rate": 1.455595121886989e-06,
+      "loss": 0.6732,
+      "step": 12369
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.0672813527426297,
+      "learning_rate": 1.4555091206444076e-06,
+      "loss": 0.6829,
+      "step": 12370
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 7.368321984083155,
+      "learning_rate": 1.455423115150519e-06,
+      "loss": 0.6868,
+      "step": 12371
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.2772983723382383,
+      "learning_rate": 1.455337105406126e-06,
+      "loss": 0.6947,
+      "step": 12372
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.509040955558913,
+      "learning_rate": 1.4552510914120318e-06,
+      "loss": 0.6426,
+      "step": 12373
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.1856723912044402,
+      "learning_rate": 1.4551650731690389e-06,
+      "loss": 0.7031,
+      "step": 12374
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.759403776170905,
+      "learning_rate": 1.4550790506779498e-06,
+      "loss": 0.6836,
+      "step": 12375
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.9235037183863368,
+      "learning_rate": 1.4549930239395677e-06,
+      "loss": 0.6641,
+      "step": 12376
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.230663736675649,
+      "learning_rate": 1.4549069929546955e-06,
+      "loss": 0.7051,
+      "step": 12377
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.7807591375201155,
+      "learning_rate": 1.4548209577241366e-06,
+      "loss": 0.6654,
+      "step": 12378
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.2488084152359842,
+      "learning_rate": 1.454734918248693e-06,
+      "loss": 0.6999,
+      "step": 12379
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.2465282063509613,
+      "learning_rate": 1.4546488745291682e-06,
+      "loss": 0.6823,
+      "step": 12380
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.670908978238409,
+      "learning_rate": 1.4545628265663657e-06,
+      "loss": 0.6484,
+      "step": 12381
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9778504386806836,
+      "learning_rate": 1.4544767743610876e-06,
+      "loss": 0.6628,
+      "step": 12382
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.4069285331804187,
+      "learning_rate": 1.454390717914138e-06,
+      "loss": 0.6784,
+      "step": 12383
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.8188331938888775,
+      "learning_rate": 1.4543046572263194e-06,
+      "loss": 0.6628,
+      "step": 12384
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.185761871521142,
+      "learning_rate": 1.4542185922984352e-06,
+      "loss": 0.6725,
+      "step": 12385
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.2288706982516169,
+      "learning_rate": 1.454132523131289e-06,
+      "loss": 0.6341,
+      "step": 12386
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.886212120417973,
+      "learning_rate": 1.4540464497256834e-06,
+      "loss": 0.6712,
+      "step": 12387
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.457020588711908,
+      "learning_rate": 1.4539603720824222e-06,
+      "loss": 0.6595,
+      "step": 12388
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.002037621406297,
+      "learning_rate": 1.4538742902023088e-06,
+      "loss": 0.6582,
+      "step": 12389
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7558344966994766,
+      "learning_rate": 1.4537882040861464e-06,
+      "loss": 0.6562,
+      "step": 12390
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.025307522416319,
+      "learning_rate": 1.4537021137347391e-06,
+      "loss": 0.6478,
+      "step": 12391
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.5560308624186145,
+      "learning_rate": 1.4536160191488893e-06,
+      "loss": 0.6478,
+      "step": 12392
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 8.612290902423858,
+      "learning_rate": 1.4535299203294012e-06,
+      "loss": 0.6914,
+      "step": 12393
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4377149750572964,
+      "learning_rate": 1.4534438172770784e-06,
+      "loss": 0.6667,
+      "step": 12394
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.243596055372939,
+      "learning_rate": 1.453357709992724e-06,
+      "loss": 0.6595,
+      "step": 12395
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.8785697854188608,
+      "learning_rate": 1.4532715984771424e-06,
+      "loss": 0.6816,
+      "step": 12396
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.845802347553367,
+      "learning_rate": 1.4531854827311368e-06,
+      "loss": 0.7005,
+      "step": 12397
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.4644227875313476,
+      "learning_rate": 1.4530993627555107e-06,
+      "loss": 0.6908,
+      "step": 12398
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.2776396885277395,
+      "learning_rate": 1.4530132385510685e-06,
+      "loss": 0.64,
+      "step": 12399
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4019412318461644,
+      "learning_rate": 1.4529271101186136e-06,
+      "loss": 0.6608,
+      "step": 12400
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9840790326260761,
+      "learning_rate": 1.4528409774589497e-06,
+      "loss": 0.6745,
+      "step": 12401
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3592035988588818,
+      "learning_rate": 1.4527548405728814e-06,
+      "loss": 0.666,
+      "step": 12402
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.127300089812723,
+      "learning_rate": 1.4526686994612117e-06,
+      "loss": 0.64,
+      "step": 12403
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.022560936055211,
+      "learning_rate": 1.4525825541247452e-06,
+      "loss": 0.6667,
+      "step": 12404
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.540494362532085,
+      "learning_rate": 1.4524964045642859e-06,
+      "loss": 0.7018,
+      "step": 12405
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.319405772991775,
+      "learning_rate": 1.452410250780637e-06,
+      "loss": 0.6908,
+      "step": 12406
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.9692966087952928,
+      "learning_rate": 1.4523240927746038e-06,
+      "loss": 0.6432,
+      "step": 12407
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.1347802290800355,
+      "learning_rate": 1.4522379305469895e-06,
+      "loss": 0.653,
+      "step": 12408
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4284465412891802,
+      "learning_rate": 1.4521517640985986e-06,
+      "loss": 0.6725,
+      "step": 12409
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0774468126662733,
+      "learning_rate": 1.4520655934302356e-06,
+      "loss": 0.6979,
+      "step": 12410
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.893985733540972,
+      "learning_rate": 1.4519794185427045e-06,
+      "loss": 0.6595,
+      "step": 12411
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.6757189997159703,
+      "learning_rate": 1.4518932394368093e-06,
+      "loss": 0.6686,
+      "step": 12412
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8730216349640032,
+      "learning_rate": 1.4518070561133544e-06,
+      "loss": 0.6771,
+      "step": 12413
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.2843981144343752,
+      "learning_rate": 1.4517208685731445e-06,
+      "loss": 0.6771,
+      "step": 12414
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8392610981143396,
+      "learning_rate": 1.4516346768169838e-06,
+      "loss": 0.6719,
+      "step": 12415
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.33434607769667,
+      "learning_rate": 1.4515484808456763e-06,
+      "loss": 0.7044,
+      "step": 12416
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.2460805388358238,
+      "learning_rate": 1.4514622806600273e-06,
+      "loss": 0.6608,
+      "step": 12417
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3899524836977712,
+      "learning_rate": 1.451376076260841e-06,
+      "loss": 0.6439,
+      "step": 12418
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.1069707589165976,
+      "learning_rate": 1.4512898676489216e-06,
+      "loss": 0.6712,
+      "step": 12419
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8324222276089908,
+      "learning_rate": 1.4512036548250741e-06,
+      "loss": 0.6523,
+      "step": 12420
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.687047909781031,
+      "learning_rate": 1.451117437790103e-06,
+      "loss": 0.724,
+      "step": 12421
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.0656555916535084,
+      "learning_rate": 1.4510312165448128e-06,
+      "loss": 0.707,
+      "step": 12422
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.1614303410097544,
+      "learning_rate": 1.4509449910900085e-06,
+      "loss": 0.6634,
+      "step": 12423
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7017602453527292,
+      "learning_rate": 1.4508587614264948e-06,
+      "loss": 0.6556,
+      "step": 12424
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.6764208861376595,
+      "learning_rate": 1.4507725275550765e-06,
+      "loss": 0.6842,
+      "step": 12425
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9112740464087864,
+      "learning_rate": 1.450686289476558e-06,
+      "loss": 0.6576,
+      "step": 12426
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.175734763963531,
+      "learning_rate": 1.4506000471917445e-06,
+      "loss": 0.6296,
+      "step": 12427
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8772470026223999,
+      "learning_rate": 1.4505138007014411e-06,
+      "loss": 0.6875,
+      "step": 12428
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.127107555250363,
+      "learning_rate": 1.4504275500064524e-06,
+      "loss": 0.6589,
+      "step": 12429
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3101168402356096,
+      "learning_rate": 1.4503412951075837e-06,
+      "loss": 0.6556,
+      "step": 12430
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.205948714967068,
+      "learning_rate": 1.45025503600564e-06,
+      "loss": 0.6523,
+      "step": 12431
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.0490133174389114,
+      "learning_rate": 1.4501687727014258e-06,
+      "loss": 0.6432,
+      "step": 12432
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.586129818393517,
+      "learning_rate": 1.450082505195747e-06,
+      "loss": 0.6823,
+      "step": 12433
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.945690660268915,
+      "learning_rate": 1.4499962334894083e-06,
+      "loss": 0.6432,
+      "step": 12434
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.2530325609265869,
+      "learning_rate": 1.4499099575832147e-06,
+      "loss": 0.7005,
+      "step": 12435
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 5.422082795971429,
+      "learning_rate": 1.449823677477972e-06,
+      "loss": 0.6842,
+      "step": 12436
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.80849807221026,
+      "learning_rate": 1.4497373931744853e-06,
+      "loss": 0.6986,
+      "step": 12437
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.4275115248838937,
+      "learning_rate": 1.449651104673559e-06,
+      "loss": 0.6764,
+      "step": 12438
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.500027093935637,
+      "learning_rate": 1.4495648119759999e-06,
+      "loss": 0.666,
+      "step": 12439
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6808298560027195,
+      "learning_rate": 1.4494785150826123e-06,
+      "loss": 0.651,
+      "step": 12440
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.4615505183194404,
+      "learning_rate": 1.449392213994202e-06,
+      "loss": 0.6764,
+      "step": 12441
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.5637310832739368,
+      "learning_rate": 1.4493059087115745e-06,
+      "loss": 0.6758,
+      "step": 12442
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.2933772903627727,
+      "learning_rate": 1.4492195992355352e-06,
+      "loss": 0.6458,
+      "step": 12443
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.543161717183826,
+      "learning_rate": 1.4491332855668899e-06,
+      "loss": 0.6484,
+      "step": 12444
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.767386580533365,
+      "learning_rate": 1.4490469677064435e-06,
+      "loss": 0.6523,
+      "step": 12445
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.778888567784385,
+      "learning_rate": 1.4489606456550024e-06,
+      "loss": 0.6647,
+      "step": 12446
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.050593781943326,
+      "learning_rate": 1.4488743194133717e-06,
+      "loss": 0.6855,
+      "step": 12447
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.518506598391988,
+      "learning_rate": 1.448787988982357e-06,
+      "loss": 0.6693,
+      "step": 12448
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.085230341601038,
+      "learning_rate": 1.4487016543627646e-06,
+      "loss": 0.6797,
+      "step": 12449
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.1971794330071301,
+      "learning_rate": 1.4486153155554e-06,
+      "loss": 0.6855,
+      "step": 12450
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.9853979700482363,
+      "learning_rate": 1.448528972561069e-06,
+      "loss": 0.6641,
+      "step": 12451
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.2559583884596903,
+      "learning_rate": 1.4484426253805775e-06,
+      "loss": 0.6725,
+      "step": 12452
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.88532485520193,
+      "learning_rate": 1.4483562740147311e-06,
+      "loss": 0.6549,
+      "step": 12453
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.031953888931476,
+      "learning_rate": 1.448269918464336e-06,
+      "loss": 0.6901,
+      "step": 12454
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9797599345950729,
+      "learning_rate": 1.4481835587301983e-06,
+      "loss": 0.653,
+      "step": 12455
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.5967065793071815,
+      "learning_rate": 1.4480971948131236e-06,
+      "loss": 0.6758,
+      "step": 12456
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9691948570329385,
+      "learning_rate": 1.4480108267139184e-06,
+      "loss": 0.6927,
+      "step": 12457
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.098600720430582,
+      "learning_rate": 1.4479244544333881e-06,
+      "loss": 0.6751,
+      "step": 12458
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.1814409692782009,
+      "learning_rate": 1.4478380779723395e-06,
+      "loss": 0.638,
+      "step": 12459
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3670916142439984,
+      "learning_rate": 1.4477516973315788e-06,
+      "loss": 0.6419,
+      "step": 12460
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.717567004848262,
+      "learning_rate": 1.4476653125119116e-06,
+      "loss": 0.6667,
+      "step": 12461
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.056241514876347,
+      "learning_rate": 1.4475789235141446e-06,
+      "loss": 0.653,
+      "step": 12462
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3800354846280465,
+      "learning_rate": 1.4474925303390837e-06,
+      "loss": 0.6764,
+      "step": 12463
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.6132226206271527,
+      "learning_rate": 1.4474061329875355e-06,
+      "loss": 0.6517,
+      "step": 12464
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.1592046782102425,
+      "learning_rate": 1.4473197314603063e-06,
+      "loss": 0.6693,
+      "step": 12465
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.2016720790677744,
+      "learning_rate": 1.4472333257582027e-06,
+      "loss": 0.6823,
+      "step": 12466
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.320916702462576,
+      "learning_rate": 1.4471469158820306e-06,
+      "loss": 0.7005,
+      "step": 12467
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.5045564406476266,
+      "learning_rate": 1.447060501832597e-06,
+      "loss": 0.6549,
+      "step": 12468
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.6884655938963564,
+      "learning_rate": 1.446974083610708e-06,
+      "loss": 0.6842,
+      "step": 12469
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.223549647286686,
+      "learning_rate": 1.4468876612171705e-06,
+      "loss": 0.6491,
+      "step": 12470
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.5638453466268443,
+      "learning_rate": 1.4468012346527912e-06,
+      "loss": 0.6582,
+      "step": 12471
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.505166932268932,
+      "learning_rate": 1.4467148039183759e-06,
+      "loss": 0.6921,
+      "step": 12472
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9709796576124676,
+      "learning_rate": 1.4466283690147326e-06,
+      "loss": 0.6855,
+      "step": 12473
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.407872282270559,
+      "learning_rate": 1.4465419299426663e-06,
+      "loss": 0.653,
+      "step": 12474
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3408356629723104,
+      "learning_rate": 1.4464554867029852e-06,
+      "loss": 0.6745,
+      "step": 12475
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.320691854643131,
+      "learning_rate": 1.4463690392964955e-06,
+      "loss": 0.6348,
+      "step": 12476
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.4587467684599067,
+      "learning_rate": 1.4462825877240042e-06,
+      "loss": 0.6576,
+      "step": 12477
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.7606234737790856,
+      "learning_rate": 1.446196131986318e-06,
+      "loss": 0.6797,
+      "step": 12478
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9828835507504571,
+      "learning_rate": 1.4461096720842435e-06,
+      "loss": 0.6875,
+      "step": 12479
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.2720832561270508,
+      "learning_rate": 1.4460232080185883e-06,
+      "loss": 0.6322,
+      "step": 12480
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.5026604544128987,
+      "learning_rate": 1.445936739790159e-06,
+      "loss": 0.6602,
+      "step": 12481
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.1640786844216429,
+      "learning_rate": 1.4458502673997628e-06,
+      "loss": 0.6484,
+      "step": 12482
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.4262943005058357,
+      "learning_rate": 1.4457637908482064e-06,
+      "loss": 0.651,
+      "step": 12483
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.2433972821897883,
+      "learning_rate": 1.4456773101362974e-06,
+      "loss": 0.6725,
+      "step": 12484
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.055048474791283,
+      "learning_rate": 1.4455908252648423e-06,
+      "loss": 0.696,
+      "step": 12485
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.410698143550614,
+      "learning_rate": 1.445504336234649e-06,
+      "loss": 0.6641,
+      "step": 12486
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.2926866094633973,
+      "learning_rate": 1.4454178430465241e-06,
+      "loss": 0.6751,
+      "step": 12487
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 6.859866655055252,
+      "learning_rate": 1.445331345701275e-06,
+      "loss": 0.6921,
+      "step": 12488
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.355775926795653,
+      "learning_rate": 1.4452448441997092e-06,
+      "loss": 0.6732,
+      "step": 12489
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.486622519457837,
+      "learning_rate": 1.445158338542634e-06,
+      "loss": 0.6829,
+      "step": 12490
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 5.747909772207964,
+      "learning_rate": 1.4450718287308565e-06,
+      "loss": 0.6888,
+      "step": 12491
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.2278229816738655,
+      "learning_rate": 1.444985314765184e-06,
+      "loss": 0.6452,
+      "step": 12492
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.2750037732010486,
+      "learning_rate": 1.4448987966464246e-06,
+      "loss": 0.6667,
+      "step": 12493
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.2442961304924807,
+      "learning_rate": 1.4448122743753855e-06,
+      "loss": 0.6732,
+      "step": 12494
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.824040588578854,
+      "learning_rate": 1.4447257479528736e-06,
+      "loss": 0.6816,
+      "step": 12495
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.23549804726312,
+      "learning_rate": 1.4446392173796973e-06,
+      "loss": 0.668,
+      "step": 12496
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.254407223657605,
+      "learning_rate": 1.4445526826566638e-06,
+      "loss": 0.6406,
+      "step": 12497
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.0033299421343163,
+      "learning_rate": 1.4444661437845808e-06,
+      "loss": 0.6706,
+      "step": 12498
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.953563963352119,
+      "learning_rate": 1.4443796007642562e-06,
+      "loss": 0.668,
+      "step": 12499
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.6479677951073497,
+      "learning_rate": 1.4442930535964973e-06,
+      "loss": 0.6829,
+      "step": 12500
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9420785849144322,
+      "learning_rate": 1.4442065022821117e-06,
+      "loss": 0.6764,
+      "step": 12501
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 4.075371943928709,
+      "learning_rate": 1.444119946821908e-06,
+      "loss": 0.6576,
+      "step": 12502
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.321831288802021,
+      "learning_rate": 1.4440333872166934e-06,
+      "loss": 0.6348,
+      "step": 12503
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.756047798931149,
+      "learning_rate": 1.4439468234672759e-06,
+      "loss": 0.6667,
+      "step": 12504
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.7852553505378075,
+      "learning_rate": 1.4438602555744634e-06,
+      "loss": 0.6908,
+      "step": 12505
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.613444780140548,
+      "learning_rate": 1.4437736835390642e-06,
+      "loss": 0.6504,
+      "step": 12506
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.95838953484507,
+      "learning_rate": 1.4436871073618857e-06,
+      "loss": 0.6706,
+      "step": 12507
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.3610608115848963,
+      "learning_rate": 1.4436005270437358e-06,
+      "loss": 0.6745,
+      "step": 12508
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.480928643468722,
+      "learning_rate": 1.4435139425854232e-06,
+      "loss": 0.6569,
+      "step": 12509
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.313870274202174,
+      "learning_rate": 1.4434273539877565e-06,
+      "loss": 0.6712,
+      "step": 12510
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.507763129447576,
+      "learning_rate": 1.4433407612515423e-06,
+      "loss": 0.6764,
+      "step": 12511
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.373338184911481,
+      "learning_rate": 1.4432541643775897e-06,
+      "loss": 0.6882,
+      "step": 12512
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.248556582236853,
+      "learning_rate": 1.4431675633667068e-06,
+      "loss": 0.6673,
+      "step": 12513
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5692631267319064,
+      "learning_rate": 1.4430809582197019e-06,
+      "loss": 0.6699,
+      "step": 12514
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8962456775121872,
+      "learning_rate": 1.4429943489373832e-06,
+      "loss": 0.6628,
+      "step": 12515
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.4639377742338495,
+      "learning_rate": 1.4429077355205591e-06,
+      "loss": 0.6816,
+      "step": 12516
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.362575275971991,
+      "learning_rate": 1.442821117970038e-06,
+      "loss": 0.6654,
+      "step": 12517
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.5771171370844232,
+      "learning_rate": 1.442734496286628e-06,
+      "loss": 0.6589,
+      "step": 12518
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.225307619925999,
+      "learning_rate": 1.442647870471138e-06,
+      "loss": 0.6654,
+      "step": 12519
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.755354409032059,
+      "learning_rate": 1.4425612405243763e-06,
+      "loss": 0.6517,
+      "step": 12520
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0400912075690814,
+      "learning_rate": 1.4424746064471514e-06,
+      "loss": 0.6882,
+      "step": 12521
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.73668467052346,
+      "learning_rate": 1.4423879682402716e-06,
+      "loss": 0.6576,
+      "step": 12522
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.0733388669867727,
+      "learning_rate": 1.4423013259045463e-06,
+      "loss": 0.6452,
+      "step": 12523
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8663234939849576,
+      "learning_rate": 1.442214679440783e-06,
+      "loss": 0.694,
+      "step": 12524
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.3064750947580546,
+      "learning_rate": 1.4421280288497913e-06,
+      "loss": 0.6855,
+      "step": 12525
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.770743498090016,
+      "learning_rate": 1.4420413741323799e-06,
+      "loss": 0.6745,
+      "step": 12526
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8015049463573158,
+      "learning_rate": 1.4419547152893567e-06,
+      "loss": 0.6712,
+      "step": 12527
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.9939878794859203,
+      "learning_rate": 1.4418680523215314e-06,
+      "loss": 0.681,
+      "step": 12528
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.2243556718038562,
+      "learning_rate": 1.4417813852297122e-06,
+      "loss": 0.694,
+      "step": 12529
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5489897589035702,
+      "learning_rate": 1.4416947140147087e-06,
+      "loss": 0.6771,
+      "step": 12530
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.30028689172695,
+      "learning_rate": 1.441608038677329e-06,
+      "loss": 0.6745,
+      "step": 12531
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8067880361236945,
+      "learning_rate": 1.4415213592183824e-06,
+      "loss": 0.6667,
+      "step": 12532
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8994451315898802,
+      "learning_rate": 1.441434675638678e-06,
+      "loss": 0.6712,
+      "step": 12533
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.3540290288516683,
+      "learning_rate": 1.4413479879390251e-06,
+      "loss": 0.6777,
+      "step": 12534
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8332594031185058,
+      "learning_rate": 1.4412612961202317e-06,
+      "loss": 0.6823,
+      "step": 12535
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.5052437168458745,
+      "learning_rate": 1.4411746001831082e-06,
+      "loss": 0.6406,
+      "step": 12536
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0290301325447977,
+      "learning_rate": 1.4410879001284631e-06,
+      "loss": 0.6771,
+      "step": 12537
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.425723402232241,
+      "learning_rate": 1.4410011959571051e-06,
+      "loss": 0.6732,
+      "step": 12538
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.5022097439387947,
+      "learning_rate": 1.4409144876698446e-06,
+      "loss": 0.6582,
+      "step": 12539
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.9001943939992163,
+      "learning_rate": 1.4408277752674897e-06,
+      "loss": 0.6758,
+      "step": 12540
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.607133407347825,
+      "learning_rate": 1.4407410587508503e-06,
+      "loss": 0.679,
+      "step": 12541
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8834046526063377,
+      "learning_rate": 1.4406543381207359e-06,
+      "loss": 0.6595,
+      "step": 12542
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9857111745590363,
+      "learning_rate": 1.4405676133779554e-06,
+      "loss": 0.7025,
+      "step": 12543
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.0653242603713333,
+      "learning_rate": 1.4404808845233185e-06,
+      "loss": 0.6615,
+      "step": 12544
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.736288276301793,
+      "learning_rate": 1.4403941515576343e-06,
+      "loss": 0.6686,
+      "step": 12545
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9380361311398617,
+      "learning_rate": 1.4403074144817127e-06,
+      "loss": 0.6888,
+      "step": 12546
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.9746262561283436,
+      "learning_rate": 1.4402206732963632e-06,
+      "loss": 0.6484,
+      "step": 12547
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.3166050841700847,
+      "learning_rate": 1.4401339280023948e-06,
+      "loss": 0.709,
+      "step": 12548
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.0628061988923159,
+      "learning_rate": 1.440047178600618e-06,
+      "loss": 0.6784,
+      "step": 12549
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.7678560400564869,
+      "learning_rate": 1.4399604250918418e-06,
+      "loss": 0.651,
+      "step": 12550
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.0139795638065414,
+      "learning_rate": 1.4398736674768756e-06,
+      "loss": 0.6699,
+      "step": 12551
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 5.866755954269344,
+      "learning_rate": 1.4397869057565302e-06,
+      "loss": 0.6803,
+      "step": 12552
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.1029200307477667,
+      "learning_rate": 1.4397001399316144e-06,
+      "loss": 0.6771,
+      "step": 12553
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.846048764038593,
+      "learning_rate": 1.4396133700029385e-06,
+      "loss": 0.6797,
+      "step": 12554
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8940650822515483,
+      "learning_rate": 1.439526595971312e-06,
+      "loss": 0.668,
+      "step": 12555
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.1804573115115742,
+      "learning_rate": 1.439439817837545e-06,
+      "loss": 0.6595,
+      "step": 12556
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.0972383743064658,
+      "learning_rate": 1.4393530356024472e-06,
+      "loss": 0.6784,
+      "step": 12557
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.945125785065361,
+      "learning_rate": 1.4392662492668285e-06,
+      "loss": 0.6823,
+      "step": 12558
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.395266403452487,
+      "learning_rate": 1.4391794588314996e-06,
+      "loss": 0.668,
+      "step": 12559
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.8682515601843726,
+      "learning_rate": 1.4390926642972696e-06,
+      "loss": 0.6536,
+      "step": 12560
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6883434453818402,
+      "learning_rate": 1.439005865664949e-06,
+      "loss": 0.6836,
+      "step": 12561
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.6909191445261555,
+      "learning_rate": 1.4389190629353478e-06,
+      "loss": 0.6732,
+      "step": 12562
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.2360019709216006,
+      "learning_rate": 1.4388322561092761e-06,
+      "loss": 0.6354,
+      "step": 12563
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.205154622961903,
+      "learning_rate": 1.438745445187544e-06,
+      "loss": 0.6628,
+      "step": 12564
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.81651945657055,
+      "learning_rate": 1.4386586301709623e-06,
+      "loss": 0.6882,
+      "step": 12565
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9702141356636155,
+      "learning_rate": 1.4385718110603407e-06,
+      "loss": 0.6589,
+      "step": 12566
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.904590547007792,
+      "learning_rate": 1.4384849878564894e-06,
+      "loss": 0.6497,
+      "step": 12567
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8173225434463613,
+      "learning_rate": 1.4383981605602189e-06,
+      "loss": 0.6602,
+      "step": 12568
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.060077097518184,
+      "learning_rate": 1.4383113291723398e-06,
+      "loss": 0.6771,
+      "step": 12569
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.176950142417727,
+      "learning_rate": 1.438224493693662e-06,
+      "loss": 0.6628,
+      "step": 12570
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.4030800945773376,
+      "learning_rate": 1.4381376541249965e-06,
+      "loss": 0.6426,
+      "step": 12571
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5847654826207236,
+      "learning_rate": 1.4380508104671534e-06,
+      "loss": 0.6458,
+      "step": 12572
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.3421766916479956,
+      "learning_rate": 1.4379639627209438e-06,
+      "loss": 0.6888,
+      "step": 12573
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5767375041503078,
+      "learning_rate": 1.4378771108871774e-06,
+      "loss": 0.6628,
+      "step": 12574
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9343908446753237,
+      "learning_rate": 1.4377902549666652e-06,
+      "loss": 0.6888,
+      "step": 12575
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.894321344675919,
+      "learning_rate": 1.437703394960218e-06,
+      "loss": 0.7337,
+      "step": 12576
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.93607782331217,
+      "learning_rate": 1.437616530868646e-06,
+      "loss": 0.6745,
+      "step": 12577
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.98934243922494,
+      "learning_rate": 1.4375296626927605e-06,
+      "loss": 0.681,
+      "step": 12578
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9774088427992862,
+      "learning_rate": 1.437442790433372e-06,
+      "loss": 0.6862,
+      "step": 12579
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.4370661984135573,
+      "learning_rate": 1.437355914091291e-06,
+      "loss": 0.6712,
+      "step": 12580
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0741700765178777,
+      "learning_rate": 1.4372690336673285e-06,
+      "loss": 0.6686,
+      "step": 12581
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.7842791418172035,
+      "learning_rate": 1.4371821491622959e-06,
+      "loss": 0.6771,
+      "step": 12582
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2060927857913637,
+      "learning_rate": 1.4370952605770032e-06,
+      "loss": 0.6784,
+      "step": 12583
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1447949331256906,
+      "learning_rate": 1.437008367912262e-06,
+      "loss": 0.7018,
+      "step": 12584
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.4070511315325773,
+      "learning_rate": 1.436921471168883e-06,
+      "loss": 0.6549,
+      "step": 12585
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0235124259179167,
+      "learning_rate": 1.4368345703476772e-06,
+      "loss": 0.627,
+      "step": 12586
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.5648860449565225,
+      "learning_rate": 1.436747665449456e-06,
+      "loss": 0.6888,
+      "step": 12587
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4051306201384222,
+      "learning_rate": 1.43666075647503e-06,
+      "loss": 0.6823,
+      "step": 12588
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5646017471125926,
+      "learning_rate": 1.4365738434252106e-06,
+      "loss": 0.6966,
+      "step": 12589
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1643267284195604,
+      "learning_rate": 1.4364869263008089e-06,
+      "loss": 0.6654,
+      "step": 12590
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2573731929949188,
+      "learning_rate": 1.4364000051026359e-06,
+      "loss": 0.6816,
+      "step": 12591
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.0132044391681387,
+      "learning_rate": 1.4363130798315033e-06,
+      "loss": 0.6738,
+      "step": 12592
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.432131208925529,
+      "learning_rate": 1.436226150488222e-06,
+      "loss": 0.6634,
+      "step": 12593
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9017062960095459,
+      "learning_rate": 1.4361392170736037e-06,
+      "loss": 0.6673,
+      "step": 12594
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7588905108065105,
+      "learning_rate": 1.436052279588459e-06,
+      "loss": 0.6602,
+      "step": 12595
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.8441746950053384,
+      "learning_rate": 1.4359653380336005e-06,
+      "loss": 0.6615,
+      "step": 12596
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.604944936418334,
+      "learning_rate": 1.4358783924098386e-06,
+      "loss": 0.6921,
+      "step": 12597
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8402019010379713,
+      "learning_rate": 1.435791442717985e-06,
+      "loss": 0.6589,
+      "step": 12598
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.440828039375082,
+      "learning_rate": 1.4357044889588515e-06,
+      "loss": 0.6953,
+      "step": 12599
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.7937164199949109,
+      "learning_rate": 1.4356175311332495e-06,
+      "loss": 0.6647,
+      "step": 12600
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.5939882606722144,
+      "learning_rate": 1.4355305692419904e-06,
+      "loss": 0.6647,
+      "step": 12601
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9692868769244196,
+      "learning_rate": 1.435443603285886e-06,
+      "loss": 0.6823,
+      "step": 12602
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9742249723404245,
+      "learning_rate": 1.4353566332657483e-06,
+      "loss": 0.6491,
+      "step": 12603
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.167606941999491,
+      "learning_rate": 1.435269659182388e-06,
+      "loss": 0.666,
+      "step": 12604
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8530220976872066,
+      "learning_rate": 1.435182681036618e-06,
+      "loss": 0.6693,
+      "step": 12605
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.0296751222755525,
+      "learning_rate": 1.4350956988292496e-06,
+      "loss": 0.6771,
+      "step": 12606
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.2764540807141975,
+      "learning_rate": 1.4350087125610942e-06,
+      "loss": 0.6562,
+      "step": 12607
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.772671837516691,
+      "learning_rate": 1.4349217222329642e-06,
+      "loss": 0.666,
+      "step": 12608
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0127849106665296,
+      "learning_rate": 1.4348347278456712e-06,
+      "loss": 0.6712,
+      "step": 12609
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.492410083852582,
+      "learning_rate": 1.4347477294000274e-06,
+      "loss": 0.6764,
+      "step": 12610
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.3626487553790483,
+      "learning_rate": 1.4346607268968445e-06,
+      "loss": 0.6842,
+      "step": 12611
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8642246906751831,
+      "learning_rate": 1.4345737203369346e-06,
+      "loss": 0.638,
+      "step": 12612
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.490397738364105,
+      "learning_rate": 1.4344867097211098e-06,
+      "loss": 0.6927,
+      "step": 12613
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 7.546005722286114,
+      "learning_rate": 1.4343996950501819e-06,
+      "loss": 0.6562,
+      "step": 12614
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.368452050983621,
+      "learning_rate": 1.4343126763249633e-06,
+      "loss": 0.6589,
+      "step": 12615
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8337737934513196,
+      "learning_rate": 1.4342256535462665e-06,
+      "loss": 0.6947,
+      "step": 12616
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.1324444992368994,
+      "learning_rate": 1.4341386267149025e-06,
+      "loss": 0.7018,
+      "step": 12617
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.007331160080461,
+      "learning_rate": 1.434051595831685e-06,
+      "loss": 0.6549,
+      "step": 12618
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.0228052731166524,
+      "learning_rate": 1.4339645608974254e-06,
+      "loss": 0.6621,
+      "step": 12619
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.1896158957821275,
+      "learning_rate": 1.4338775219129357e-06,
+      "loss": 0.6777,
+      "step": 12620
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9354930720411854,
+      "learning_rate": 1.4337904788790293e-06,
+      "loss": 0.6673,
+      "step": 12621
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.743948656475399,
+      "learning_rate": 1.433703431796518e-06,
+      "loss": 0.6576,
+      "step": 12622
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.624733412396031,
+      "learning_rate": 1.433616380666214e-06,
+      "loss": 0.6934,
+      "step": 12623
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 4.092091295383152,
+      "learning_rate": 1.4335293254889297e-06,
+      "loss": 0.6862,
+      "step": 12624
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.038065145504709,
+      "learning_rate": 1.4334422662654785e-06,
+      "loss": 0.6816,
+      "step": 12625
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8980795838497362,
+      "learning_rate": 1.4333552029966717e-06,
+      "loss": 0.6921,
+      "step": 12626
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.3583505078417613,
+      "learning_rate": 1.4332681356833227e-06,
+      "loss": 0.6823,
+      "step": 12627
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 5.942289375741922,
+      "learning_rate": 1.4331810643262439e-06,
+      "loss": 0.6647,
+      "step": 12628
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.0047696772232269,
+      "learning_rate": 1.433093988926248e-06,
+      "loss": 0.6784,
+      "step": 12629
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.520065852133935,
+      "learning_rate": 1.4330069094841474e-06,
+      "loss": 0.7109,
+      "step": 12630
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9226486952216164,
+      "learning_rate": 1.4329198260007551e-06,
+      "loss": 0.6621,
+      "step": 12631
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.1898386172065347,
+      "learning_rate": 1.4328327384768838e-06,
+      "loss": 0.6816,
+      "step": 12632
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4192513130506603,
+      "learning_rate": 1.4327456469133464e-06,
+      "loss": 0.7018,
+      "step": 12633
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.460812550146282,
+      "learning_rate": 1.4326585513109555e-06,
+      "loss": 0.679,
+      "step": 12634
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9624934202149165,
+      "learning_rate": 1.4325714516705241e-06,
+      "loss": 0.6829,
+      "step": 12635
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.230577221953142,
+      "learning_rate": 1.4324843479928648e-06,
+      "loss": 0.6888,
+      "step": 12636
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5077756016926258,
+      "learning_rate": 1.4323972402787911e-06,
+      "loss": 0.6986,
+      "step": 12637
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4105561433903921,
+      "learning_rate": 1.4323101285291158e-06,
+      "loss": 0.6849,
+      "step": 12638
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.614899394375778,
+      "learning_rate": 1.432223012744652e-06,
+      "loss": 0.6536,
+      "step": 12639
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.1884186835246617,
+      "learning_rate": 1.432135892926212e-06,
+      "loss": 0.6836,
+      "step": 12640
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7217469712589699,
+      "learning_rate": 1.4320487690746096e-06,
+      "loss": 0.6647,
+      "step": 12641
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7215483421138434,
+      "learning_rate": 1.4319616411906585e-06,
+      "loss": 0.6803,
+      "step": 12642
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.81429302586655,
+      "learning_rate": 1.4318745092751705e-06,
+      "loss": 0.6634,
+      "step": 12643
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6481250316517788,
+      "learning_rate": 1.43178737332896e-06,
+      "loss": 0.6686,
+      "step": 12644
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7182377105832082,
+      "learning_rate": 1.4317002333528397e-06,
+      "loss": 0.6758,
+      "step": 12645
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1984175059368667,
+      "learning_rate": 1.4316130893476227e-06,
+      "loss": 0.6693,
+      "step": 12646
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.6966179017134224,
+      "learning_rate": 1.4315259413141225e-06,
+      "loss": 0.707,
+      "step": 12647
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.145689089751425,
+      "learning_rate": 1.4314387892531525e-06,
+      "loss": 0.6992,
+      "step": 12648
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.326709459687918,
+      "learning_rate": 1.4313516331655265e-06,
+      "loss": 0.666,
+      "step": 12649
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.62562079422434,
+      "learning_rate": 1.4312644730520573e-06,
+      "loss": 0.6797,
+      "step": 12650
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9027058495265996,
+      "learning_rate": 1.4311773089135585e-06,
+      "loss": 0.6667,
+      "step": 12651
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1479355794465336,
+      "learning_rate": 1.431090140750844e-06,
+      "loss": 0.6595,
+      "step": 12652
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.5330969181385963,
+      "learning_rate": 1.4310029685647274e-06,
+      "loss": 0.681,
+      "step": 12653
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.7669010644112366,
+      "learning_rate": 1.4309157923560213e-06,
+      "loss": 0.6797,
+      "step": 12654
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.6802848507932517,
+      "learning_rate": 1.4308286121255406e-06,
+      "loss": 0.6484,
+      "step": 12655
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5289490910784596,
+      "learning_rate": 1.430741427874098e-06,
+      "loss": 0.6751,
+      "step": 12656
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7992649854487311,
+      "learning_rate": 1.4306542396025075e-06,
+      "loss": 0.6836,
+      "step": 12657
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.078577453545664,
+      "learning_rate": 1.430567047311583e-06,
+      "loss": 0.6725,
+      "step": 12658
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2384106113655997,
+      "learning_rate": 1.4304798510021383e-06,
+      "loss": 0.6615,
+      "step": 12659
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.4360712614299245,
+      "learning_rate": 1.4303926506749871e-06,
+      "loss": 0.6712,
+      "step": 12660
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.4335587629153648,
+      "learning_rate": 1.430305446330943e-06,
+      "loss": 0.6615,
+      "step": 12661
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.371377673763259,
+      "learning_rate": 1.4302182379708203e-06,
+      "loss": 0.6777,
+      "step": 12662
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9957099285368467,
+      "learning_rate": 1.4301310255954327e-06,
+      "loss": 0.681,
+      "step": 12663
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7356259628415585,
+      "learning_rate": 1.430043809205594e-06,
+      "loss": 0.6641,
+      "step": 12664
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5034781144660627,
+      "learning_rate": 1.4299565888021185e-06,
+      "loss": 0.6777,
+      "step": 12665
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.6702859724737253,
+      "learning_rate": 1.4298693643858204e-06,
+      "loss": 0.6602,
+      "step": 12666
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.9413133512991246,
+      "learning_rate": 1.429782135957513e-06,
+      "loss": 0.668,
+      "step": 12667
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.8643791679577477,
+      "learning_rate": 1.4296949035180116e-06,
+      "loss": 0.6699,
+      "step": 12668
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1807945880777444,
+      "learning_rate": 1.4296076670681292e-06,
+      "loss": 0.6992,
+      "step": 12669
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.1899190545259675,
+      "learning_rate": 1.4295204266086804e-06,
+      "loss": 0.6589,
+      "step": 12670
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.710313190566166,
+      "learning_rate": 1.4294331821404797e-06,
+      "loss": 0.6758,
+      "step": 12671
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0341826571891455,
+      "learning_rate": 1.429345933664341e-06,
+      "loss": 0.6693,
+      "step": 12672
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2619013184730898,
+      "learning_rate": 1.4292586811810787e-06,
+      "loss": 0.6628,
+      "step": 12673
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 4.822532571369567,
+      "learning_rate": 1.4291714246915074e-06,
+      "loss": 0.6849,
+      "step": 12674
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0143199551857882,
+      "learning_rate": 1.429084164196441e-06,
+      "loss": 0.6335,
+      "step": 12675
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.5900012744377325,
+      "learning_rate": 1.4289968996966944e-06,
+      "loss": 0.6602,
+      "step": 12676
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.3742022269836434,
+      "learning_rate": 1.4289096311930815e-06,
+      "loss": 0.6178,
+      "step": 12677
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9924404672782514,
+      "learning_rate": 1.4288223586864173e-06,
+      "loss": 0.6875,
+      "step": 12678
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.4241098032902069,
+      "learning_rate": 1.4287350821775158e-06,
+      "loss": 0.651,
+      "step": 12679
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.609135397296484,
+      "learning_rate": 1.4286478016671922e-06,
+      "loss": 0.696,
+      "step": 12680
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6304452040659572,
+      "learning_rate": 1.4285605171562606e-06,
+      "loss": 0.6888,
+      "step": 12681
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1566376205986806,
+      "learning_rate": 1.428473228645536e-06,
+      "loss": 0.6576,
+      "step": 12682
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9673779351331424,
+      "learning_rate": 1.4283859361358323e-06,
+      "loss": 0.6986,
+      "step": 12683
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0890378937705893,
+      "learning_rate": 1.4282986396279654e-06,
+      "loss": 0.6465,
+      "step": 12684
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.478699758828585,
+      "learning_rate": 1.4282113391227494e-06,
+      "loss": 0.6732,
+      "step": 12685
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2125562285909288,
+      "learning_rate": 1.4281240346209987e-06,
+      "loss": 0.6784,
+      "step": 12686
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2774046139894424,
+      "learning_rate": 1.4280367261235287e-06,
+      "loss": 0.6745,
+      "step": 12687
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.1867956319906159,
+      "learning_rate": 1.427949413631154e-06,
+      "loss": 0.7005,
+      "step": 12688
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.965173704914945,
+      "learning_rate": 1.42786209714469e-06,
+      "loss": 0.6699,
+      "step": 12689
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6019287004848963,
+      "learning_rate": 1.4277747766649506e-06,
+      "loss": 0.6829,
+      "step": 12690
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.702922387550265,
+      "learning_rate": 1.4276874521927518e-06,
+      "loss": 0.6849,
+      "step": 12691
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.8924461620052029,
+      "learning_rate": 1.427600123728908e-06,
+      "loss": 0.6641,
+      "step": 12692
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.101597986356412,
+      "learning_rate": 1.4275127912742343e-06,
+      "loss": 0.6667,
+      "step": 12693
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.473736188570242,
+      "learning_rate": 1.4274254548295463e-06,
+      "loss": 0.6517,
+      "step": 12694
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.990996177310839,
+      "learning_rate": 1.4273381143956585e-06,
+      "loss": 0.666,
+      "step": 12695
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.9214614387080842,
+      "learning_rate": 1.4272507699733862e-06,
+      "loss": 0.666,
+      "step": 12696
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.4693210677234,
+      "learning_rate": 1.4271634215635448e-06,
+      "loss": 0.6777,
+      "step": 12697
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.809803772218447,
+      "learning_rate": 1.4270760691669495e-06,
+      "loss": 0.6641,
+      "step": 12698
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.256121715189284,
+      "learning_rate": 1.4269887127844152e-06,
+      "loss": 0.6758,
+      "step": 12699
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 5.018544180879667,
+      "learning_rate": 1.4269013524167579e-06,
+      "loss": 0.7161,
+      "step": 12700
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 3.283227748356394,
+      "learning_rate": 1.4268139880647923e-06,
+      "loss": 0.7051,
+      "step": 12701
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5524200477270582,
+      "learning_rate": 1.4267266197293336e-06,
+      "loss": 0.6634,
+      "step": 12702
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.380626646596437,
+      "learning_rate": 1.4266392474111983e-06,
+      "loss": 0.6654,
+      "step": 12703
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0402069140252834,
+      "learning_rate": 1.4265518711112009e-06,
+      "loss": 0.6589,
+      "step": 12704
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 6.067516708503343,
+      "learning_rate": 1.426464490830157e-06,
+      "loss": 0.6882,
+      "step": 12705
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.212501726235734,
+      "learning_rate": 1.4263771065688825e-06,
+      "loss": 0.6504,
+      "step": 12706
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9428620711370828,
+      "learning_rate": 1.4262897183281926e-06,
+      "loss": 0.6738,
+      "step": 12707
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8820354869494984,
+      "learning_rate": 1.4262023261089034e-06,
+      "loss": 0.6855,
+      "step": 12708
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.308027430023218,
+      "learning_rate": 1.4261149299118297e-06,
+      "loss": 0.6751,
+      "step": 12709
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.8669784950856747,
+      "learning_rate": 1.4260275297377881e-06,
+      "loss": 0.6953,
+      "step": 12710
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.4459517815961083,
+      "learning_rate": 1.425940125587594e-06,
+      "loss": 0.6615,
+      "step": 12711
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.6792662476479117,
+      "learning_rate": 1.4258527174620625e-06,
+      "loss": 0.6882,
+      "step": 12712
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.1831735637288827,
+      "learning_rate": 1.42576530536201e-06,
+      "loss": 0.6699,
+      "step": 12713
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.6387269783345153,
+      "learning_rate": 1.4256778892882524e-06,
+      "loss": 0.6634,
+      "step": 12714
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.018773192636667,
+      "learning_rate": 1.4255904692416056e-06,
+      "loss": 0.7018,
+      "step": 12715
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 8.451516978362141,
+      "learning_rate": 1.4255030452228852e-06,
+      "loss": 0.7051,
+      "step": 12716
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.3907951904895464,
+      "learning_rate": 1.4254156172329071e-06,
+      "loss": 0.666,
+      "step": 12717
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7791976449715916,
+      "learning_rate": 1.4253281852724875e-06,
+      "loss": 0.6673,
+      "step": 12718
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5997027665253245,
+      "learning_rate": 1.4252407493424423e-06,
+      "loss": 0.6816,
+      "step": 12719
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.9879839465390745,
+      "learning_rate": 1.4251533094435876e-06,
+      "loss": 0.6549,
+      "step": 12720
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1356776321615174,
+      "learning_rate": 1.4250658655767397e-06,
+      "loss": 0.6693,
+      "step": 12721
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6647905830394203,
+      "learning_rate": 1.4249784177427143e-06,
+      "loss": 0.6921,
+      "step": 12722
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8150325827735764,
+      "learning_rate": 1.4248909659423276e-06,
+      "loss": 0.679,
+      "step": 12723
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.140826224035399,
+      "learning_rate": 1.4248035101763962e-06,
+      "loss": 0.6719,
+      "step": 12724
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.409775501761359,
+      "learning_rate": 1.4247160504457361e-06,
+      "loss": 0.709,
+      "step": 12725
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.613073084397558,
+      "learning_rate": 1.4246285867511635e-06,
+      "loss": 0.6647,
+      "step": 12726
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.3553974144954153,
+      "learning_rate": 1.4245411190934945e-06,
+      "loss": 0.6562,
+      "step": 12727
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4875458525697474,
+      "learning_rate": 1.4244536474735458e-06,
+      "loss": 0.6738,
+      "step": 12728
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.4608520277306667,
+      "learning_rate": 1.4243661718921337e-06,
+      "loss": 0.679,
+      "step": 12729
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.068085797169004,
+      "learning_rate": 1.4242786923500745e-06,
+      "loss": 0.666,
+      "step": 12730
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.898976796491099,
+      "learning_rate": 1.4241912088481848e-06,
+      "loss": 0.7051,
+      "step": 12731
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.5487879773036877,
+      "learning_rate": 1.4241037213872811e-06,
+      "loss": 0.6641,
+      "step": 12732
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.9555766755679003,
+      "learning_rate": 1.4240162299681798e-06,
+      "loss": 0.6719,
+      "step": 12733
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.2734963962361614,
+      "learning_rate": 1.4239287345916977e-06,
+      "loss": 0.694,
+      "step": 12734
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8083589455554778,
+      "learning_rate": 1.423841235258651e-06,
+      "loss": 0.6895,
+      "step": 12735
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9387652395665186,
+      "learning_rate": 1.4237537319698568e-06,
+      "loss": 0.6784,
+      "step": 12736
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6903170439586284,
+      "learning_rate": 1.4236662247261314e-06,
+      "loss": 0.6868,
+      "step": 12737
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7826734340825222,
+      "learning_rate": 1.4235787135282915e-06,
+      "loss": 0.679,
+      "step": 12738
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.241993478189037,
+      "learning_rate": 1.423491198377154e-06,
+      "loss": 0.6738,
+      "step": 12739
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5986956771191057,
+      "learning_rate": 1.423403679273536e-06,
+      "loss": 0.6576,
+      "step": 12740
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7526370645555619,
+      "learning_rate": 1.423316156218254e-06,
+      "loss": 0.6738,
+      "step": 12741
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.936583332830843,
+      "learning_rate": 1.4232286292121248e-06,
+      "loss": 0.6973,
+      "step": 12742
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.764801304978914,
+      "learning_rate": 1.4231410982559653e-06,
+      "loss": 0.6901,
+      "step": 12743
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.8252619747658465,
+      "learning_rate": 1.4230535633505924e-06,
+      "loss": 0.6797,
+      "step": 12744
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.6664565348784355,
+      "learning_rate": 1.4229660244968233e-06,
+      "loss": 0.6693,
+      "step": 12745
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.543722878153807,
+      "learning_rate": 1.4228784816954748e-06,
+      "loss": 0.6836,
+      "step": 12746
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.263278176007791,
+      "learning_rate": 1.4227909349473643e-06,
+      "loss": 0.6751,
+      "step": 12747
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.6953176981659257,
+      "learning_rate": 1.4227033842533085e-06,
+      "loss": 0.6875,
+      "step": 12748
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9460256808336505,
+      "learning_rate": 1.4226158296141244e-06,
+      "loss": 0.6888,
+      "step": 12749
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 4.248479467949668,
+      "learning_rate": 1.4225282710306297e-06,
+      "loss": 0.6569,
+      "step": 12750
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.926793026965161,
+      "learning_rate": 1.4224407085036413e-06,
+      "loss": 0.6914,
+      "step": 12751
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.1135652796665,
+      "learning_rate": 1.422353142033976e-06,
+      "loss": 0.6693,
+      "step": 12752
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.285448919385564,
+      "learning_rate": 1.4222655716224518e-06,
+      "loss": 0.6712,
+      "step": 12753
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.5140820492280502,
+      "learning_rate": 1.4221779972698857e-06,
+      "loss": 0.6536,
+      "step": 12754
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.699768815261376,
+      "learning_rate": 1.422090418977095e-06,
+      "loss": 0.6712,
+      "step": 12755
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.0929759798951935,
+      "learning_rate": 1.422002836744897e-06,
+      "loss": 0.666,
+      "step": 12756
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.880090106128239,
+      "learning_rate": 1.4219152505741091e-06,
+      "loss": 0.6823,
+      "step": 12757
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.4202476048058577,
+      "learning_rate": 1.4218276604655493e-06,
+      "loss": 0.6745,
+      "step": 12758
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 6.0021838187678975,
+      "learning_rate": 1.421740066420034e-06,
+      "loss": 0.6465,
+      "step": 12759
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.220890857200297,
+      "learning_rate": 1.421652468438382e-06,
+      "loss": 0.6608,
+      "step": 12760
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4320712525761068,
+      "learning_rate": 1.4215648665214099e-06,
+      "loss": 0.6816,
+      "step": 12761
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4441200143845128,
+      "learning_rate": 1.4214772606699355e-06,
+      "loss": 0.6771,
+      "step": 12762
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.2074169609215877,
+      "learning_rate": 1.4213896508847767e-06,
+      "loss": 0.6797,
+      "step": 12763
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.0639142496986387,
+      "learning_rate": 1.421302037166751e-06,
+      "loss": 0.668,
+      "step": 12764
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.7634009576313927,
+      "learning_rate": 1.421214419516676e-06,
+      "loss": 0.6751,
+      "step": 12765
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.357626045206837,
+      "learning_rate": 1.4211267979353697e-06,
+      "loss": 0.6634,
+      "step": 12766
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.2535089010633653,
+      "learning_rate": 1.42103917242365e-06,
+      "loss": 0.6745,
+      "step": 12767
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.551705141482578,
+      "learning_rate": 1.420951542982334e-06,
+      "loss": 0.6647,
+      "step": 12768
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 5.9693119594959,
+      "learning_rate": 1.4208639096122404e-06,
+      "loss": 0.6406,
+      "step": 12769
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.489743734972277,
+      "learning_rate": 1.4207762723141862e-06,
+      "loss": 0.6406,
+      "step": 12770
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.315395606411171,
+      "learning_rate": 1.4206886310889907e-06,
+      "loss": 0.7331,
+      "step": 12771
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 3.513914468461589,
+      "learning_rate": 1.4206009859374703e-06,
+      "loss": 0.6406,
+      "step": 12772
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.0243000108802602,
+      "learning_rate": 1.420513336860444e-06,
+      "loss": 0.6569,
+      "step": 12773
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6905272811814687,
+      "learning_rate": 1.4204256838587295e-06,
+      "loss": 0.6647,
+      "step": 12774
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.148245886143196,
+      "learning_rate": 1.4203380269331447e-06,
+      "loss": 0.653,
+      "step": 12775
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.938812601227677,
+      "learning_rate": 1.4202503660845082e-06,
+      "loss": 0.6556,
+      "step": 12776
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.0432524661689,
+      "learning_rate": 1.4201627013136378e-06,
+      "loss": 0.6803,
+      "step": 12777
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3317426626234679,
+      "learning_rate": 1.4200750326213516e-06,
+      "loss": 0.6471,
+      "step": 12778
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1866035732723301,
+      "learning_rate": 1.4199873600084683e-06,
+      "loss": 0.6784,
+      "step": 12779
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.6930581728903316,
+      "learning_rate": 1.4198996834758053e-06,
+      "loss": 0.6836,
+      "step": 12780
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.040030080700013,
+      "learning_rate": 1.4198120030241817e-06,
+      "loss": 0.6914,
+      "step": 12781
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2775139468735266,
+      "learning_rate": 1.4197243186544156e-06,
+      "loss": 0.6335,
+      "step": 12782
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.0237889647815557,
+      "learning_rate": 1.4196366303673252e-06,
+      "loss": 0.6576,
+      "step": 12783
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 6.122202798556475,
+      "learning_rate": 1.4195489381637291e-06,
+      "loss": 0.6536,
+      "step": 12784
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.6235425374943415,
+      "learning_rate": 1.4194612420444459e-06,
+      "loss": 0.6615,
+      "step": 12785
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.223915184201507,
+      "learning_rate": 1.4193735420102932e-06,
+      "loss": 0.6745,
+      "step": 12786
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.7784223683139535,
+      "learning_rate": 1.4192858380620909e-06,
+      "loss": 0.6283,
+      "step": 12787
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.005670763683182,
+      "learning_rate": 1.4191981302006563e-06,
+      "loss": 0.6849,
+      "step": 12788
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.036029416424164,
+      "learning_rate": 1.4191104184268087e-06,
+      "loss": 0.6543,
+      "step": 12789
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.797979299755235,
+      "learning_rate": 1.4190227027413663e-06,
+      "loss": 0.7188,
+      "step": 12790
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3210712541555576,
+      "learning_rate": 1.418934983145148e-06,
+      "loss": 0.6413,
+      "step": 12791
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6966839911345606,
+      "learning_rate": 1.4188472596389727e-06,
+      "loss": 0.7044,
+      "step": 12792
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.223707150222009,
+      "learning_rate": 1.4187595322236584e-06,
+      "loss": 0.6641,
+      "step": 12793
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.003132416703462,
+      "learning_rate": 1.4186718009000249e-06,
+      "loss": 0.6829,
+      "step": 12794
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.100443520206562,
+      "learning_rate": 1.4185840656688902e-06,
+      "loss": 0.6829,
+      "step": 12795
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.122683266659331,
+      "learning_rate": 1.4184963265310734e-06,
+      "loss": 0.6465,
+      "step": 12796
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.1329489000970607,
+      "learning_rate": 1.4184085834873933e-06,
+      "loss": 0.6921,
+      "step": 12797
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2130310509818363,
+      "learning_rate": 1.4183208365386691e-06,
+      "loss": 0.6536,
+      "step": 12798
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.526173007930855,
+      "learning_rate": 1.4182330856857192e-06,
+      "loss": 0.6758,
+      "step": 12799
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.0862990799623717,
+      "learning_rate": 1.4181453309293634e-06,
+      "loss": 0.6829,
+      "step": 12800
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3588527404862987,
+      "learning_rate": 1.41805757227042e-06,
+      "loss": 0.6725,
+      "step": 12801
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.1053358021847703,
+      "learning_rate": 1.4179698097097082e-06,
+      "loss": 0.6725,
+      "step": 12802
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.4249152663141573,
+      "learning_rate": 1.4178820432480474e-06,
+      "loss": 0.6914,
+      "step": 12803
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2123743402477163,
+      "learning_rate": 1.4177942728862565e-06,
+      "loss": 0.681,
+      "step": 12804
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.2578381340597096,
+      "learning_rate": 1.4177064986251544e-06,
+      "loss": 0.6725,
+      "step": 12805
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.075482908936755,
+      "learning_rate": 1.417618720465561e-06,
+      "loss": 0.7057,
+      "step": 12806
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.0314485990593532,
+      "learning_rate": 1.417530938408295e-06,
+      "loss": 0.6803,
+      "step": 12807
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.598870063716184,
+      "learning_rate": 1.417443152454176e-06,
+      "loss": 0.653,
+      "step": 12808
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6529315316335704,
+      "learning_rate": 1.4173553626040226e-06,
+      "loss": 0.6634,
+      "step": 12809
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.5055810502735896,
+      "learning_rate": 1.4172675688586551e-06,
+      "loss": 0.6836,
+      "step": 12810
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.6123977152678988,
+      "learning_rate": 1.4171797712188927e-06,
+      "loss": 0.6582,
+      "step": 12811
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.132333056721575,
+      "learning_rate": 1.417091969685554e-06,
+      "loss": 0.6842,
+      "step": 12812
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.0028698697017726,
+      "learning_rate": 1.4170041642594595e-06,
+      "loss": 0.6497,
+      "step": 12813
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.4729649558904385,
+      "learning_rate": 1.4169163549414282e-06,
+      "loss": 0.6816,
+      "step": 12814
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 5.3770967237542,
+      "learning_rate": 1.4168285417322797e-06,
+      "loss": 0.6901,
+      "step": 12815
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8332202824314414,
+      "learning_rate": 1.4167407246328334e-06,
+      "loss": 0.6647,
+      "step": 12816
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5531462931256732,
+      "learning_rate": 1.4166529036439092e-06,
+      "loss": 0.6582,
+      "step": 12817
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9643552618447554,
+      "learning_rate": 1.4165650787663265e-06,
+      "loss": 0.6738,
+      "step": 12818
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.0278036777324715,
+      "learning_rate": 1.4164772500009052e-06,
+      "loss": 0.6738,
+      "step": 12819
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.3128279954682527,
+      "learning_rate": 1.4163894173484646e-06,
+      "loss": 0.6829,
+      "step": 12820
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.3340747681425427,
+      "learning_rate": 1.4163015808098252e-06,
+      "loss": 0.6582,
+      "step": 12821
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1719114589162707,
+      "learning_rate": 1.416213740385806e-06,
+      "loss": 0.6745,
+      "step": 12822
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1174859070991194,
+      "learning_rate": 1.4161258960772272e-06,
+      "loss": 0.6654,
+      "step": 12823
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.1924237224818084,
+      "learning_rate": 1.416038047884909e-06,
+      "loss": 0.6999,
+      "step": 12824
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.655616635525228,
+      "learning_rate": 1.4159501958096705e-06,
+      "loss": 0.6862,
+      "step": 12825
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.998199281477778,
+      "learning_rate": 1.4158623398523321e-06,
+      "loss": 0.7051,
+      "step": 12826
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.167181590111423,
+      "learning_rate": 1.4157744800137137e-06,
+      "loss": 0.6621,
+      "step": 12827
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8909382256695663,
+      "learning_rate": 1.4156866162946353e-06,
+      "loss": 0.6803,
+      "step": 12828
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.6075231180899392,
+      "learning_rate": 1.415598748695917e-06,
+      "loss": 0.6823,
+      "step": 12829
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8145782356447447,
+      "learning_rate": 1.4155108772183789e-06,
+      "loss": 0.6621,
+      "step": 12830
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.9332498596677217,
+      "learning_rate": 1.4154230018628408e-06,
+      "loss": 0.6725,
+      "step": 12831
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.064503614994999,
+      "learning_rate": 1.4153351226301234e-06,
+      "loss": 0.6641,
+      "step": 12832
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7823019366358687,
+      "learning_rate": 1.4152472395210461e-06,
+      "loss": 0.6647,
+      "step": 12833
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3206736326995125,
+      "learning_rate": 1.41515935253643e-06,
+      "loss": 0.6719,
+      "step": 12834
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.8204597575098326,
+      "learning_rate": 1.415071461677095e-06,
+      "loss": 0.6895,
+      "step": 12835
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 3.730970696853117,
+      "learning_rate": 1.414983566943861e-06,
+      "loss": 0.6647,
+      "step": 12836
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2174835178238852,
+      "learning_rate": 1.414895668337549e-06,
+      "loss": 0.6764,
+      "step": 12837
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.550310293747304,
+      "learning_rate": 1.4148077658589788e-06,
+      "loss": 0.6393,
+      "step": 12838
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 4.008411015393533,
+      "learning_rate": 1.4147198595089708e-06,
+      "loss": 0.6712,
+      "step": 12839
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.819962634928022,
+      "learning_rate": 1.4146319492883465e-06,
+      "loss": 0.6699,
+      "step": 12840
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.3571819828946903,
+      "learning_rate": 1.414544035197925e-06,
+      "loss": 0.6647,
+      "step": 12841
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.22479939515595,
+      "learning_rate": 1.4144561172385274e-06,
+      "loss": 0.6947,
+      "step": 12842
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9650257867846695,
+      "learning_rate": 1.4143681954109743e-06,
+      "loss": 0.6712,
+      "step": 12843
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.2730346073129004,
+      "learning_rate": 1.414280269716086e-06,
+      "loss": 0.653,
+      "step": 12844
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.4331414847104473,
+      "learning_rate": 1.4141923401546834e-06,
+      "loss": 0.6875,
+      "step": 12845
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.0043417730133763,
+      "learning_rate": 1.414104406727587e-06,
+      "loss": 0.6992,
+      "step": 12846
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.0591226138249645,
+      "learning_rate": 1.4140164694356177e-06,
+      "loss": 0.6536,
+      "step": 12847
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9342736579633029,
+      "learning_rate": 1.413928528279596e-06,
+      "loss": 0.6608,
+      "step": 12848
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.462075955160419,
+      "learning_rate": 1.4138405832603426e-06,
+      "loss": 0.681,
+      "step": 12849
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6472135320647452,
+      "learning_rate": 1.4137526343786786e-06,
+      "loss": 0.666,
+      "step": 12850
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.3013694368811815,
+      "learning_rate": 1.4136646816354246e-06,
+      "loss": 0.6706,
+      "step": 12851
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.044690145817431,
+      "learning_rate": 1.4135767250314015e-06,
+      "loss": 0.6628,
+      "step": 12852
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1990118591686472,
+      "learning_rate": 1.4134887645674308e-06,
+      "loss": 0.6836,
+      "step": 12853
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.0986219459123128,
+      "learning_rate": 1.4134008002443322e-06,
+      "loss": 0.6621,
+      "step": 12854
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.05927310124199,
+      "learning_rate": 1.4133128320629272e-06,
+      "loss": 0.6654,
+      "step": 12855
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.3873157022074376,
+      "learning_rate": 1.4132248600240378e-06,
+      "loss": 0.6882,
+      "step": 12856
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.237008892052023,
+      "learning_rate": 1.413136884128484e-06,
+      "loss": 0.6634,
+      "step": 12857
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.8517635012471803,
+      "learning_rate": 1.4130489043770868e-06,
+      "loss": 0.7038,
+      "step": 12858
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.101315478313295,
+      "learning_rate": 1.4129609207706676e-06,
+      "loss": 0.666,
+      "step": 12859
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.2127645355203613,
+      "learning_rate": 1.412872933310048e-06,
+      "loss": 0.6732,
+      "step": 12860
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.0698964675030822,
+      "learning_rate": 1.4127849419960484e-06,
+      "loss": 0.6621,
+      "step": 12861
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.0245245096816675,
+      "learning_rate": 1.4126969468294907e-06,
+      "loss": 0.7155,
+      "step": 12862
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.615125776529782,
+      "learning_rate": 1.4126089478111958e-06,
+      "loss": 0.6784,
+      "step": 12863
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.961480876184843,
+      "learning_rate": 1.4125209449419853e-06,
+      "loss": 0.6543,
+      "step": 12864
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2600715043981305,
+      "learning_rate": 1.4124329382226802e-06,
+      "loss": 0.6882,
+      "step": 12865
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.3809724563953707,
+      "learning_rate": 1.412344927654102e-06,
+      "loss": 0.6895,
+      "step": 12866
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.914697069775161,
+      "learning_rate": 1.4122569132370721e-06,
+      "loss": 0.6764,
+      "step": 12867
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.341231478099604,
+      "learning_rate": 1.4121688949724119e-06,
+      "loss": 0.7005,
+      "step": 12868
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.7018001302534818,
+      "learning_rate": 1.412080872860943e-06,
+      "loss": 0.7064,
+      "step": 12869
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.287566811220623,
+      "learning_rate": 1.411992846903487e-06,
+      "loss": 0.6992,
+      "step": 12870
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2292971035048894,
+      "learning_rate": 1.411904817100865e-06,
+      "loss": 0.6732,
+      "step": 12871
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.334554331780839,
+      "learning_rate": 1.4118167834538995e-06,
+      "loss": 0.6686,
+      "step": 12872
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7928656233812511,
+      "learning_rate": 1.411728745963411e-06,
+      "loss": 0.6686,
+      "step": 12873
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.823522285369333,
+      "learning_rate": 1.4116407046302222e-06,
+      "loss": 0.6745,
+      "step": 12874
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9204302544142324,
+      "learning_rate": 1.4115526594551537e-06,
+      "loss": 0.6582,
+      "step": 12875
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.37381560617873,
+      "learning_rate": 1.4114646104390283e-06,
+      "loss": 0.6432,
+      "step": 12876
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8279753927100064,
+      "learning_rate": 1.4113765575826671e-06,
+      "loss": 0.6699,
+      "step": 12877
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8477913008783866,
+      "learning_rate": 1.411288500886892e-06,
+      "loss": 0.6829,
+      "step": 12878
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.0663699515885163,
+      "learning_rate": 1.411200440352525e-06,
+      "loss": 0.6738,
+      "step": 12879
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9853245211016814,
+      "learning_rate": 1.4111123759803881e-06,
+      "loss": 0.6849,
+      "step": 12880
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8503554192209817,
+      "learning_rate": 1.4110243077713028e-06,
+      "loss": 0.668,
+      "step": 12881
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9145972571625973,
+      "learning_rate": 1.4109362357260916e-06,
+      "loss": 0.6654,
+      "step": 12882
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.85663776928085,
+      "learning_rate": 1.410848159845576e-06,
+      "loss": 0.653,
+      "step": 12883
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.766156145194664,
+      "learning_rate": 1.4107600801305781e-06,
+      "loss": 0.6497,
+      "step": 12884
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.8200247909413705,
+      "learning_rate": 1.4106719965819203e-06,
+      "loss": 0.6829,
+      "step": 12885
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 6.221062866506149,
+      "learning_rate": 1.4105839092004244e-06,
+      "loss": 0.6738,
+      "step": 12886
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.6127121222340897,
+      "learning_rate": 1.4104958179869124e-06,
+      "loss": 0.6855,
+      "step": 12887
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.732377403608726,
+      "learning_rate": 1.4104077229422068e-06,
+      "loss": 0.6556,
+      "step": 12888
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.042267278025713,
+      "learning_rate": 1.4103196240671292e-06,
+      "loss": 0.6816,
+      "step": 12889
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7676605747048009,
+      "learning_rate": 1.410231521362503e-06,
+      "loss": 0.6523,
+      "step": 12890
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.9687530656254271,
+      "learning_rate": 1.410143414829149e-06,
+      "loss": 0.6686,
+      "step": 12891
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.42470392888078,
+      "learning_rate": 1.4100553044678905e-06,
+      "loss": 0.679,
+      "step": 12892
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.837885231750733,
+      "learning_rate": 1.4099671902795497e-06,
+      "loss": 0.6478,
+      "step": 12893
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.251078782206496,
+      "learning_rate": 1.4098790722649487e-06,
+      "loss": 0.6647,
+      "step": 12894
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.4179307024056789,
+      "learning_rate": 1.40979095042491e-06,
+      "loss": 0.6667,
+      "step": 12895
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1391668040495924,
+      "learning_rate": 1.4097028247602564e-06,
+      "loss": 0.6667,
+      "step": 12896
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.757791684892955,
+      "learning_rate": 1.40961469527181e-06,
+      "loss": 0.6764,
+      "step": 12897
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 5.198972752996393,
+      "learning_rate": 1.4095265619603934e-06,
+      "loss": 0.6556,
+      "step": 12898
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.181530786616308,
+      "learning_rate": 1.409438424826829e-06,
+      "loss": 0.6628,
+      "step": 12899
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.300637119618101,
+      "learning_rate": 1.4093502838719397e-06,
+      "loss": 0.6465,
+      "step": 12900
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.3791507735309054,
+      "learning_rate": 1.4092621390965482e-06,
+      "loss": 0.6966,
+      "step": 12901
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.7854005524549554,
+      "learning_rate": 1.4091739905014768e-06,
+      "loss": 0.6634,
+      "step": 12902
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.4909999305237913,
+      "learning_rate": 1.4090858380875485e-06,
+      "loss": 0.627,
+      "step": 12903
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 3.6774974249949293,
+      "learning_rate": 1.4089976818555856e-06,
+      "loss": 0.6328,
+      "step": 12904
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 4.612259098741676,
+      "learning_rate": 1.4089095218064112e-06,
+      "loss": 0.668,
+      "step": 12905
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.1935824165342845,
+      "learning_rate": 1.4088213579408487e-06,
+      "loss": 0.6745,
+      "step": 12906
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.6945984563008722,
+      "learning_rate": 1.4087331902597194e-06,
+      "loss": 0.6406,
+      "step": 12907
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.2283066976881316,
+      "learning_rate": 1.4086450187638477e-06,
+      "loss": 0.7096,
+      "step": 12908
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5290579629443508,
+      "learning_rate": 1.4085568434540556e-06,
+      "loss": 0.6719,
+      "step": 12909
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.2807302060445385,
+      "learning_rate": 1.4084686643311666e-06,
+      "loss": 0.694,
+      "step": 12910
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.562921725962534,
+      "learning_rate": 1.4083804813960031e-06,
+      "loss": 0.6875,
+      "step": 12911
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.3633999926594944,
+      "learning_rate": 1.4082922946493887e-06,
+      "loss": 0.666,
+      "step": 12912
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.675531277705713,
+      "learning_rate": 1.408204104092146e-06,
+      "loss": 0.6829,
+      "step": 12913
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.859949010856096,
+      "learning_rate": 1.4081159097250984e-06,
+      "loss": 0.6406,
+      "step": 12914
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.933057954094087,
+      "learning_rate": 1.4080277115490691e-06,
+      "loss": 0.6777,
+      "step": 12915
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.4741039960710514,
+      "learning_rate": 1.407939509564881e-06,
+      "loss": 0.6764,
+      "step": 12916
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1129720583027036,
+      "learning_rate": 1.4078513037733573e-06,
+      "loss": 0.6862,
+      "step": 12917
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.606389899155841,
+      "learning_rate": 1.407763094175321e-06,
+      "loss": 0.6738,
+      "step": 12918
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1909920244091108,
+      "learning_rate": 1.4076748807715966e-06,
+      "loss": 0.6628,
+      "step": 12919
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0837231991234668,
+      "learning_rate": 1.4075866635630055e-06,
+      "loss": 0.6361,
+      "step": 12920
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.8043390509718726,
+      "learning_rate": 1.4074984425503725e-06,
+      "loss": 0.6608,
+      "step": 12921
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.120562407144169,
+      "learning_rate": 1.4074102177345205e-06,
+      "loss": 0.6641,
+      "step": 12922
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.513662488142481,
+      "learning_rate": 1.407321989116273e-06,
+      "loss": 0.6628,
+      "step": 12923
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.5659922662661594,
+      "learning_rate": 1.4072337566964533e-06,
+      "loss": 0.6517,
+      "step": 12924
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.468196060997647,
+      "learning_rate": 1.4071455204758845e-06,
+      "loss": 0.6855,
+      "step": 12925
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.2931550142613448,
+      "learning_rate": 1.4070572804553908e-06,
+      "loss": 0.6719,
+      "step": 12926
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.1938515045200218,
+      "learning_rate": 1.4069690366357958e-06,
+      "loss": 0.6556,
+      "step": 12927
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.153564590367651,
+      "learning_rate": 1.4068807890179225e-06,
+      "loss": 0.6973,
+      "step": 12928
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.089524777761413,
+      "learning_rate": 1.4067925376025949e-06,
+      "loss": 0.6745,
+      "step": 12929
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.7027666680720435,
+      "learning_rate": 1.4067042823906364e-06,
+      "loss": 0.6921,
+      "step": 12930
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.3198041203747306,
+      "learning_rate": 1.406616023382871e-06,
+      "loss": 0.6693,
+      "step": 12931
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.3375118359636917,
+      "learning_rate": 1.4065277605801224e-06,
+      "loss": 0.6426,
+      "step": 12932
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7991693204871708,
+      "learning_rate": 1.4064394939832142e-06,
+      "loss": 0.653,
+      "step": 12933
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.4204940914895943,
+      "learning_rate": 1.4063512235929703e-06,
+      "loss": 0.6654,
+      "step": 12934
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0726416915806214,
+      "learning_rate": 1.4062629494102142e-06,
+      "loss": 0.668,
+      "step": 12935
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1390988874169017,
+      "learning_rate": 1.4061746714357706e-06,
+      "loss": 0.6497,
+      "step": 12936
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.912124290578713,
+      "learning_rate": 1.4060863896704622e-06,
+      "loss": 0.6706,
+      "step": 12937
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.83689640047716,
+      "learning_rate": 1.4059981041151142e-06,
+      "loss": 0.6517,
+      "step": 12938
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5775828480944358,
+      "learning_rate": 1.4059098147705499e-06,
+      "loss": 0.6921,
+      "step": 12939
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8116653915536467,
+      "learning_rate": 1.4058215216375932e-06,
+      "loss": 0.6465,
+      "step": 12940
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.861114840372033,
+      "learning_rate": 1.4057332247170684e-06,
+      "loss": 0.6458,
+      "step": 12941
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7743631419151435,
+      "learning_rate": 1.4056449240097994e-06,
+      "loss": 0.6569,
+      "step": 12942
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.6151783281245145,
+      "learning_rate": 1.405556619516611e-06,
+      "loss": 0.7064,
+      "step": 12943
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9770039889617208,
+      "learning_rate": 1.4054683112383261e-06,
+      "loss": 0.6452,
+      "step": 12944
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.213411949919454,
+      "learning_rate": 1.40537999917577e-06,
+      "loss": 0.6673,
+      "step": 12945
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.0124629760695494,
+      "learning_rate": 1.4052916833297666e-06,
+      "loss": 0.6979,
+      "step": 12946
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.814460883813714,
+      "learning_rate": 1.4052033637011395e-06,
+      "loss": 0.6908,
+      "step": 12947
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.054591646304856,
+      "learning_rate": 1.4051150402907142e-06,
+      "loss": 0.6595,
+      "step": 12948
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1597905891884412,
+      "learning_rate": 1.405026713099314e-06,
+      "loss": 0.6543,
+      "step": 12949
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.8573500206671358,
+      "learning_rate": 1.404938382127764e-06,
+      "loss": 0.6523,
+      "step": 12950
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.9791796231654553,
+      "learning_rate": 1.404850047376888e-06,
+      "loss": 0.668,
+      "step": 12951
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.375457972653346,
+      "learning_rate": 1.404761708847511e-06,
+      "loss": 0.7188,
+      "step": 12952
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.133665185235993,
+      "learning_rate": 1.4046733665404568e-06,
+      "loss": 0.6178,
+      "step": 12953
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.9647435860642644,
+      "learning_rate": 1.4045850204565505e-06,
+      "loss": 0.6615,
+      "step": 12954
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.7788507786010523,
+      "learning_rate": 1.4044966705966164e-06,
+      "loss": 0.6875,
+      "step": 12955
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.060748404922829,
+      "learning_rate": 1.4044083169614793e-06,
+      "loss": 0.6413,
+      "step": 12956
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0790896448478269,
+      "learning_rate": 1.4043199595519632e-06,
+      "loss": 0.6602,
+      "step": 12957
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1388737651521157,
+      "learning_rate": 1.4042315983688933e-06,
+      "loss": 0.6725,
+      "step": 12958
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9893272889693202,
+      "learning_rate": 1.4041432334130944e-06,
+      "loss": 0.6589,
+      "step": 12959
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.804782873637079,
+      "learning_rate": 1.4040548646853907e-06,
+      "loss": 0.6901,
+      "step": 12960
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.279653938813796,
+      "learning_rate": 1.4039664921866073e-06,
+      "loss": 0.6523,
+      "step": 12961
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.9700494256811287,
+      "learning_rate": 1.403878115917569e-06,
+      "loss": 0.7044,
+      "step": 12962
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1083376894018637,
+      "learning_rate": 1.4037897358791003e-06,
+      "loss": 0.6758,
+      "step": 12963
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.4326179670383925,
+      "learning_rate": 1.4037013520720264e-06,
+      "loss": 0.7253,
+      "step": 12964
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.549961608925722,
+      "learning_rate": 1.403612964497172e-06,
+      "loss": 0.6784,
+      "step": 12965
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 3.527402710336413,
+      "learning_rate": 1.4035245731553622e-06,
+      "loss": 0.6823,
+      "step": 12966
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0567569261194847,
+      "learning_rate": 1.403436178047422e-06,
+      "loss": 0.6523,
+      "step": 12967
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.4717973074802444,
+      "learning_rate": 1.403347779174176e-06,
+      "loss": 0.6836,
+      "step": 12968
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.577390632978276,
+      "learning_rate": 1.4032593765364497e-06,
+      "loss": 0.6777,
+      "step": 12969
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 5.716679138820634,
+      "learning_rate": 1.403170970135068e-06,
+      "loss": 0.6895,
+      "step": 12970
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.5320530849235738,
+      "learning_rate": 1.4030825599708557e-06,
+      "loss": 0.6803,
+      "step": 12971
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.3360911186466393,
+      "learning_rate": 1.4029941460446385e-06,
+      "loss": 0.6595,
+      "step": 12972
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.9061917119765863,
+      "learning_rate": 1.4029057283572412e-06,
+      "loss": 0.6517,
+      "step": 12973
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 4.294093206649457,
+      "learning_rate": 1.4028173069094894e-06,
+      "loss": 0.6686,
+      "step": 12974
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.130839699090959,
+      "learning_rate": 1.4027288817022074e-06,
+      "loss": 0.6484,
+      "step": 12975
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3606001508811214,
+      "learning_rate": 1.4026404527362217e-06,
+      "loss": 0.6465,
+      "step": 12976
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.007869332511742,
+      "learning_rate": 1.4025520200123569e-06,
+      "loss": 0.6517,
+      "step": 12977
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.680112088534043,
+      "learning_rate": 1.4024635835314383e-06,
+      "loss": 0.6738,
+      "step": 12978
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.4838448230457835,
+      "learning_rate": 1.4023751432942917e-06,
+      "loss": 0.6523,
+      "step": 12979
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9250743742039668,
+      "learning_rate": 1.4022866993017424e-06,
+      "loss": 0.6719,
+      "step": 12980
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9871500624402798,
+      "learning_rate": 1.4021982515546154e-06,
+      "loss": 0.6784,
+      "step": 12981
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9700511344183806,
+      "learning_rate": 1.402109800053737e-06,
+      "loss": 0.694,
+      "step": 12982
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.7781085955182716,
+      "learning_rate": 1.4020213447999318e-06,
+      "loss": 0.7057,
+      "step": 12983
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3076299645107081,
+      "learning_rate": 1.401932885794026e-06,
+      "loss": 0.6927,
+      "step": 12984
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4825276980398217,
+      "learning_rate": 1.401844423036845e-06,
+      "loss": 0.6335,
+      "step": 12985
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.882649728652866,
+      "learning_rate": 1.4017559565292147e-06,
+      "loss": 0.6862,
+      "step": 12986
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.6843360979354545,
+      "learning_rate": 1.4016674862719603e-06,
+      "loss": 0.6921,
+      "step": 12987
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.115115484543992,
+      "learning_rate": 1.401579012265908e-06,
+      "loss": 0.6706,
+      "step": 12988
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8322551333400028,
+      "learning_rate": 1.401490534511883e-06,
+      "loss": 0.6979,
+      "step": 12989
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7319100197810424,
+      "learning_rate": 1.4014020530107116e-06,
+      "loss": 0.6628,
+      "step": 12990
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5307410278525715,
+      "learning_rate": 1.401313567763219e-06,
+      "loss": 0.6693,
+      "step": 12991
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9073259085412586,
+      "learning_rate": 1.4012250787702316e-06,
+      "loss": 0.6973,
+      "step": 12992
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.3005241580722213,
+      "learning_rate": 1.401136586032575e-06,
+      "loss": 0.6413,
+      "step": 12993
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.4706502333614084,
+      "learning_rate": 1.401048089551075e-06,
+      "loss": 0.6882,
+      "step": 12994
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8382078504064175,
+      "learning_rate": 1.400959589326558e-06,
+      "loss": 0.6732,
+      "step": 12995
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.5986676067567465,
+      "learning_rate": 1.4008710853598495e-06,
+      "loss": 0.6777,
+      "step": 12996
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.007064566821334,
+      "learning_rate": 1.4007825776517756e-06,
+      "loss": 0.6576,
+      "step": 12997
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.4382264318520352,
+      "learning_rate": 1.4006940662031628e-06,
+      "loss": 0.6895,
+      "step": 12998
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.618156518032833,
+      "learning_rate": 1.400605551014837e-06,
+      "loss": 0.6777,
+      "step": 12999
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.004898749359389,
+      "learning_rate": 1.4005170320876234e-06,
+      "loss": 0.6784,
+      "step": 13000
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0499232472263913,
+      "learning_rate": 1.4004285094223496e-06,
+      "loss": 0.6569,
+      "step": 13001
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.8426641609247096,
+      "learning_rate": 1.4003399830198407e-06,
+      "loss": 0.6784,
+      "step": 13002
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.1201565253939374,
+      "learning_rate": 1.4002514528809234e-06,
+      "loss": 0.6745,
+      "step": 13003
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.8895878746251799,
+      "learning_rate": 1.4001629190064241e-06,
+      "loss": 0.6829,
+      "step": 13004
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.3717456429961645,
+      "learning_rate": 1.400074381397169e-06,
+      "loss": 0.6816,
+      "step": 13005
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 5.519106208365375,
+      "learning_rate": 1.3999858400539842e-06,
+      "loss": 0.6947,
+      "step": 13006
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6335650651043576,
+      "learning_rate": 1.3998972949776959e-06,
+      "loss": 0.6751,
+      "step": 13007
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.176157581354294,
+      "learning_rate": 1.3998087461691308e-06,
+      "loss": 0.6947,
+      "step": 13008
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7433732386565102,
+      "learning_rate": 1.3997201936291157e-06,
+      "loss": 0.6634,
+      "step": 13009
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.9388220081738089,
+      "learning_rate": 1.3996316373584764e-06,
+      "loss": 0.6849,
+      "step": 13010
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.208537379164724,
+      "learning_rate": 1.39954307735804e-06,
+      "loss": 0.6647,
+      "step": 13011
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5988103687199673,
+      "learning_rate": 1.3994545136286325e-06,
+      "loss": 0.681,
+      "step": 13012
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8733823380783254,
+      "learning_rate": 1.3993659461710806e-06,
+      "loss": 0.6738,
+      "step": 13013
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.054657689386541,
+      "learning_rate": 1.399277374986211e-06,
+      "loss": 0.6628,
+      "step": 13014
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8481767913505336,
+      "learning_rate": 1.3991888000748503e-06,
+      "loss": 0.6497,
+      "step": 13015
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.5160707374221145,
+      "learning_rate": 1.3991002214378253e-06,
+      "loss": 0.6927,
+      "step": 13016
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.5268229650791163,
+      "learning_rate": 1.399011639075963e-06,
+      "loss": 0.6979,
+      "step": 13017
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.309707657598006,
+      "learning_rate": 1.3989230529900892e-06,
+      "loss": 0.6745,
+      "step": 13018
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.679439791284444,
+      "learning_rate": 1.3988344631810317e-06,
+      "loss": 0.6693,
+      "step": 13019
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.2073480615895846,
+      "learning_rate": 1.3987458696496166e-06,
+      "loss": 0.6582,
+      "step": 13020
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.2081267374945885,
+      "learning_rate": 1.398657272396671e-06,
+      "loss": 0.6576,
+      "step": 13021
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 4.022079936230449,
+      "learning_rate": 1.3985686714230224e-06,
+      "loss": 0.679,
+      "step": 13022
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9160815211878806,
+      "learning_rate": 1.3984800667294967e-06,
+      "loss": 0.6816,
+      "step": 13023
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.2946104780944943,
+      "learning_rate": 1.3983914583169214e-06,
+      "loss": 0.6569,
+      "step": 13024
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.373213272268625,
+      "learning_rate": 1.3983028461861232e-06,
+      "loss": 0.6647,
+      "step": 13025
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8708884974728588,
+      "learning_rate": 1.3982142303379295e-06,
+      "loss": 0.6621,
+      "step": 13026
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.1847308598845583,
+      "learning_rate": 1.398125610773167e-06,
+      "loss": 0.6803,
+      "step": 13027
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0822332260290375,
+      "learning_rate": 1.3980369874926632e-06,
+      "loss": 0.6471,
+      "step": 13028
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.646547094555071,
+      "learning_rate": 1.3979483604972448e-06,
+      "loss": 0.6686,
+      "step": 13029
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.688953708640687,
+      "learning_rate": 1.3978597297877393e-06,
+      "loss": 0.6777,
+      "step": 13030
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0232760253975535,
+      "learning_rate": 1.3977710953649734e-06,
+      "loss": 0.6556,
+      "step": 13031
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7486162464727228,
+      "learning_rate": 1.3976824572297752e-06,
+      "loss": 0.6725,
+      "step": 13032
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.7157962853955957,
+      "learning_rate": 1.3975938153829712e-06,
+      "loss": 0.6771,
+      "step": 13033
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.2108884511592215,
+      "learning_rate": 1.397505169825389e-06,
+      "loss": 0.6517,
+      "step": 13034
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.833517210487151,
+      "learning_rate": 1.397416520557856e-06,
+      "loss": 0.6667,
+      "step": 13035
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0509131802620912,
+      "learning_rate": 1.3973278675811992e-06,
+      "loss": 0.6536,
+      "step": 13036
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5672019231656136,
+      "learning_rate": 1.3972392108962463e-06,
+      "loss": 0.6953,
+      "step": 13037
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0102873526197675,
+      "learning_rate": 1.397150550503825e-06,
+      "loss": 0.6589,
+      "step": 13038
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 3.4165500699195763,
+      "learning_rate": 1.3970618864047623e-06,
+      "loss": 0.668,
+      "step": 13039
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3413084775334365,
+      "learning_rate": 1.396973218599886e-06,
+      "loss": 0.6667,
+      "step": 13040
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3711380218950424,
+      "learning_rate": 1.3968845470900234e-06,
+      "loss": 0.666,
+      "step": 13041
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.7089897460489842,
+      "learning_rate": 1.3967958718760024e-06,
+      "loss": 0.6673,
+      "step": 13042
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.280695729365668,
+      "learning_rate": 1.3967071929586502e-06,
+      "loss": 0.6732,
+      "step": 13043
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.9949438645596862,
+      "learning_rate": 1.3966185103387949e-06,
+      "loss": 0.6602,
+      "step": 13044
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8783663103416446,
+      "learning_rate": 1.3965298240172639e-06,
+      "loss": 0.6647,
+      "step": 13045
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.446974862494971,
+      "learning_rate": 1.396441133994885e-06,
+      "loss": 0.6953,
+      "step": 13046
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1010098251538805,
+      "learning_rate": 1.3963524402724858e-06,
+      "loss": 0.6745,
+      "step": 13047
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.1651497188246407,
+      "learning_rate": 1.3962637428508944e-06,
+      "loss": 0.6634,
+      "step": 13048
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.2143425374059116,
+      "learning_rate": 1.3961750417309385e-06,
+      "loss": 0.6608,
+      "step": 13049
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 5.0568914488888295,
+      "learning_rate": 1.3960863369134455e-06,
+      "loss": 0.7168,
+      "step": 13050
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2885133260773307,
+      "learning_rate": 1.395997628399244e-06,
+      "loss": 0.7012,
+      "step": 13051
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2896791254118218,
+      "learning_rate": 1.3959089161891617e-06,
+      "loss": 0.6367,
+      "step": 13052
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.051908601790371,
+      "learning_rate": 1.3958202002840263e-06,
+      "loss": 0.6803,
+      "step": 13053
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.9873603722836064,
+      "learning_rate": 1.3957314806846661e-06,
+      "loss": 0.6608,
+      "step": 13054
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9760100013890756,
+      "learning_rate": 1.395642757391909e-06,
+      "loss": 0.6562,
+      "step": 13055
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.8317980113382655,
+      "learning_rate": 1.395554030406583e-06,
+      "loss": 0.6849,
+      "step": 13056
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.8945998118287872,
+      "learning_rate": 1.395465299729516e-06,
+      "loss": 0.6771,
+      "step": 13057
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.2662912855430304,
+      "learning_rate": 1.3953765653615368e-06,
+      "loss": 0.6784,
+      "step": 13058
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6826006105588256,
+      "learning_rate": 1.3952878273034728e-06,
+      "loss": 0.6673,
+      "step": 13059
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 6.696906772554908,
+      "learning_rate": 1.3951990855561527e-06,
+      "loss": 0.6719,
+      "step": 13060
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2750814783292144,
+      "learning_rate": 1.3951103401204047e-06,
+      "loss": 0.6576,
+      "step": 13061
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.902225770075876,
+      "learning_rate": 1.3950215909970568e-06,
+      "loss": 0.6986,
+      "step": 13062
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.0064871111464058,
+      "learning_rate": 1.3949328381869375e-06,
+      "loss": 0.6634,
+      "step": 13063
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.1457112451343336,
+      "learning_rate": 1.3948440816908751e-06,
+      "loss": 0.7057,
+      "step": 13064
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.7333004143801638,
+      "learning_rate": 1.394755321509698e-06,
+      "loss": 0.6816,
+      "step": 13065
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5229215328055532,
+      "learning_rate": 1.3946665576442345e-06,
+      "loss": 0.6901,
+      "step": 13066
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9535683409885687,
+      "learning_rate": 1.394577790095313e-06,
+      "loss": 0.6576,
+      "step": 13067
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.330870794994635,
+      "learning_rate": 1.3944890188637626e-06,
+      "loss": 0.6536,
+      "step": 13068
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.9571913183692617,
+      "learning_rate": 1.3944002439504109e-06,
+      "loss": 0.6699,
+      "step": 13069
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.4119003786679363,
+      "learning_rate": 1.3943114653560868e-06,
+      "loss": 0.6719,
+      "step": 13070
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.0163179258725976,
+      "learning_rate": 1.394222683081619e-06,
+      "loss": 0.6895,
+      "step": 13071
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2855142167763955,
+      "learning_rate": 1.3941338971278363e-06,
+      "loss": 0.6908,
+      "step": 13072
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.546058476378657,
+      "learning_rate": 1.3940451074955665e-06,
+      "loss": 0.6719,
+      "step": 13073
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.6357448835039996,
+      "learning_rate": 1.3939563141856392e-06,
+      "loss": 0.679,
+      "step": 13074
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8564914232953225,
+      "learning_rate": 1.393867517198883e-06,
+      "loss": 0.6576,
+      "step": 13075
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9071785373658633,
+      "learning_rate": 1.393778716536126e-06,
+      "loss": 0.6732,
+      "step": 13076
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.300317376278429,
+      "learning_rate": 1.3936899121981978e-06,
+      "loss": 0.6732,
+      "step": 13077
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1988623480537373,
+      "learning_rate": 1.3936011041859266e-06,
+      "loss": 0.6745,
+      "step": 13078
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.3798099795015593,
+      "learning_rate": 1.3935122925001414e-06,
+      "loss": 0.6745,
+      "step": 13079
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0058123472701588,
+      "learning_rate": 1.3934234771416713e-06,
+      "loss": 0.6758,
+      "step": 13080
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8129554475061969,
+      "learning_rate": 1.3933346581113452e-06,
+      "loss": 0.6829,
+      "step": 13081
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.7484809883221795,
+      "learning_rate": 1.3932458354099919e-06,
+      "loss": 0.6875,
+      "step": 13082
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.519004972379638,
+      "learning_rate": 1.3931570090384403e-06,
+      "loss": 0.7064,
+      "step": 13083
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.160026838056003,
+      "learning_rate": 1.3930681789975196e-06,
+      "loss": 0.6842,
+      "step": 13084
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.8859898723258035,
+      "learning_rate": 1.3929793452880592e-06,
+      "loss": 0.6921,
+      "step": 13085
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.078322265771461,
+      "learning_rate": 1.3928905079108874e-06,
+      "loss": 0.6608,
+      "step": 13086
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8314592901630763,
+      "learning_rate": 1.3928016668668335e-06,
+      "loss": 0.6868,
+      "step": 13087
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.011609729059014,
+      "learning_rate": 1.3927128221567277e-06,
+      "loss": 0.6829,
+      "step": 13088
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.1792506195218064,
+      "learning_rate": 1.392623973781398e-06,
+      "loss": 0.6738,
+      "step": 13089
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.3903704741312586,
+      "learning_rate": 1.392535121741674e-06,
+      "loss": 0.6615,
+      "step": 13090
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1796127725228467,
+      "learning_rate": 1.392446266038385e-06,
+      "loss": 0.6419,
+      "step": 13091
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8068751588803452,
+      "learning_rate": 1.39235740667236e-06,
+      "loss": 0.6562,
+      "step": 13092
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.036172919737143,
+      "learning_rate": 1.392268543644429e-06,
+      "loss": 0.681,
+      "step": 13093
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2745799955675845,
+      "learning_rate": 1.3921796769554207e-06,
+      "loss": 0.6823,
+      "step": 13094
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.143881210542724,
+      "learning_rate": 1.3920908066061652e-06,
+      "loss": 0.7005,
+      "step": 13095
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.38398899856708,
+      "learning_rate": 1.3920019325974915e-06,
+      "loss": 0.6764,
+      "step": 13096
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.0902923372854936,
+      "learning_rate": 1.3919130549302286e-06,
+      "loss": 0.6745,
+      "step": 13097
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.9406895781887081,
+      "learning_rate": 1.391824173605207e-06,
+      "loss": 0.6751,
+      "step": 13098
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.2753990452346144,
+      "learning_rate": 1.3917352886232556e-06,
+      "loss": 0.668,
+      "step": 13099
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 3.284880165958848,
+      "learning_rate": 1.3916463999852037e-06,
+      "loss": 0.7096,
+      "step": 13100
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.795728699650396,
+      "learning_rate": 1.391557507691882e-06,
+      "loss": 0.6647,
+      "step": 13101
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.6248243021559343,
+      "learning_rate": 1.391468611744119e-06,
+      "loss": 0.6849,
+      "step": 13102
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0637146523898546,
+      "learning_rate": 1.3913797121427448e-06,
+      "loss": 0.6771,
+      "step": 13103
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1357029981223377,
+      "learning_rate": 1.3912908088885895e-06,
+      "loss": 0.6823,
+      "step": 13104
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.5272284356302166,
+      "learning_rate": 1.3912019019824823e-06,
+      "loss": 0.681,
+      "step": 13105
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.2746619645100434,
+      "learning_rate": 1.391112991425253e-06,
+      "loss": 0.6647,
+      "step": 13106
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 2.6063988878298967,
+      "learning_rate": 1.3910240772177318e-06,
+      "loss": 0.6699,
+      "step": 13107
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 4.1265737781436345,
+      "learning_rate": 1.3909351593607482e-06,
+      "loss": 0.6816,
+      "step": 13108
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.83113343939977,
+      "learning_rate": 1.3908462378551323e-06,
+      "loss": 0.681,
+      "step": 13109
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.515268103637715,
+      "learning_rate": 1.3907573127017137e-06,
+      "loss": 0.6777,
+      "step": 13110
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5252399816598072,
+      "learning_rate": 1.3906683839013225e-06,
+      "loss": 0.6432,
+      "step": 13111
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5511555494471279,
+      "learning_rate": 1.390579451454789e-06,
+      "loss": 0.668,
+      "step": 13112
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8325315577252432,
+      "learning_rate": 1.3904905153629425e-06,
+      "loss": 0.6706,
+      "step": 13113
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8408072520673661,
+      "learning_rate": 1.390401575626614e-06,
+      "loss": 0.6517,
+      "step": 13114
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.511899073547349,
+      "learning_rate": 1.3903126322466328e-06,
+      "loss": 0.6556,
+      "step": 13115
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8922270045263905,
+      "learning_rate": 1.3902236852238292e-06,
+      "loss": 0.6634,
+      "step": 13116
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.76920131473609,
+      "learning_rate": 1.3901347345590335e-06,
+      "loss": 0.6777,
+      "step": 13117
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.2878478321892874,
+      "learning_rate": 1.3900457802530762e-06,
+      "loss": 0.6706,
+      "step": 13118
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.892206592002078,
+      "learning_rate": 1.3899568223067863e-06,
+      "loss": 0.694,
+      "step": 13119
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.5890120773584577,
+      "learning_rate": 1.3898678607209954e-06,
+      "loss": 0.6595,
+      "step": 13120
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.3933638462878735,
+      "learning_rate": 1.3897788954965335e-06,
+      "loss": 0.653,
+      "step": 13121
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.459006509337071,
+      "learning_rate": 1.3896899266342306e-06,
+      "loss": 0.6953,
+      "step": 13122
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.735114632504842,
+      "learning_rate": 1.389600954134917e-06,
+      "loss": 0.6628,
+      "step": 13123
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.75920770254098,
+      "learning_rate": 1.3895119779994234e-06,
+      "loss": 0.6556,
+      "step": 13124
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.2474646574131643,
+      "learning_rate": 1.3894229982285799e-06,
+      "loss": 0.668,
+      "step": 13125
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.8630650397340593,
+      "learning_rate": 1.3893340148232168e-06,
+      "loss": 0.7116,
+      "step": 13126
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1045986978929963,
+      "learning_rate": 1.3892450277841655e-06,
+      "loss": 0.6816,
+      "step": 13127
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9421509254326832,
+      "learning_rate": 1.3891560371122556e-06,
+      "loss": 0.6484,
+      "step": 13128
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.7513313619330004,
+      "learning_rate": 1.389067042808318e-06,
+      "loss": 0.6849,
+      "step": 13129
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.105890477144193,
+      "learning_rate": 1.3889780448731835e-06,
+      "loss": 0.6875,
+      "step": 13130
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.4650765368949865,
+      "learning_rate": 1.3888890433076825e-06,
+      "loss": 0.6784,
+      "step": 13131
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.7582644012913577,
+      "learning_rate": 1.3888000381126453e-06,
+      "loss": 0.6836,
+      "step": 13132
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 4.0050203163696985,
+      "learning_rate": 1.3887110292889033e-06,
+      "loss": 0.6719,
+      "step": 13133
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.872927287679603,
+      "learning_rate": 1.388622016837287e-06,
+      "loss": 0.6602,
+      "step": 13134
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6479763643269787,
+      "learning_rate": 1.3885330007586264e-06,
+      "loss": 0.6732,
+      "step": 13135
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.701994989382979,
+      "learning_rate": 1.3884439810537537e-06,
+      "loss": 0.6862,
+      "step": 13136
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.7879132325794633,
+      "learning_rate": 1.3883549577234984e-06,
+      "loss": 0.638,
+      "step": 13137
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9656468801344567,
+      "learning_rate": 1.3882659307686923e-06,
+      "loss": 0.6908,
+      "step": 13138
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.3783591551739733,
+      "learning_rate": 1.3881769001901658e-06,
+      "loss": 0.6543,
+      "step": 13139
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.5278843684696577,
+      "learning_rate": 1.3880878659887498e-06,
+      "loss": 0.6829,
+      "step": 13140
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1453765078810896,
+      "learning_rate": 1.387998828165276e-06,
+      "loss": 0.6576,
+      "step": 13141
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.5579320663803673,
+      "learning_rate": 1.3879097867205743e-06,
+      "loss": 0.6478,
+      "step": 13142
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.99353246488745,
+      "learning_rate": 1.3878207416554763e-06,
+      "loss": 0.6439,
+      "step": 13143
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9580310242903272,
+      "learning_rate": 1.387731692970813e-06,
+      "loss": 0.6621,
+      "step": 13144
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2975781717233958,
+      "learning_rate": 1.3876426406674157e-06,
+      "loss": 0.679,
+      "step": 13145
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9398149725905756,
+      "learning_rate": 1.3875535847461156e-06,
+      "loss": 0.623,
+      "step": 13146
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9265996880535987,
+      "learning_rate": 1.3874645252077433e-06,
+      "loss": 0.6908,
+      "step": 13147
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1245037137277158,
+      "learning_rate": 1.38737546205313e-06,
+      "loss": 0.6849,
+      "step": 13148
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.0208729632213278,
+      "learning_rate": 1.3872863952831078e-06,
+      "loss": 0.6693,
+      "step": 13149
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.195400548741567,
+      "learning_rate": 1.387197324898507e-06,
+      "loss": 0.6478,
+      "step": 13150
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.6188282028523768,
+      "learning_rate": 1.3871082509001598e-06,
+      "loss": 0.6927,
+      "step": 13151
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.893684783272169,
+      "learning_rate": 1.3870191732888965e-06,
+      "loss": 0.6621,
+      "step": 13152
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.502459587131113,
+      "learning_rate": 1.3869300920655492e-06,
+      "loss": 0.6745,
+      "step": 13153
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2050482330638326,
+      "learning_rate": 1.3868410072309495e-06,
+      "loss": 0.6725,
+      "step": 13154
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.2608567759179077,
+      "learning_rate": 1.3867519187859283e-06,
+      "loss": 0.6673,
+      "step": 13155
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.4409139276091314,
+      "learning_rate": 1.3866628267313173e-06,
+      "loss": 0.696,
+      "step": 13156
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.059443438028437,
+      "learning_rate": 1.3865737310679478e-06,
+      "loss": 0.6862,
+      "step": 13157
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.255717521821227,
+      "learning_rate": 1.3864846317966512e-06,
+      "loss": 0.6686,
+      "step": 13158
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.050919089923011,
+      "learning_rate": 1.3863955289182596e-06,
+      "loss": 0.6745,
+      "step": 13159
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1311902968362824,
+      "learning_rate": 1.3863064224336044e-06,
+      "loss": 0.6517,
+      "step": 13160
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1562801470574693,
+      "learning_rate": 1.3862173123435173e-06,
+      "loss": 0.6999,
+      "step": 13161
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1731222596664979,
+      "learning_rate": 1.3861281986488297e-06,
+      "loss": 0.6751,
+      "step": 13162
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1856178564877975,
+      "learning_rate": 1.3860390813503734e-06,
+      "loss": 0.6517,
+      "step": 13163
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.642262272088208,
+      "learning_rate": 1.3859499604489803e-06,
+      "loss": 0.6504,
+      "step": 13164
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 5.810885422932143,
+      "learning_rate": 1.3858608359454824e-06,
+      "loss": 0.6608,
+      "step": 13165
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1638360631666165,
+      "learning_rate": 1.3857717078407108e-06,
+      "loss": 0.6829,
+      "step": 13166
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.5395766179572554,
+      "learning_rate": 1.385682576135498e-06,
+      "loss": 0.6882,
+      "step": 13167
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.777646845296438,
+      "learning_rate": 1.3855934408306755e-06,
+      "loss": 0.7051,
+      "step": 13168
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.011235879218841,
+      "learning_rate": 1.3855043019270752e-06,
+      "loss": 0.6986,
+      "step": 13169
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.33711726509822,
+      "learning_rate": 1.3854151594255292e-06,
+      "loss": 0.6654,
+      "step": 13170
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.7830676772235372,
+      "learning_rate": 1.3853260133268695e-06,
+      "loss": 0.6758,
+      "step": 13171
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9769644179965717,
+      "learning_rate": 1.385236863631928e-06,
+      "loss": 0.6777,
+      "step": 13172
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 3.616049647981659,
+      "learning_rate": 1.3851477103415368e-06,
+      "loss": 0.6829,
+      "step": 13173
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.8542830946942923,
+      "learning_rate": 1.385058553456528e-06,
+      "loss": 0.6738,
+      "step": 13174
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9778737772585646,
+      "learning_rate": 1.3849693929777337e-06,
+      "loss": 0.679,
+      "step": 13175
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9761992751399015,
+      "learning_rate": 1.3848802289059857e-06,
+      "loss": 0.6582,
+      "step": 13176
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.307165336539664,
+      "learning_rate": 1.3847910612421169e-06,
+      "loss": 0.6901,
+      "step": 13177
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.2284093407386933,
+      "learning_rate": 1.384701889986959e-06,
+      "loss": 0.6725,
+      "step": 13178
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.386481402686784,
+      "learning_rate": 1.384612715141344e-06,
+      "loss": 0.6576,
+      "step": 13179
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.558277434142353,
+      "learning_rate": 1.384523536706105e-06,
+      "loss": 0.6823,
+      "step": 13180
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.021679971990669,
+      "learning_rate": 1.3844343546820736e-06,
+      "loss": 0.6986,
+      "step": 13181
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.2828272483272602,
+      "learning_rate": 1.3843451690700823e-06,
+      "loss": 0.6589,
+      "step": 13182
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.938842276573048,
+      "learning_rate": 1.3842559798709636e-06,
+      "loss": 0.6738,
+      "step": 13183
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.3382242041803145,
+      "learning_rate": 1.38416678708555e-06,
+      "loss": 0.6758,
+      "step": 13184
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.2621561651880437,
+      "learning_rate": 1.3840775907146735e-06,
+      "loss": 0.6901,
+      "step": 13185
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.415910599888298,
+      "learning_rate": 1.3839883907591673e-06,
+      "loss": 0.7057,
+      "step": 13186
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.683747381356965,
+      "learning_rate": 1.3838991872198632e-06,
+      "loss": 0.6719,
+      "step": 13187
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.1429588189113264,
+      "learning_rate": 1.3838099800975943e-06,
+      "loss": 0.6452,
+      "step": 13188
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9169902145502226,
+      "learning_rate": 1.3837207693931925e-06,
+      "loss": 0.6875,
+      "step": 13189
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5902472384227149,
+      "learning_rate": 1.3836315551074908e-06,
+      "loss": 0.6758,
+      "step": 13190
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5066903903787097,
+      "learning_rate": 1.3835423372413224e-06,
+      "loss": 0.6458,
+      "step": 13191
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.9072444809831086,
+      "learning_rate": 1.383453115795519e-06,
+      "loss": 0.6751,
+      "step": 13192
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8288474631435048,
+      "learning_rate": 1.383363890770914e-06,
+      "loss": 0.6732,
+      "step": 13193
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9207657035623377,
+      "learning_rate": 1.3832746621683397e-06,
+      "loss": 0.6823,
+      "step": 13194
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8593329290881322,
+      "learning_rate": 1.3831854299886292e-06,
+      "loss": 0.6647,
+      "step": 13195
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8849002397650583,
+      "learning_rate": 1.383096194232615e-06,
+      "loss": 0.679,
+      "step": 13196
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8262313012117473,
+      "learning_rate": 1.3830069549011306e-06,
+      "loss": 0.6576,
+      "step": 13197
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8321438931415024,
+      "learning_rate": 1.382917711995008e-06,
+      "loss": 0.6582,
+      "step": 13198
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.2861276828948836,
+      "learning_rate": 1.3828284655150807e-06,
+      "loss": 0.6634,
+      "step": 13199
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.0593119209990085,
+      "learning_rate": 1.3827392154621816e-06,
+      "loss": 0.627,
+      "step": 13200
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6799424806487475,
+      "learning_rate": 1.3826499618371432e-06,
+      "loss": 0.6732,
+      "step": 13201
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.0958256786057488,
+      "learning_rate": 1.3825607046407991e-06,
+      "loss": 0.7057,
+      "step": 13202
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.4683261633534404,
+      "learning_rate": 1.3824714438739819e-06,
+      "loss": 0.7005,
+      "step": 13203
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6675271842554675,
+      "learning_rate": 1.3823821795375253e-06,
+      "loss": 0.6361,
+      "step": 13204
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.104977478502781,
+      "learning_rate": 1.3822929116322617e-06,
+      "loss": 0.7103,
+      "step": 13205
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8580870766708084,
+      "learning_rate": 1.3822036401590245e-06,
+      "loss": 0.6836,
+      "step": 13206
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.3191930401877574,
+      "learning_rate": 1.3821143651186473e-06,
+      "loss": 0.6517,
+      "step": 13207
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 3.7072252179864393,
+      "learning_rate": 1.3820250865119624e-06,
+      "loss": 0.668,
+      "step": 13208
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.801533170507398,
+      "learning_rate": 1.3819358043398041e-06,
+      "loss": 0.6686,
+      "step": 13209
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.600072449429403,
+      "learning_rate": 1.3818465186030051e-06,
+      "loss": 0.6836,
+      "step": 13210
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.9973352192429075,
+      "learning_rate": 1.3817572293023984e-06,
+      "loss": 0.6732,
+      "step": 13211
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.102436434116146,
+      "learning_rate": 1.381667936438818e-06,
+      "loss": 0.6634,
+      "step": 13212
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.138691923276484,
+      "learning_rate": 1.3815786400130969e-06,
+      "loss": 0.679,
+      "step": 13213
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.355499941238161,
+      "learning_rate": 1.3814893400260686e-06,
+      "loss": 0.6615,
+      "step": 13214
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7803387635489443,
+      "learning_rate": 1.3814000364785666e-06,
+      "loss": 0.6862,
+      "step": 13215
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8595632769294927,
+      "learning_rate": 1.381310729371424e-06,
+      "loss": 0.7012,
+      "step": 13216
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9312728380072386,
+      "learning_rate": 1.3812214187054753e-06,
+      "loss": 0.6647,
+      "step": 13217
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.3746614296055113,
+      "learning_rate": 1.3811321044815528e-06,
+      "loss": 0.6387,
+      "step": 13218
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.069975599103567,
+      "learning_rate": 1.381042786700491e-06,
+      "loss": 0.6908,
+      "step": 13219
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.626390451918204,
+      "learning_rate": 1.3809534653631233e-06,
+      "loss": 0.6868,
+      "step": 13220
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.2954322121679716,
+      "learning_rate": 1.3808641404702828e-06,
+      "loss": 0.6908,
+      "step": 13221
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.01057032131821,
+      "learning_rate": 1.3807748120228037e-06,
+      "loss": 0.6725,
+      "step": 13222
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.1227540416570734,
+      "learning_rate": 1.38068548002152e-06,
+      "loss": 0.653,
+      "step": 13223
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.390908874139006,
+      "learning_rate": 1.3805961444672644e-06,
+      "loss": 0.6764,
+      "step": 13224
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.9616962287143727,
+      "learning_rate": 1.380506805360872e-06,
+      "loss": 0.651,
+      "step": 13225
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8888456168572931,
+      "learning_rate": 1.3804174627031752e-06,
+      "loss": 0.6719,
+      "step": 13226
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.909777968783425,
+      "learning_rate": 1.380328116495009e-06,
+      "loss": 0.6576,
+      "step": 13227
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.134587105690488,
+      "learning_rate": 1.380238766737207e-06,
+      "loss": 0.6595,
+      "step": 13228
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.9374478901656317,
+      "learning_rate": 1.3801494134306027e-06,
+      "loss": 0.6445,
+      "step": 13229
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.509260690381801,
+      "learning_rate": 1.3800600565760306e-06,
+      "loss": 0.6615,
+      "step": 13230
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 5.582166487221471,
+      "learning_rate": 1.379970696174324e-06,
+      "loss": 0.6764,
+      "step": 13231
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4523925444344588,
+      "learning_rate": 1.3798813322263174e-06,
+      "loss": 0.6628,
+      "step": 13232
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.3488575819484434,
+      "learning_rate": 1.379791964732845e-06,
+      "loss": 0.6934,
+      "step": 13233
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.7373157307341387,
+      "learning_rate": 1.3797025936947403e-06,
+      "loss": 0.625,
+      "step": 13234
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.589474380546973,
+      "learning_rate": 1.3796132191128378e-06,
+      "loss": 0.6777,
+      "step": 13235
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6291456711514318,
+      "learning_rate": 1.3795238409879718e-06,
+      "loss": 0.6693,
+      "step": 13236
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8459192483856257,
+      "learning_rate": 1.379434459320976e-06,
+      "loss": 0.668,
+      "step": 13237
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.6896872640313128,
+      "learning_rate": 1.379345074112685e-06,
+      "loss": 0.651,
+      "step": 13238
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.1165031960664347,
+      "learning_rate": 1.3792556853639323e-06,
+      "loss": 0.6634,
+      "step": 13239
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 6.7326071732991615,
+      "learning_rate": 1.3791662930755535e-06,
+      "loss": 0.6764,
+      "step": 13240
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8039944212146068,
+      "learning_rate": 1.3790768972483816e-06,
+      "loss": 0.6569,
+      "step": 13241
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 4.998151718522531,
+      "learning_rate": 1.3789874978832518e-06,
+      "loss": 0.6484,
+      "step": 13242
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.2699263832191847,
+      "learning_rate": 1.378898094980998e-06,
+      "loss": 0.6452,
+      "step": 13243
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.1257531021236558,
+      "learning_rate": 1.378808688542455e-06,
+      "loss": 0.6543,
+      "step": 13244
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.8488761017809914,
+      "learning_rate": 1.3787192785684568e-06,
+      "loss": 0.6641,
+      "step": 13245
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.7355229766545706,
+      "learning_rate": 1.378629865059838e-06,
+      "loss": 0.653,
+      "step": 13246
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.021425019059304,
+      "learning_rate": 1.3785404480174334e-06,
+      "loss": 0.6927,
+      "step": 13247
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.070262046879961,
+      "learning_rate": 1.3784510274420772e-06,
+      "loss": 0.7148,
+      "step": 13248
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.1041741496967825,
+      "learning_rate": 1.3783616033346044e-06,
+      "loss": 0.6842,
+      "step": 13249
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6015826401050397,
+      "learning_rate": 1.3782721756958488e-06,
+      "loss": 0.7122,
+      "step": 13250
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.6188704274664474,
+      "learning_rate": 1.3781827445266458e-06,
+      "loss": 0.6803,
+      "step": 13251
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.659733327596371,
+      "learning_rate": 1.3780933098278298e-06,
+      "loss": 0.696,
+      "step": 13252
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.399589102844746,
+      "learning_rate": 1.3780038716002356e-06,
+      "loss": 0.6712,
+      "step": 13253
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.618261303007728,
+      "learning_rate": 1.3779144298446976e-06,
+      "loss": 0.694,
+      "step": 13254
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.102374742242893,
+      "learning_rate": 1.3778249845620508e-06,
+      "loss": 0.6797,
+      "step": 13255
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9937846902012626,
+      "learning_rate": 1.3777355357531305e-06,
+      "loss": 0.6536,
+      "step": 13256
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.035406199203448,
+      "learning_rate": 1.3776460834187705e-06,
+      "loss": 0.694,
+      "step": 13257
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.1731304764743404,
+      "learning_rate": 1.3775566275598062e-06,
+      "loss": 0.6836,
+      "step": 13258
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.058191174411447,
+      "learning_rate": 1.377467168177073e-06,
+      "loss": 0.6582,
+      "step": 13259
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9263133254462386,
+      "learning_rate": 1.3773777052714049e-06,
+      "loss": 0.6947,
+      "step": 13260
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3832537570799932,
+      "learning_rate": 1.3772882388436372e-06,
+      "loss": 0.6784,
+      "step": 13261
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.9243734346148775,
+      "learning_rate": 1.3771987688946054e-06,
+      "loss": 0.6667,
+      "step": 13262
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5609477456954146,
+      "learning_rate": 1.377109295425144e-06,
+      "loss": 0.6432,
+      "step": 13263
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.2768954469290736,
+      "learning_rate": 1.377019818436088e-06,
+      "loss": 0.6823,
+      "step": 13264
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.357458466186097,
+      "learning_rate": 1.3769303379282729e-06,
+      "loss": 0.6615,
+      "step": 13265
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9566169395869201,
+      "learning_rate": 1.3768408539025335e-06,
+      "loss": 0.6576,
+      "step": 13266
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0763298479764194,
+      "learning_rate": 1.3767513663597053e-06,
+      "loss": 0.6777,
+      "step": 13267
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.2481481873562215,
+      "learning_rate": 1.376661875300623e-06,
+      "loss": 0.6927,
+      "step": 13268
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.2038459544430773,
+      "learning_rate": 1.376572380726122e-06,
+      "loss": 0.6966,
+      "step": 13269
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.2275084763437998,
+      "learning_rate": 1.3764828826370383e-06,
+      "loss": 0.668,
+      "step": 13270
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.7620992518429532,
+      "learning_rate": 1.376393381034206e-06,
+      "loss": 0.6699,
+      "step": 13271
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.757648119250422,
+      "learning_rate": 1.3763038759184612e-06,
+      "loss": 0.666,
+      "step": 13272
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.4978808083360518,
+      "learning_rate": 1.376214367290639e-06,
+      "loss": 0.6654,
+      "step": 13273
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.710096718105727,
+      "learning_rate": 1.3761248551515747e-06,
+      "loss": 0.6758,
+      "step": 13274
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.874266718725227,
+      "learning_rate": 1.376035339502104e-06,
+      "loss": 0.6895,
+      "step": 13275
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 8.802380678391776,
+      "learning_rate": 1.3759458203430625e-06,
+      "loss": 0.681,
+      "step": 13276
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0881647222985695,
+      "learning_rate": 1.375856297675285e-06,
+      "loss": 0.6875,
+      "step": 13277
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.699570107710257,
+      "learning_rate": 1.3757667714996074e-06,
+      "loss": 0.6582,
+      "step": 13278
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.089729899036906,
+      "learning_rate": 1.3756772418168656e-06,
+      "loss": 0.653,
+      "step": 13279
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 5.033867755203355,
+      "learning_rate": 1.3755877086278946e-06,
+      "loss": 0.6686,
+      "step": 13280
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.631992229564578,
+      "learning_rate": 1.3754981719335305e-06,
+      "loss": 0.6406,
+      "step": 13281
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3273298580707524,
+      "learning_rate": 1.3754086317346087e-06,
+      "loss": 0.6725,
+      "step": 13282
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.817658332164304,
+      "learning_rate": 1.375319088031965e-06,
+      "loss": 0.6888,
+      "step": 13283
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.7770073396707415,
+      "learning_rate": 1.375229540826435e-06,
+      "loss": 0.6641,
+      "step": 13284
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.0716009919769895,
+      "learning_rate": 1.3751399901188544e-06,
+      "loss": 0.6484,
+      "step": 13285
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.43780828162356,
+      "learning_rate": 1.3750504359100595e-06,
+      "loss": 0.6497,
+      "step": 13286
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.8650094260022902,
+      "learning_rate": 1.3749608782008852e-06,
+      "loss": 0.6888,
+      "step": 13287
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9644277255485094,
+      "learning_rate": 1.3748713169921681e-06,
+      "loss": 0.6647,
+      "step": 13288
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.842778221065664,
+      "learning_rate": 1.3747817522847437e-06,
+      "loss": 0.6908,
+      "step": 13289
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.789481057226158,
+      "learning_rate": 1.3746921840794483e-06,
+      "loss": 0.6751,
+      "step": 13290
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.5488648129120421,
+      "learning_rate": 1.3746026123771176e-06,
+      "loss": 0.6868,
+      "step": 13291
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.785253129590963,
+      "learning_rate": 1.3745130371785873e-06,
+      "loss": 0.6751,
+      "step": 13292
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.5832963219647125,
+      "learning_rate": 1.374423458484694e-06,
+      "loss": 0.6966,
+      "step": 13293
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.354381722712488,
+      "learning_rate": 1.3743338762962733e-06,
+      "loss": 0.6354,
+      "step": 13294
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.1021972470310977,
+      "learning_rate": 1.3742442906141614e-06,
+      "loss": 0.6966,
+      "step": 13295
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.0614199037667507,
+      "learning_rate": 1.3741547014391946e-06,
+      "loss": 0.679,
+      "step": 13296
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 3.056095575632454,
+      "learning_rate": 1.3740651087722088e-06,
+      "loss": 0.6536,
+      "step": 13297
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.3070367543700327,
+      "learning_rate": 1.3739755126140402e-06,
+      "loss": 0.6576,
+      "step": 13298
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0152699929450704,
+      "learning_rate": 1.3738859129655255e-06,
+      "loss": 0.6445,
+      "step": 13299
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.1163610901901246,
+      "learning_rate": 1.3737963098274999e-06,
+      "loss": 0.6654,
+      "step": 13300
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0499746661614564,
+      "learning_rate": 1.3737067032008003e-06,
+      "loss": 0.6654,
+      "step": 13301
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.9353606879792813,
+      "learning_rate": 1.3736170930862635e-06,
+      "loss": 0.6875,
+      "step": 13302
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.003301557320665,
+      "learning_rate": 1.3735274794847249e-06,
+      "loss": 0.6816,
+      "step": 13303
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.23036483228985,
+      "learning_rate": 1.3734378623970215e-06,
+      "loss": 0.6523,
+      "step": 13304
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3899511646650842,
+      "learning_rate": 1.3733482418239892e-06,
+      "loss": 0.6712,
+      "step": 13305
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.064867458512888,
+      "learning_rate": 1.373258617766465e-06,
+      "loss": 0.6543,
+      "step": 13306
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.858328235066125,
+      "learning_rate": 1.3731689902252853e-06,
+      "loss": 0.6667,
+      "step": 13307
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.8996613315131026,
+      "learning_rate": 1.3730793592012862e-06,
+      "loss": 0.6576,
+      "step": 13308
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 4.678003663147057,
+      "learning_rate": 1.3729897246953044e-06,
+      "loss": 0.696,
+      "step": 13309
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.305548481810089,
+      "learning_rate": 1.3729000867081766e-06,
+      "loss": 0.6465,
+      "step": 13310
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.5522399547204957,
+      "learning_rate": 1.3728104452407391e-06,
+      "loss": 0.7116,
+      "step": 13311
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.489158571895533,
+      "learning_rate": 1.372720800293829e-06,
+      "loss": 0.6816,
+      "step": 13312
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0093891550719245,
+      "learning_rate": 1.3726311518682827e-06,
+      "loss": 0.6771,
+      "step": 13313
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.886891855366701,
+      "learning_rate": 1.3725414999649368e-06,
+      "loss": 0.6543,
+      "step": 13314
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.0813051924326076,
+      "learning_rate": 1.3724518445846283e-06,
+      "loss": 0.6471,
+      "step": 13315
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.582863008794248,
+      "learning_rate": 1.3723621857281935e-06,
+      "loss": 0.6406,
+      "step": 13316
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0340821514673189,
+      "learning_rate": 1.3722725233964696e-06,
+      "loss": 0.6738,
+      "step": 13317
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.591077042940746,
+      "learning_rate": 1.3721828575902934e-06,
+      "loss": 0.6706,
+      "step": 13318
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.059074701342683,
+      "learning_rate": 1.3720931883105018e-06,
+      "loss": 0.681,
+      "step": 13319
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6838743365793072,
+      "learning_rate": 1.3720035155579314e-06,
+      "loss": 0.6641,
+      "step": 13320
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6731511728244466,
+      "learning_rate": 1.3719138393334193e-06,
+      "loss": 0.6615,
+      "step": 13321
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0563700090164885,
+      "learning_rate": 1.3718241596378023e-06,
+      "loss": 0.6784,
+      "step": 13322
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1991153790518887,
+      "learning_rate": 1.371734476471918e-06,
+      "loss": 0.6621,
+      "step": 13323
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.718640907902632,
+      "learning_rate": 1.3716447898366025e-06,
+      "loss": 0.6797,
+      "step": 13324
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.835088149494721,
+      "learning_rate": 1.3715550997326934e-06,
+      "loss": 0.7051,
+      "step": 13325
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.2761772954355273,
+      "learning_rate": 1.371465406161028e-06,
+      "loss": 0.694,
+      "step": 13326
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.07643460516648,
+      "learning_rate": 1.3713757091224424e-06,
+      "loss": 0.6699,
+      "step": 13327
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1962835950519988,
+      "learning_rate": 1.3712860086177754e-06,
+      "loss": 0.6497,
+      "step": 13328
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.2613731809682571,
+      "learning_rate": 1.3711963046478627e-06,
+      "loss": 0.6829,
+      "step": 13329
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.6921291457833245,
+      "learning_rate": 1.371106597213542e-06,
+      "loss": 0.6738,
+      "step": 13330
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.720586016430287,
+      "learning_rate": 1.3710168863156508e-06,
+      "loss": 0.6953,
+      "step": 13331
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0675449927796752,
+      "learning_rate": 1.3709271719550261e-06,
+      "loss": 0.6745,
+      "step": 13332
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0302533126270066,
+      "learning_rate": 1.3708374541325052e-06,
+      "loss": 0.6966,
+      "step": 13333
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6473826418684943,
+      "learning_rate": 1.3707477328489258e-06,
+      "loss": 0.6706,
+      "step": 13334
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.1563281180216032,
+      "learning_rate": 1.3706580081051248e-06,
+      "loss": 0.6829,
+      "step": 13335
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.0477044387291232,
+      "learning_rate": 1.3705682799019402e-06,
+      "loss": 0.6842,
+      "step": 13336
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.884627313059669,
+      "learning_rate": 1.3704785482402085e-06,
+      "loss": 0.653,
+      "step": 13337
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.2096985696856613,
+      "learning_rate": 1.370388813120768e-06,
+      "loss": 0.6725,
+      "step": 13338
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.097450134258271,
+      "learning_rate": 1.3702990745444565e-06,
+      "loss": 0.6855,
+      "step": 13339
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.605796077392864,
+      "learning_rate": 1.3702093325121102e-06,
+      "loss": 0.6849,
+      "step": 13340
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.897061892899434,
+      "learning_rate": 1.3701195870245678e-06,
+      "loss": 0.6621,
+      "step": 13341
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.181825934164225,
+      "learning_rate": 1.3700298380826666e-06,
+      "loss": 0.6875,
+      "step": 13342
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.905343796200434,
+      "learning_rate": 1.369940085687244e-06,
+      "loss": 0.6836,
+      "step": 13343
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9240881910645153,
+      "learning_rate": 1.369850329839138e-06,
+      "loss": 0.6823,
+      "step": 13344
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.514690309054045,
+      "learning_rate": 1.3697605705391861e-06,
+      "loss": 0.6764,
+      "step": 13345
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.6297466953694904,
+      "learning_rate": 1.3696708077882263e-06,
+      "loss": 0.6803,
+      "step": 13346
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8218421503855059,
+      "learning_rate": 1.369581041587096e-06,
+      "loss": 0.6901,
+      "step": 13347
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8825880652501386,
+      "learning_rate": 1.369491271936633e-06,
+      "loss": 0.6901,
+      "step": 13348
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.4939776151787463,
+      "learning_rate": 1.3694014988376759e-06,
+      "loss": 0.6602,
+      "step": 13349
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9144035641625052,
+      "learning_rate": 1.3693117222910612e-06,
+      "loss": 0.6549,
+      "step": 13350
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.570049009202544,
+      "learning_rate": 1.3692219422976276e-06,
+      "loss": 0.6615,
+      "step": 13351
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.2684007597761346,
+      "learning_rate": 1.3691321588582135e-06,
+      "loss": 0.6686,
+      "step": 13352
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 4.0266285680708025,
+      "learning_rate": 1.369042371973656e-06,
+      "loss": 0.6999,
+      "step": 13353
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.866712352377824,
+      "learning_rate": 1.3689525816447932e-06,
+      "loss": 0.6777,
+      "step": 13354
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.8640627217734163,
+      "learning_rate": 1.3688627878724634e-06,
+      "loss": 0.6784,
+      "step": 13355
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8754017238451621,
+      "learning_rate": 1.3687729906575049e-06,
+      "loss": 0.6693,
+      "step": 13356
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.0858277914218544,
+      "learning_rate": 1.3686831900007555e-06,
+      "loss": 0.666,
+      "step": 13357
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.640494937004661,
+      "learning_rate": 1.3685933859030527e-06,
+      "loss": 0.6523,
+      "step": 13358
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.6811650214130116,
+      "learning_rate": 1.3685035783652357e-06,
+      "loss": 0.6647,
+      "step": 13359
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.587422106713824,
+      "learning_rate": 1.3684137673881422e-06,
+      "loss": 0.6797,
+      "step": 13360
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.39731253451383,
+      "learning_rate": 1.3683239529726102e-06,
+      "loss": 0.666,
+      "step": 13361
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 3.1901316770757666,
+      "learning_rate": 1.3682341351194784e-06,
+      "loss": 0.7005,
+      "step": 13362
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.7875737451126295,
+      "learning_rate": 1.3681443138295847e-06,
+      "loss": 0.6667,
+      "step": 13363
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.9375115697339558,
+      "learning_rate": 1.3680544891037675e-06,
+      "loss": 0.6615,
+      "step": 13364
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.396071146978566,
+      "learning_rate": 1.3679646609428654e-06,
+      "loss": 0.6777,
+      "step": 13365
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7652370658712304,
+      "learning_rate": 1.3678748293477167e-06,
+      "loss": 0.6693,
+      "step": 13366
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.58327577805397,
+      "learning_rate": 1.3677849943191593e-06,
+      "loss": 0.681,
+      "step": 13367
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.600959676445185,
+      "learning_rate": 1.3676951558580323e-06,
+      "loss": 0.6621,
+      "step": 13368
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.186983465733605,
+      "learning_rate": 1.3676053139651737e-06,
+      "loss": 0.6504,
+      "step": 13369
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.7735822015108633,
+      "learning_rate": 1.3675154686414225e-06,
+      "loss": 0.7012,
+      "step": 13370
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9061463660838851,
+      "learning_rate": 1.3674256198876165e-06,
+      "loss": 0.6862,
+      "step": 13371
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.8657058905155948,
+      "learning_rate": 1.367335767704595e-06,
+      "loss": 0.6699,
+      "step": 13372
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.9725580143207441,
+      "learning_rate": 1.3672459120931962e-06,
+      "loss": 0.6719,
+      "step": 13373
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 6.551824194807773,
+      "learning_rate": 1.3671560530542588e-06,
+      "loss": 0.6979,
+      "step": 13374
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 5.282033879657559,
+      "learning_rate": 1.3670661905886216e-06,
+      "loss": 0.6842,
+      "step": 13375
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.2781966863413294,
+      "learning_rate": 1.366976324697123e-06,
+      "loss": 0.6478,
+      "step": 13376
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.414941476746597,
+      "learning_rate": 1.366886455380602e-06,
+      "loss": 0.6836,
+      "step": 13377
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.199621770820516,
+      "learning_rate": 1.3667965826398975e-06,
+      "loss": 0.6862,
+      "step": 13378
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.8464577866235123,
+      "learning_rate": 1.3667067064758478e-06,
+      "loss": 0.6543,
+      "step": 13379
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.8211721480751106,
+      "learning_rate": 1.3666168268892919e-06,
+      "loss": 0.6602,
+      "step": 13380
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.243672650725814,
+      "learning_rate": 1.366526943881069e-06,
+      "loss": 0.6842,
+      "step": 13381
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.155969946173069,
+      "learning_rate": 1.3664370574520177e-06,
+      "loss": 0.6452,
+      "step": 13382
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.607868425745373,
+      "learning_rate": 1.3663471676029765e-06,
+      "loss": 0.7116,
+      "step": 13383
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.2485148801931,
+      "learning_rate": 1.366257274334785e-06,
+      "loss": 0.6732,
+      "step": 13384
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.2165569383817414,
+      "learning_rate": 1.3661673776482822e-06,
+      "loss": 0.6986,
+      "step": 13385
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.650310739425231,
+      "learning_rate": 1.3660774775443068e-06,
+      "loss": 0.6849,
+      "step": 13386
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.414167357097979,
+      "learning_rate": 1.3659875740236978e-06,
+      "loss": 0.6732,
+      "step": 13387
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9012788372904439,
+      "learning_rate": 1.3658976670872944e-06,
+      "loss": 0.6543,
+      "step": 13388
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1033193309081588,
+      "learning_rate": 1.365807756735936e-06,
+      "loss": 0.6875,
+      "step": 13389
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.694457143387877,
+      "learning_rate": 1.3657178429704611e-06,
+      "loss": 0.6784,
+      "step": 13390
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.773032758725273,
+      "learning_rate": 1.3656279257917097e-06,
+      "loss": 0.6771,
+      "step": 13391
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.5411449345100476,
+      "learning_rate": 1.36553800520052e-06,
+      "loss": 0.7122,
+      "step": 13392
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.5527849039399553,
+      "learning_rate": 1.3654480811977321e-06,
+      "loss": 0.6628,
+      "step": 13393
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.808326163653883,
+      "learning_rate": 1.3653581537841848e-06,
+      "loss": 0.6719,
+      "step": 13394
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8729978611972425,
+      "learning_rate": 1.3652682229607177e-06,
+      "loss": 0.6647,
+      "step": 13395
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.6744640250792675,
+      "learning_rate": 1.3651782887281695e-06,
+      "loss": 0.6934,
+      "step": 13396
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.322234669837583,
+      "learning_rate": 1.3650883510873806e-06,
+      "loss": 0.6823,
+      "step": 13397
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.1033614974321817,
+      "learning_rate": 1.3649984100391893e-06,
+      "loss": 0.668,
+      "step": 13398
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.5002626100104335,
+      "learning_rate": 1.3649084655844362e-06,
+      "loss": 0.6738,
+      "step": 13399
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.5369907874539597,
+      "learning_rate": 1.3648185177239597e-06,
+      "loss": 0.7018,
+      "step": 13400
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.992853586195895,
+      "learning_rate": 1.3647285664585996e-06,
+      "loss": 0.6862,
+      "step": 13401
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 4.414146407717461,
+      "learning_rate": 1.364638611789196e-06,
+      "loss": 0.6771,
+      "step": 13402
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.442659437170646,
+      "learning_rate": 1.3645486537165876e-06,
+      "loss": 0.6465,
+      "step": 13403
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.336888397494825,
+      "learning_rate": 1.3644586922416143e-06,
+      "loss": 0.6973,
+      "step": 13404
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4464509134913854,
+      "learning_rate": 1.3643687273651163e-06,
+      "loss": 0.6673,
+      "step": 13405
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9202480502794188,
+      "learning_rate": 1.3642787590879323e-06,
+      "loss": 0.6751,
+      "step": 13406
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.8892149041164337,
+      "learning_rate": 1.3641887874109025e-06,
+      "loss": 0.6882,
+      "step": 13407
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.45240322177311,
+      "learning_rate": 1.364098812334867e-06,
+      "loss": 0.6888,
+      "step": 13408
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.4194919127395194,
+      "learning_rate": 1.3640088338606646e-06,
+      "loss": 0.6348,
+      "step": 13409
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.739544894447946,
+      "learning_rate": 1.3639188519891355e-06,
+      "loss": 0.666,
+      "step": 13410
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.0815525556602728,
+      "learning_rate": 1.3638288667211195e-06,
+      "loss": 0.6556,
+      "step": 13411
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7520992652987862,
+      "learning_rate": 1.363738878057457e-06,
+      "loss": 0.6602,
+      "step": 13412
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.5533341776805285,
+      "learning_rate": 1.363648885998987e-06,
+      "loss": 0.6452,
+      "step": 13413
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 5.04805144618195,
+      "learning_rate": 1.3635588905465498e-06,
+      "loss": 0.6823,
+      "step": 13414
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 6.195126617433276,
+      "learning_rate": 1.3634688917009857e-06,
+      "loss": 0.6745,
+      "step": 13415
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.3045118637573614,
+      "learning_rate": 1.363378889463134e-06,
+      "loss": 0.6758,
+      "step": 13416
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.5894611901585247,
+      "learning_rate": 1.3632888838338348e-06,
+      "loss": 0.6595,
+      "step": 13417
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9940049821484593,
+      "learning_rate": 1.3631988748139289e-06,
+      "loss": 0.6556,
+      "step": 13418
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.853336487761782,
+      "learning_rate": 1.363108862404255e-06,
+      "loss": 0.6582,
+      "step": 13419
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.221589295582053,
+      "learning_rate": 1.3630188466056547e-06,
+      "loss": 0.6296,
+      "step": 13420
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1312783527648897,
+      "learning_rate": 1.3629288274189668e-06,
+      "loss": 0.6667,
+      "step": 13421
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.097484221321375,
+      "learning_rate": 1.3628388048450325e-06,
+      "loss": 0.6569,
+      "step": 13422
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1270642906670085,
+      "learning_rate": 1.3627487788846914e-06,
+      "loss": 0.6393,
+      "step": 13423
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.6992002997625466,
+      "learning_rate": 1.3626587495387838e-06,
+      "loss": 0.6764,
+      "step": 13424
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.8017140930994748,
+      "learning_rate": 1.36256871680815e-06,
+      "loss": 0.6478,
+      "step": 13425
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1541448997313373,
+      "learning_rate": 1.3624786806936307e-06,
+      "loss": 0.6497,
+      "step": 13426
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3722441122906721,
+      "learning_rate": 1.3623886411960653e-06,
+      "loss": 0.6725,
+      "step": 13427
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.3283604592918463,
+      "learning_rate": 1.3622985983162947e-06,
+      "loss": 0.6868,
+      "step": 13428
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.8582732152476003,
+      "learning_rate": 1.3622085520551596e-06,
+      "loss": 0.6543,
+      "step": 13429
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.001036374145405,
+      "learning_rate": 1.3621185024134997e-06,
+      "loss": 0.6732,
+      "step": 13430
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 3.351069610145807,
+      "learning_rate": 1.362028449392156e-06,
+      "loss": 0.6354,
+      "step": 13431
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4913614646109223,
+      "learning_rate": 1.3619383929919687e-06,
+      "loss": 0.6484,
+      "step": 13432
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6375565313602714,
+      "learning_rate": 1.3618483332137783e-06,
+      "loss": 0.6569,
+      "step": 13433
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.5946337671342525,
+      "learning_rate": 1.3617582700584256e-06,
+      "loss": 0.6387,
+      "step": 13434
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.9616074742914806,
+      "learning_rate": 1.361668203526751e-06,
+      "loss": 0.6536,
+      "step": 13435
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3348811244568353,
+      "learning_rate": 1.3615781336195949e-06,
+      "loss": 0.6797,
+      "step": 13436
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.156104207635896,
+      "learning_rate": 1.361488060337798e-06,
+      "loss": 0.6198,
+      "step": 13437
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.873429853183892,
+      "learning_rate": 1.3613979836822012e-06,
+      "loss": 0.6204,
+      "step": 13438
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.263062646500634,
+      "learning_rate": 1.3613079036536451e-06,
+      "loss": 0.6751,
+      "step": 13439
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.7759578756169916,
+      "learning_rate": 1.3612178202529702e-06,
+      "loss": 0.6654,
+      "step": 13440
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.7180515752559813,
+      "learning_rate": 1.3611277334810176e-06,
+      "loss": 0.6439,
+      "step": 13441
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6700757479218697,
+      "learning_rate": 1.3610376433386281e-06,
+      "loss": 0.6341,
+      "step": 13442
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4536343049884402,
+      "learning_rate": 1.3609475498266419e-06,
+      "loss": 0.6445,
+      "step": 13443
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.6178177837838845,
+      "learning_rate": 1.3608574529459007e-06,
+      "loss": 0.7005,
+      "step": 13444
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.9887823695770415,
+      "learning_rate": 1.360767352697245e-06,
+      "loss": 0.6641,
+      "step": 13445
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.1973958273094327,
+      "learning_rate": 1.3606772490815152e-06,
+      "loss": 0.6849,
+      "step": 13446
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.5943815588031862,
+      "learning_rate": 1.3605871420995532e-06,
+      "loss": 0.64,
+      "step": 13447
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4298661551836425,
+      "learning_rate": 1.3604970317521996e-06,
+      "loss": 0.6309,
+      "step": 13448
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4618494905422166,
+      "learning_rate": 1.3604069180402949e-06,
+      "loss": 0.6497,
+      "step": 13449
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.603148842087315,
+      "learning_rate": 1.3603168009646808e-06,
+      "loss": 0.6764,
+      "step": 13450
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.208394333409434,
+      "learning_rate": 1.360226680526198e-06,
+      "loss": 0.7012,
+      "step": 13451
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.525988780284267,
+      "learning_rate": 1.360136556725688e-06,
+      "loss": 0.6725,
+      "step": 13452
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.482289715468606,
+      "learning_rate": 1.3600464295639912e-06,
+      "loss": 0.6374,
+      "step": 13453
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.48783799937286,
+      "learning_rate": 1.3599562990419496e-06,
+      "loss": 0.6283,
+      "step": 13454
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.468512963462477,
+      "learning_rate": 1.3598661651604042e-06,
+      "loss": 0.6667,
+      "step": 13455
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.4949948757751295,
+      "learning_rate": 1.3597760279201955e-06,
+      "loss": 0.6354,
+      "step": 13456
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.053857077033095,
+      "learning_rate": 1.3596858873221656e-06,
+      "loss": 0.6471,
+      "step": 13457
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.457955699122341,
+      "learning_rate": 1.3595957433671557e-06,
+      "loss": 0.6777,
+      "step": 13458
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.542726860227372,
+      "learning_rate": 1.3595055960560067e-06,
+      "loss": 0.64,
+      "step": 13459
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.3579458898256087,
+      "learning_rate": 1.35941544538956e-06,
+      "loss": 0.6302,
+      "step": 13460
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.813516704623326,
+      "learning_rate": 1.3593252913686578e-06,
+      "loss": 0.6738,
+      "step": 13461
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.284799478668429,
+      "learning_rate": 1.35923513399414e-06,
+      "loss": 0.638,
+      "step": 13462
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.761669209633376,
+      "learning_rate": 1.3591449732668497e-06,
+      "loss": 0.6478,
+      "step": 13463
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.6937769442263333,
+      "learning_rate": 1.359054809187627e-06,
+      "loss": 0.623,
+      "step": 13464
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.466975290226957,
+      "learning_rate": 1.3589646417573144e-06,
+      "loss": 0.6217,
+      "step": 13465
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.645386352396363,
+      "learning_rate": 1.3588744709767532e-06,
+      "loss": 0.6602,
+      "step": 13466
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.900885818514905,
+      "learning_rate": 1.3587842968467844e-06,
+      "loss": 0.6992,
+      "step": 13467
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6918653626571019,
+      "learning_rate": 1.3586941193682505e-06,
+      "loss": 0.6302,
+      "step": 13468
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.00267798958568,
+      "learning_rate": 1.3586039385419921e-06,
+      "loss": 0.6426,
+      "step": 13469
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.701333600554813,
+      "learning_rate": 1.3585137543688516e-06,
+      "loss": 0.6641,
+      "step": 13470
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.716804802580609,
+      "learning_rate": 1.358423566849671e-06,
+      "loss": 0.6732,
+      "step": 13471
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.351858070845399,
+      "learning_rate": 1.358333375985291e-06,
+      "loss": 0.653,
+      "step": 13472
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.4998803702801038,
+      "learning_rate": 1.358243181776554e-06,
+      "loss": 0.666,
+      "step": 13473
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.579321672317965,
+      "learning_rate": 1.3581529842243018e-06,
+      "loss": 0.6172,
+      "step": 13474
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.998884281541847,
+      "learning_rate": 1.3580627833293763e-06,
+      "loss": 0.6172,
+      "step": 13475
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.6212234222910067,
+      "learning_rate": 1.357972579092619e-06,
+      "loss": 0.6445,
+      "step": 13476
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.404092806333264,
+      "learning_rate": 1.3578823715148718e-06,
+      "loss": 0.6478,
+      "step": 13477
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.124140734836512,
+      "learning_rate": 1.357792160596977e-06,
+      "loss": 0.6484,
+      "step": 13478
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.189092964042026,
+      "learning_rate": 1.3577019463397763e-06,
+      "loss": 0.6751,
+      "step": 13479
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.3496016861893105,
+      "learning_rate": 1.3576117287441116e-06,
+      "loss": 0.7018,
+      "step": 13480
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.153376981597459,
+      "learning_rate": 1.3575215078108254e-06,
+      "loss": 0.584,
+      "step": 13481
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.913446407711714,
+      "learning_rate": 1.357431283540759e-06,
+      "loss": 0.6549,
+      "step": 13482
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.7080358514362428,
+      "learning_rate": 1.3573410559347545e-06,
+      "loss": 0.6543,
+      "step": 13483
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.444049578678295,
+      "learning_rate": 1.357250824993655e-06,
+      "loss": 0.6764,
+      "step": 13484
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.8145553316947172,
+      "learning_rate": 1.3571605907183014e-06,
+      "loss": 0.6888,
+      "step": 13485
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.8347032489401274,
+      "learning_rate": 1.3570703531095368e-06,
+      "loss": 0.6296,
+      "step": 13486
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.181386785584049,
+      "learning_rate": 1.3569801121682027e-06,
+      "loss": 0.627,
+      "step": 13487
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.293098376986487,
+      "learning_rate": 1.356889867895142e-06,
+      "loss": 0.6576,
+      "step": 13488
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.90058951054711,
+      "learning_rate": 1.3567996202911963e-06,
+      "loss": 0.6374,
+      "step": 13489
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.9557798169317724,
+      "learning_rate": 1.3567093693572083e-06,
+      "loss": 0.6615,
+      "step": 13490
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 5.381423929586669,
+      "learning_rate": 1.3566191150940203e-06,
+      "loss": 0.6536,
+      "step": 13491
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.2022510330470526,
+      "learning_rate": 1.3565288575024746e-06,
+      "loss": 0.6452,
+      "step": 13492
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.008155853838501,
+      "learning_rate": 1.356438596583413e-06,
+      "loss": 0.6497,
+      "step": 13493
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.535238195586314,
+      "learning_rate": 1.356348332337679e-06,
+      "loss": 0.6621,
+      "step": 13494
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6652024456762562,
+      "learning_rate": 1.3562580647661146e-06,
+      "loss": 0.6348,
+      "step": 13495
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.415134047253328,
+      "learning_rate": 1.3561677938695617e-06,
+      "loss": 0.6927,
+      "step": 13496
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.4848864412027747,
+      "learning_rate": 1.3560775196488636e-06,
+      "loss": 0.651,
+      "step": 13497
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.9746988999192507,
+      "learning_rate": 1.3559872421048626e-06,
+      "loss": 0.6361,
+      "step": 13498
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.0332403039536393,
+      "learning_rate": 1.3558969612384007e-06,
+      "loss": 0.6387,
+      "step": 13499
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.506641171226551,
+      "learning_rate": 1.3558066770503214e-06,
+      "loss": 0.6667,
+      "step": 13500
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.4972403864361676,
+      "learning_rate": 1.355716389541467e-06,
+      "loss": 0.6641,
+      "step": 13501
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6877139430194366,
+      "learning_rate": 1.35562609871268e-06,
+      "loss": 0.6517,
+      "step": 13502
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.567589004524895,
+      "learning_rate": 1.3555358045648029e-06,
+      "loss": 0.6712,
+      "step": 13503
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6016923631571005,
+      "learning_rate": 1.355445507098679e-06,
+      "loss": 0.6641,
+      "step": 13504
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 6.834955302751593,
+      "learning_rate": 1.3553552063151508e-06,
+      "loss": 0.6484,
+      "step": 13505
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 4.8770183394677105,
+      "learning_rate": 1.3552649022150605e-06,
+      "loss": 0.6595,
+      "step": 13506
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.8991308816990733,
+      "learning_rate": 1.355174594799252e-06,
+      "loss": 0.6445,
+      "step": 13507
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.871245494043392,
+      "learning_rate": 1.3550842840685674e-06,
+      "loss": 0.6276,
+      "step": 13508
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.4088083234612143,
+      "learning_rate": 1.3549939700238498e-06,
+      "loss": 0.6504,
+      "step": 13509
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.6116216229587577,
+      "learning_rate": 1.354903652665942e-06,
+      "loss": 0.6628,
+      "step": 13510
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.470062384602289,
+      "learning_rate": 1.3548133319956871e-06,
+      "loss": 0.6465,
+      "step": 13511
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 9.899797391267114,
+      "learning_rate": 1.354723008013928e-06,
+      "loss": 0.6973,
+      "step": 13512
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.485916192562916,
+      "learning_rate": 1.3546326807215076e-06,
+      "loss": 0.6432,
+      "step": 13513
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.348616511416443,
+      "learning_rate": 1.3545423501192694e-06,
+      "loss": 0.6908,
+      "step": 13514
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.6914406595098184,
+      "learning_rate": 1.3544520162080556e-06,
+      "loss": 0.6497,
+      "step": 13515
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.961816630297597,
+      "learning_rate": 1.35436167898871e-06,
+      "loss": 0.6465,
+      "step": 13516
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.8148352163827774,
+      "learning_rate": 1.3542713384620756e-06,
+      "loss": 0.6113,
+      "step": 13517
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.878010554969716,
+      "learning_rate": 1.3541809946289954e-06,
+      "loss": 0.6608,
+      "step": 13518
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.8320779856914853,
+      "learning_rate": 1.3540906474903126e-06,
+      "loss": 0.6432,
+      "step": 13519
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6325591003479345,
+      "learning_rate": 1.3540002970468701e-06,
+      "loss": 0.61,
+      "step": 13520
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.395115822787504,
+      "learning_rate": 1.3539099432995124e-06,
+      "loss": 0.6445,
+      "step": 13521
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.2119266928441554,
+      "learning_rate": 1.3538195862490812e-06,
+      "loss": 0.6764,
+      "step": 13522
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.4314813174083487,
+      "learning_rate": 1.3537292258964208e-06,
+      "loss": 0.6549,
+      "step": 13523
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6673938556401158,
+      "learning_rate": 1.3536388622423744e-06,
+      "loss": 0.6816,
+      "step": 13524
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.2036352670885604,
+      "learning_rate": 1.3535484952877848e-06,
+      "loss": 0.6816,
+      "step": 13525
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.0219324008900528,
+      "learning_rate": 1.3534581250334962e-06,
+      "loss": 0.6452,
+      "step": 13526
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.420789763809331,
+      "learning_rate": 1.3533677514803514e-06,
+      "loss": 0.6341,
+      "step": 13527
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.923023180020789,
+      "learning_rate": 1.3532773746291942e-06,
+      "loss": 0.6569,
+      "step": 13528
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.15607778572406,
+      "learning_rate": 1.353186994480868e-06,
+      "loss": 0.6797,
+      "step": 13529
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.656820424317417,
+      "learning_rate": 1.3530966110362163e-06,
+      "loss": 0.7031,
+      "step": 13530
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.667808680148717,
+      "learning_rate": 1.3530062242960828e-06,
+      "loss": 0.6354,
+      "step": 13531
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.1455163712556833,
+      "learning_rate": 1.3529158342613112e-06,
+      "loss": 0.6328,
+      "step": 13532
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.6623518126746126,
+      "learning_rate": 1.3528254409327444e-06,
+      "loss": 0.6452,
+      "step": 13533
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6919186001500208,
+      "learning_rate": 1.352735044311227e-06,
+      "loss": 0.625,
+      "step": 13534
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.3832320805921294,
+      "learning_rate": 1.352644644397602e-06,
+      "loss": 0.6777,
+      "step": 13535
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.412585328361205,
+      "learning_rate": 1.3525542411927132e-06,
+      "loss": 0.6296,
+      "step": 13536
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.0863770609466235,
+      "learning_rate": 1.352463834697405e-06,
+      "loss": 0.6484,
+      "step": 13537
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.5949618099548464,
+      "learning_rate": 1.3523734249125206e-06,
+      "loss": 0.6484,
+      "step": 13538
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.696104528354575,
+      "learning_rate": 1.3522830118389034e-06,
+      "loss": 0.6536,
+      "step": 13539
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.1606102793197763,
+      "learning_rate": 1.352192595477398e-06,
+      "loss": 0.6354,
+      "step": 13540
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.31761412399972,
+      "learning_rate": 1.352102175828848e-06,
+      "loss": 0.6992,
+      "step": 13541
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6913871974204782,
+      "learning_rate": 1.3520117528940973e-06,
+      "loss": 0.6882,
+      "step": 13542
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.9172939135700018,
+      "learning_rate": 1.3519213266739894e-06,
+      "loss": 0.6458,
+      "step": 13543
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.596482882453033,
+      "learning_rate": 1.351830897169369e-06,
+      "loss": 0.6413,
+      "step": 13544
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.536351704371488,
+      "learning_rate": 1.3517404643810796e-06,
+      "loss": 0.6576,
+      "step": 13545
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.9676366365553908,
+      "learning_rate": 1.3516500283099653e-06,
+      "loss": 0.6191,
+      "step": 13546
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.8318532965934393,
+      "learning_rate": 1.3515595889568703e-06,
+      "loss": 0.6322,
+      "step": 13547
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.7023185194398178,
+      "learning_rate": 1.3514691463226384e-06,
+      "loss": 0.6283,
+      "step": 13548
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.333206234125659,
+      "learning_rate": 1.351378700408114e-06,
+      "loss": 0.6478,
+      "step": 13549
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.822913247212875,
+      "learning_rate": 1.351288251214141e-06,
+      "loss": 0.653,
+      "step": 13550
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.7661261334332425,
+      "learning_rate": 1.351197798741564e-06,
+      "loss": 0.7096,
+      "step": 13551
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.3648055881822705,
+      "learning_rate": 1.3511073429912266e-06,
+      "loss": 0.6602,
+      "step": 13552
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.77264072688375,
+      "learning_rate": 1.351016883963973e-06,
+      "loss": 0.638,
+      "step": 13553
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.9843479601974854,
+      "learning_rate": 1.3509264216606485e-06,
+      "loss": 0.7051,
+      "step": 13554
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.5101560789948794,
+      "learning_rate": 1.3508359560820964e-06,
+      "loss": 0.6478,
+      "step": 13555
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.6920940122703017,
+      "learning_rate": 1.3507454872291607e-06,
+      "loss": 0.6478,
+      "step": 13556
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.753377499151465,
+      "learning_rate": 1.350655015102687e-06,
+      "loss": 0.6523,
+      "step": 13557
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.5131123582362753,
+      "learning_rate": 1.350564539703519e-06,
+      "loss": 0.6725,
+      "step": 13558
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.718029479530143,
+      "learning_rate": 1.350474061032501e-06,
+      "loss": 0.6198,
+      "step": 13559
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.0172188373229107,
+      "learning_rate": 1.3503835790904776e-06,
+      "loss": 0.6289,
+      "step": 13560
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.9829611466534387,
+      "learning_rate": 1.3502930938782934e-06,
+      "loss": 0.6335,
+      "step": 13561
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.674264100843808,
+      "learning_rate": 1.3502026053967926e-06,
+      "loss": 0.5918,
+      "step": 13562
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.7536962457054264,
+      "learning_rate": 1.3501121136468202e-06,
+      "loss": 0.6322,
+      "step": 13563
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.8212517511147444,
+      "learning_rate": 1.3500216186292202e-06,
+      "loss": 0.666,
+      "step": 13564
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.599450351244365,
+      "learning_rate": 1.3499311203448375e-06,
+      "loss": 0.6517,
+      "step": 13565
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.215882231960931,
+      "learning_rate": 1.3498406187945168e-06,
+      "loss": 0.6706,
+      "step": 13566
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6099484126608719,
+      "learning_rate": 1.3497501139791027e-06,
+      "loss": 0.627,
+      "step": 13567
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.7576085838337492,
+      "learning_rate": 1.3496596058994398e-06,
+      "loss": 0.6139,
+      "step": 13568
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.0097210584009195,
+      "learning_rate": 1.3495690945563726e-06,
+      "loss": 0.6471,
+      "step": 13569
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.9947511070036303,
+      "learning_rate": 1.3494785799507467e-06,
+      "loss": 0.6289,
+      "step": 13570
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 4.8817279784436804,
+      "learning_rate": 1.3493880620834061e-06,
+      "loss": 0.6543,
+      "step": 13571
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.0044256894700783,
+      "learning_rate": 1.3492975409551956e-06,
+      "loss": 0.6276,
+      "step": 13572
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 5.6411080657927775,
+      "learning_rate": 1.3492070165669604e-06,
+      "loss": 0.6517,
+      "step": 13573
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.231286019274484,
+      "learning_rate": 1.3491164889195451e-06,
+      "loss": 0.6491,
+      "step": 13574
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.873614672036474,
+      "learning_rate": 1.3490259580137948e-06,
+      "loss": 0.6699,
+      "step": 13575
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.9128377018143128,
+      "learning_rate": 1.3489354238505547e-06,
+      "loss": 0.6413,
+      "step": 13576
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.3109924083436613,
+      "learning_rate": 1.348844886430669e-06,
+      "loss": 0.6745,
+      "step": 13577
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.9073320173440975,
+      "learning_rate": 1.3487543457549833e-06,
+      "loss": 0.6634,
+      "step": 13578
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.6304559531358516,
+      "learning_rate": 1.3486638018243424e-06,
+      "loss": 0.6706,
+      "step": 13579
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.767241266447287,
+      "learning_rate": 1.3485732546395915e-06,
+      "loss": 0.6589,
+      "step": 13580
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.625602580507604,
+      "learning_rate": 1.3484827042015754e-06,
+      "loss": 0.6322,
+      "step": 13581
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.5915828723948473,
+      "learning_rate": 1.3483921505111395e-06,
+      "loss": 0.6178,
+      "step": 13582
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.6792960070698864,
+      "learning_rate": 1.3483015935691287e-06,
+      "loss": 0.6491,
+      "step": 13583
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.995016376397095,
+      "learning_rate": 1.3482110333763888e-06,
+      "loss": 0.6562,
+      "step": 13584
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0402994212966927,
+      "learning_rate": 1.3481204699337641e-06,
+      "loss": 0.6328,
+      "step": 13585
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.769781314878707,
+      "learning_rate": 1.3480299032421e-06,
+      "loss": 0.6393,
+      "step": 13586
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.1327600908769595,
+      "learning_rate": 1.3479393333022429e-06,
+      "loss": 0.6361,
+      "step": 13587
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.18135785523164,
+      "learning_rate": 1.3478487601150365e-06,
+      "loss": 0.6699,
+      "step": 13588
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.9593203740492222,
+      "learning_rate": 1.3477581836813271e-06,
+      "loss": 0.6758,
+      "step": 13589
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6749204663981274,
+      "learning_rate": 1.34766760400196e-06,
+      "loss": 0.6354,
+      "step": 13590
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.7933463190029055,
+      "learning_rate": 1.34757702107778e-06,
+      "loss": 0.6419,
+      "step": 13591
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.316523585567213,
+      "learning_rate": 1.3474864349096333e-06,
+      "loss": 0.6237,
+      "step": 13592
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.137338069847435,
+      "learning_rate": 1.347395845498365e-06,
+      "loss": 0.612,
+      "step": 13593
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.961744294315645,
+      "learning_rate": 1.34730525284482e-06,
+      "loss": 0.5977,
+      "step": 13594
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.252924629266121,
+      "learning_rate": 1.347214656949845e-06,
+      "loss": 0.6901,
+      "step": 13595
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.120440525250097,
+      "learning_rate": 1.3471240578142846e-06,
+      "loss": 0.6445,
+      "step": 13596
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.445534157069433,
+      "learning_rate": 1.3470334554389849e-06,
+      "loss": 0.6686,
+      "step": 13597
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.177427939044865,
+      "learning_rate": 1.346942849824791e-06,
+      "loss": 0.6641,
+      "step": 13598
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.330516817307844,
+      "learning_rate": 1.3468522409725488e-06,
+      "loss": 0.6367,
+      "step": 13599
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.027652814952941,
+      "learning_rate": 1.3467616288831044e-06,
+      "loss": 0.6589,
+      "step": 13600
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.421718713819332,
+      "learning_rate": 1.3466710135573025e-06,
+      "loss": 0.6667,
+      "step": 13601
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.448471129858495,
+      "learning_rate": 1.3465803949959894e-06,
+      "loss": 0.6413,
+      "step": 13602
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.479960480332824,
+      "learning_rate": 1.3464897732000112e-06,
+      "loss": 0.6465,
+      "step": 13603
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0688288303690117,
+      "learning_rate": 1.3463991481702128e-06,
+      "loss": 0.6608,
+      "step": 13604
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.351059720801898,
+      "learning_rate": 1.346308519907441e-06,
+      "loss": 0.6126,
+      "step": 13605
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.507303487185487,
+      "learning_rate": 1.3462178884125408e-06,
+      "loss": 0.6113,
+      "step": 13606
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.9230231764373185,
+      "learning_rate": 1.3461272536863588e-06,
+      "loss": 0.6738,
+      "step": 13607
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 5.455222887898202,
+      "learning_rate": 1.3460366157297403e-06,
+      "loss": 0.6419,
+      "step": 13608
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.9258340712697852,
+      "learning_rate": 1.3459459745435315e-06,
+      "loss": 0.6784,
+      "step": 13609
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.8887076713585744,
+      "learning_rate": 1.3458553301285782e-06,
+      "loss": 0.668,
+      "step": 13610
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.017466925915035,
+      "learning_rate": 1.3457646824857267e-06,
+      "loss": 0.6445,
+      "step": 13611
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.2301850329724036,
+      "learning_rate": 1.3456740316158228e-06,
+      "loss": 0.6068,
+      "step": 13612
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.8020039106776704,
+      "learning_rate": 1.3455833775197128e-06,
+      "loss": 0.6185,
+      "step": 13613
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.1008264190828134,
+      "learning_rate": 1.3454927201982423e-06,
+      "loss": 0.6771,
+      "step": 13614
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.451574464134144,
+      "learning_rate": 1.3454020596522575e-06,
+      "loss": 0.6602,
+      "step": 13615
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.8902364534892222,
+      "learning_rate": 1.345311395882605e-06,
+      "loss": 0.6016,
+      "step": 13616
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0422697774503167,
+      "learning_rate": 1.3452207288901307e-06,
+      "loss": 0.6458,
+      "step": 13617
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.1073769204835697,
+      "learning_rate": 1.345130058675681e-06,
+      "loss": 0.6738,
+      "step": 13618
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.079491570253361,
+      "learning_rate": 1.3450393852401013e-06,
+      "loss": 0.6751,
+      "step": 13619
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0311845629097256,
+      "learning_rate": 1.3449487085842392e-06,
+      "loss": 0.653,
+      "step": 13620
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.586192308123405,
+      "learning_rate": 1.3448580287089401e-06,
+      "loss": 0.6621,
+      "step": 13621
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.881761665360499,
+      "learning_rate": 1.3447673456150504e-06,
+      "loss": 0.6309,
+      "step": 13622
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.2744506419122095,
+      "learning_rate": 1.3446766593034167e-06,
+      "loss": 0.6758,
+      "step": 13623
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.9405985274598723,
+      "learning_rate": 1.344585969774885e-06,
+      "loss": 0.6465,
+      "step": 13624
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.5976155032370283,
+      "learning_rate": 1.344495277030302e-06,
+      "loss": 0.6582,
+      "step": 13625
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0308963675281597,
+      "learning_rate": 1.3444045810705144e-06,
+      "loss": 0.6458,
+      "step": 13626
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.441859652990548,
+      "learning_rate": 1.3443138818963684e-06,
+      "loss": 0.6504,
+      "step": 13627
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.391849976567222,
+      "learning_rate": 1.3442231795087103e-06,
+      "loss": 0.6413,
+      "step": 13628
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.057611458100045,
+      "learning_rate": 1.3441324739083869e-06,
+      "loss": 0.6523,
+      "step": 13629
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.4155478536062445,
+      "learning_rate": 1.3440417650962445e-06,
+      "loss": 0.6283,
+      "step": 13630
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.7396859650212455,
+      "learning_rate": 1.34395105307313e-06,
+      "loss": 0.653,
+      "step": 13631
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.1255784916056295,
+      "learning_rate": 1.34386033783989e-06,
+      "loss": 0.6764,
+      "step": 13632
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.65322170336649,
+      "learning_rate": 1.343769619397371e-06,
+      "loss": 0.638,
+      "step": 13633
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 6.560587317709281,
+      "learning_rate": 1.34367889774642e-06,
+      "loss": 0.7005,
+      "step": 13634
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0804470243444064,
+      "learning_rate": 1.3435881728878828e-06,
+      "loss": 0.6159,
+      "step": 13635
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.8392653457238732,
+      "learning_rate": 1.3434974448226072e-06,
+      "loss": 0.6771,
+      "step": 13636
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.2044550314198474,
+      "learning_rate": 1.3434067135514396e-06,
+      "loss": 0.6374,
+      "step": 13637
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 4.100648670417481,
+      "learning_rate": 1.3433159790752266e-06,
+      "loss": 0.6191,
+      "step": 13638
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.8905548363971671,
+      "learning_rate": 1.3432252413948155e-06,
+      "loss": 0.6504,
+      "step": 13639
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.91542200156108,
+      "learning_rate": 1.3431345005110528e-06,
+      "loss": 0.6673,
+      "step": 13640
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.6574434980740245,
+      "learning_rate": 1.3430437564247851e-06,
+      "loss": 0.6634,
+      "step": 13641
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.3568785464030166,
+      "learning_rate": 1.3429530091368602e-06,
+      "loss": 0.6758,
+      "step": 13642
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6161190135481849,
+      "learning_rate": 1.3428622586481244e-06,
+      "loss": 0.6719,
+      "step": 13643
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.1029842160331103,
+      "learning_rate": 1.3427715049594243e-06,
+      "loss": 0.6667,
+      "step": 13644
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.4592413690753654,
+      "learning_rate": 1.342680748071608e-06,
+      "loss": 0.6322,
+      "step": 13645
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.6849659595866038,
+      "learning_rate": 1.342589987985522e-06,
+      "loss": 0.6432,
+      "step": 13646
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.3421926749496986,
+      "learning_rate": 1.342499224702013e-06,
+      "loss": 0.6582,
+      "step": 13647
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.944160164736662,
+      "learning_rate": 1.3424084582219286e-06,
+      "loss": 0.6107,
+      "step": 13648
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.4641776760617957,
+      "learning_rate": 1.3423176885461156e-06,
+      "loss": 0.6354,
+      "step": 13649
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.918197188711974,
+      "learning_rate": 1.3422269156754217e-06,
+      "loss": 0.6862,
+      "step": 13650
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.56564979061439,
+      "learning_rate": 1.3421361396106935e-06,
+      "loss": 0.6914,
+      "step": 13651
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.031921699328621,
+      "learning_rate": 1.3420453603527785e-06,
+      "loss": 0.6387,
+      "step": 13652
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.7010822218741508,
+      "learning_rate": 1.341954577902524e-06,
+      "loss": 0.6413,
+      "step": 13653
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.567985459280802,
+      "learning_rate": 1.3418637922607768e-06,
+      "loss": 0.6257,
+      "step": 13654
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.7976663865085367,
+      "learning_rate": 1.3417730034283852e-06,
+      "loss": 0.6478,
+      "step": 13655
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.069008763696545,
+      "learning_rate": 1.3416822114061957e-06,
+      "loss": 0.625,
+      "step": 13656
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.799907505463558,
+      "learning_rate": 1.3415914161950556e-06,
+      "loss": 0.6191,
+      "step": 13657
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.337906445823828,
+      "learning_rate": 1.341500617795813e-06,
+      "loss": 0.6445,
+      "step": 13658
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.962731822503073,
+      "learning_rate": 1.3414098162093146e-06,
+      "loss": 0.6406,
+      "step": 13659
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.1621500775730764,
+      "learning_rate": 1.3413190114364083e-06,
+      "loss": 0.6406,
+      "step": 13660
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.8812615057699749,
+      "learning_rate": 1.3412282034779416e-06,
+      "loss": 0.6712,
+      "step": 13661
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.5023556443106263,
+      "learning_rate": 1.3411373923347617e-06,
+      "loss": 0.6699,
+      "step": 13662
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.576934915486853,
+      "learning_rate": 1.3410465780077166e-06,
+      "loss": 0.627,
+      "step": 13663
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.345652434664925,
+      "learning_rate": 1.3409557604976537e-06,
+      "loss": 0.6478,
+      "step": 13664
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.3704047805758393,
+      "learning_rate": 1.34086493980542e-06,
+      "loss": 0.6458,
+      "step": 13665
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.900817671629744,
+      "learning_rate": 1.3407741159318643e-06,
+      "loss": 0.6439,
+      "step": 13666
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.2289949747386424,
+      "learning_rate": 1.3406832888778335e-06,
+      "loss": 0.666,
+      "step": 13667
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.861322690971462,
+      "learning_rate": 1.340592458644175e-06,
+      "loss": 0.6361,
+      "step": 13668
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.990015800287051,
+      "learning_rate": 1.3405016252317376e-06,
+      "loss": 0.6361,
+      "step": 13669
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.043840025673521,
+      "learning_rate": 1.3404107886413677e-06,
+      "loss": 0.6406,
+      "step": 13670
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.405602057709659,
+      "learning_rate": 1.3403199488739142e-06,
+      "loss": 0.6478,
+      "step": 13671
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.173715584358305,
+      "learning_rate": 1.3402291059302244e-06,
+      "loss": 0.6367,
+      "step": 13672
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.354034848662325,
+      "learning_rate": 1.3401382598111466e-06,
+      "loss": 0.6126,
+      "step": 13673
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.094412621867122,
+      "learning_rate": 1.3400474105175281e-06,
+      "loss": 0.6634,
+      "step": 13674
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.677459319634354,
+      "learning_rate": 1.339956558050217e-06,
+      "loss": 0.6354,
+      "step": 13675
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.4370683281202945,
+      "learning_rate": 1.3398657024100613e-06,
+      "loss": 0.6549,
+      "step": 13676
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.169722701655665,
+      "learning_rate": 1.339774843597909e-06,
+      "loss": 0.696,
+      "step": 13677
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.1538596480795595,
+      "learning_rate": 1.339683981614608e-06,
+      "loss": 0.6107,
+      "step": 13678
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.762284906617368,
+      "learning_rate": 1.3395931164610066e-06,
+      "loss": 0.6276,
+      "step": 13679
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.2322311426843,
+      "learning_rate": 1.3395022481379524e-06,
+      "loss": 0.6458,
+      "step": 13680
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.2280066433612373,
+      "learning_rate": 1.3394113766462934e-06,
+      "loss": 0.6253,
+      "step": 13681
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 6.303022305255659,
+      "learning_rate": 1.3393205019868783e-06,
+      "loss": 0.6478,
+      "step": 13682
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.420730449125169,
+      "learning_rate": 1.339229624160555e-06,
+      "loss": 0.6159,
+      "step": 13683
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.164787163900372,
+      "learning_rate": 1.3391387431681715e-06,
+      "loss": 0.6217,
+      "step": 13684
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.305873909570769,
+      "learning_rate": 1.3390478590105761e-06,
+      "loss": 0.6836,
+      "step": 13685
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.89937764416694,
+      "learning_rate": 1.3389569716886171e-06,
+      "loss": 0.6491,
+      "step": 13686
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.1489332874788114,
+      "learning_rate": 1.338866081203143e-06,
+      "loss": 0.6484,
+      "step": 13687
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.2196457048383675,
+      "learning_rate": 1.3387751875550013e-06,
+      "loss": 0.6693,
+      "step": 13688
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.455308178235422,
+      "learning_rate": 1.338684290745041e-06,
+      "loss": 0.679,
+      "step": 13689
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.3837874377704975,
+      "learning_rate": 1.3385933907741105e-06,
+      "loss": 0.6341,
+      "step": 13690
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.0196031397927143,
+      "learning_rate": 1.3385024876430575e-06,
+      "loss": 0.6165,
+      "step": 13691
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.112708374550765,
+      "learning_rate": 1.338411581352731e-06,
+      "loss": 0.6263,
+      "step": 13692
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.1740802120894536,
+      "learning_rate": 1.3383206719039795e-06,
+      "loss": 0.6419,
+      "step": 13693
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.638155809568693,
+      "learning_rate": 1.338229759297651e-06,
+      "loss": 0.6484,
+      "step": 13694
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.1336769174905106,
+      "learning_rate": 1.3381388435345944e-06,
+      "loss": 0.6673,
+      "step": 13695
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.8082614207142194,
+      "learning_rate": 1.3380479246156582e-06,
+      "loss": 0.6165,
+      "step": 13696
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.2067790786976147,
+      "learning_rate": 1.3379570025416905e-06,
+      "loss": 0.6582,
+      "step": 13697
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.2619857989557794,
+      "learning_rate": 1.3378660773135405e-06,
+      "loss": 0.6283,
+      "step": 13698
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.001803797347964,
+      "learning_rate": 1.3377751489320564e-06,
+      "loss": 0.6725,
+      "step": 13699
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 5.018416256047323,
+      "learning_rate": 1.337684217398087e-06,
+      "loss": 0.64,
+      "step": 13700
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.114418913011732,
+      "learning_rate": 1.3375932827124808e-06,
+      "loss": 0.6257,
+      "step": 13701
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.060203008605947,
+      "learning_rate": 1.3375023448760865e-06,
+      "loss": 0.6445,
+      "step": 13702
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.8497979736026418,
+      "learning_rate": 1.3374114038897535e-06,
+      "loss": 0.679,
+      "step": 13703
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.4705868782351446,
+      "learning_rate": 1.3373204597543295e-06,
+      "loss": 0.653,
+      "step": 13704
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.2822749590883564,
+      "learning_rate": 1.337229512470664e-06,
+      "loss": 0.6445,
+      "step": 13705
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.8487676814131548,
+      "learning_rate": 1.3371385620396058e-06,
+      "loss": 0.6328,
+      "step": 13706
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.0550929687949018,
+      "learning_rate": 1.3370476084620035e-06,
+      "loss": 0.6178,
+      "step": 13707
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 4.550968643117786,
+      "learning_rate": 1.336956651738706e-06,
+      "loss": 0.6576,
+      "step": 13708
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.932576745887366,
+      "learning_rate": 1.3368656918705623e-06,
+      "loss": 0.6224,
+      "step": 13709
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.0940887483484136,
+      "learning_rate": 1.3367747288584212e-06,
+      "loss": 0.6068,
+      "step": 13710
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.4040239465018023,
+      "learning_rate": 1.336683762703132e-06,
+      "loss": 0.6361,
+      "step": 13711
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.569205565207677,
+      "learning_rate": 1.3365927934055433e-06,
+      "loss": 0.6738,
+      "step": 13712
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.487051757377246,
+      "learning_rate": 1.3365018209665043e-06,
+      "loss": 0.6582,
+      "step": 13713
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.020977465726065,
+      "learning_rate": 1.3364108453868642e-06,
+      "loss": 0.6452,
+      "step": 13714
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.650372780829897,
+      "learning_rate": 1.3363198666674716e-06,
+      "loss": 0.6211,
+      "step": 13715
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.4311410873760955,
+      "learning_rate": 1.3362288848091763e-06,
+      "loss": 0.6712,
+      "step": 13716
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.6980299377670907,
+      "learning_rate": 1.3361378998128267e-06,
+      "loss": 0.6146,
+      "step": 13717
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.292034140353089,
+      "learning_rate": 1.3360469116792726e-06,
+      "loss": 0.6191,
+      "step": 13718
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.552265316475421,
+      "learning_rate": 1.3359559204093632e-06,
+      "loss": 0.6725,
+      "step": 13719
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.139734184674025,
+      "learning_rate": 1.3358649260039472e-06,
+      "loss": 0.6523,
+      "step": 13720
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.4379516441129487,
+      "learning_rate": 1.335773928463874e-06,
+      "loss": 0.681,
+      "step": 13721
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.0039240088647436,
+      "learning_rate": 1.3356829277899935e-06,
+      "loss": 0.6875,
+      "step": 13722
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.373176455996195,
+      "learning_rate": 1.3355919239831539e-06,
+      "loss": 0.6478,
+      "step": 13723
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.696592267570243,
+      "learning_rate": 1.3355009170442056e-06,
+      "loss": 0.6016,
+      "step": 13724
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.5001481231023477,
+      "learning_rate": 1.3354099069739976e-06,
+      "loss": 0.6432,
+      "step": 13725
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.035917399997758,
+      "learning_rate": 1.335318893773379e-06,
+      "loss": 0.6354,
+      "step": 13726
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.8958680579087837,
+      "learning_rate": 1.3352278774432e-06,
+      "loss": 0.651,
+      "step": 13727
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.025354131648789,
+      "learning_rate": 1.335136857984309e-06,
+      "loss": 0.6497,
+      "step": 13728
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.0701323691593596,
+      "learning_rate": 1.3350458353975565e-06,
+      "loss": 0.6576,
+      "step": 13729
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.064590581063908,
+      "learning_rate": 1.3349548096837913e-06,
+      "loss": 0.6608,
+      "step": 13730
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.342009152350213,
+      "learning_rate": 1.3348637808438632e-06,
+      "loss": 0.6491,
+      "step": 13731
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.209710456678855,
+      "learning_rate": 1.3347727488786226e-06,
+      "loss": 0.6126,
+      "step": 13732
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.53618282028034,
+      "learning_rate": 1.3346817137889175e-06,
+      "loss": 0.6393,
+      "step": 13733
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.1480934021159332,
+      "learning_rate": 1.3345906755755985e-06,
+      "loss": 0.6426,
+      "step": 13734
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 6.12436155481708,
+      "learning_rate": 1.3344996342395151e-06,
+      "loss": 0.6621,
+      "step": 13735
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.000173997238445,
+      "learning_rate": 1.334408589781517e-06,
+      "loss": 0.6562,
+      "step": 13736
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.9458230127065925,
+      "learning_rate": 1.3343175422024544e-06,
+      "loss": 0.623,
+      "step": 13737
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.198038763606718,
+      "learning_rate": 1.3342264915031762e-06,
+      "loss": 0.6855,
+      "step": 13738
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.0759357452308325,
+      "learning_rate": 1.3341354376845328e-06,
+      "loss": 0.6517,
+      "step": 13739
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.999288399322119,
+      "learning_rate": 1.3340443807473737e-06,
+      "loss": 0.6393,
+      "step": 13740
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.7318654630655805,
+      "learning_rate": 1.3339533206925489e-06,
+      "loss": 0.6009,
+      "step": 13741
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.9947101039645376,
+      "learning_rate": 1.3338622575209082e-06,
+      "loss": 0.6602,
+      "step": 13742
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.717738855698955,
+      "learning_rate": 1.3337711912333017e-06,
+      "loss": 0.6908,
+      "step": 13743
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.0163456164846005,
+      "learning_rate": 1.3336801218305787e-06,
+      "loss": 0.6432,
+      "step": 13744
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.604799771906471,
+      "learning_rate": 1.3335890493135902e-06,
+      "loss": 0.61,
+      "step": 13745
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.4692135617741404,
+      "learning_rate": 1.333497973683185e-06,
+      "loss": 0.6406,
+      "step": 13746
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.698273425177567,
+      "learning_rate": 1.333406894940214e-06,
+      "loss": 0.6094,
+      "step": 13747
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.1415963633373587,
+      "learning_rate": 1.3333158130855271e-06,
+      "loss": 0.6302,
+      "step": 13748
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.423675392963359,
+      "learning_rate": 1.3332247281199743e-06,
+      "loss": 0.5863,
+      "step": 13749
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.568879387224677,
+      "learning_rate": 1.3331336400444053e-06,
+      "loss": 0.6484,
+      "step": 13750
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.950694392643686,
+      "learning_rate": 1.3330425488596708e-06,
+      "loss": 0.623,
+      "step": 13751
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.2266230498857733,
+      "learning_rate": 1.3329514545666206e-06,
+      "loss": 0.6471,
+      "step": 13752
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.1411565087142237,
+      "learning_rate": 1.3328603571661051e-06,
+      "loss": 0.6087,
+      "step": 13753
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.360469994863931,
+      "learning_rate": 1.3327692566589743e-06,
+      "loss": 0.6296,
+      "step": 13754
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.80461448795364,
+      "learning_rate": 1.3326781530460788e-06,
+      "loss": 0.6758,
+      "step": 13755
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.872506177856161,
+      "learning_rate": 1.3325870463282689e-06,
+      "loss": 0.638,
+      "step": 13756
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.7822645971020954,
+      "learning_rate": 1.3324959365063944e-06,
+      "loss": 0.6217,
+      "step": 13757
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.503527699843004,
+      "learning_rate": 1.332404823581306e-06,
+      "loss": 0.6771,
+      "step": 13758
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.9922114615363715,
+      "learning_rate": 1.3323137075538537e-06,
+      "loss": 0.6478,
+      "step": 13759
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.285791822841517,
+      "learning_rate": 1.3322225884248886e-06,
+      "loss": 0.6335,
+      "step": 13760
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.4668785695824695,
+      "learning_rate": 1.3321314661952605e-06,
+      "loss": 0.6491,
+      "step": 13761
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.4484359478441897,
+      "learning_rate": 1.33204034086582e-06,
+      "loss": 0.6849,
+      "step": 13762
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.7656980804181335,
+      "learning_rate": 1.3319492124374177e-06,
+      "loss": 0.5957,
+      "step": 13763
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.167070430655269,
+      "learning_rate": 1.3318580809109042e-06,
+      "loss": 0.7031,
+      "step": 13764
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.8492835597132156,
+      "learning_rate": 1.3317669462871294e-06,
+      "loss": 0.6504,
+      "step": 13765
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.0882802553706648,
+      "learning_rate": 1.331675808566945e-06,
+      "loss": 0.6152,
+      "step": 13766
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 5.845562998256177,
+      "learning_rate": 1.3315846677512003e-06,
+      "loss": 0.7246,
+      "step": 13767
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.2132766847120373,
+      "learning_rate": 1.3314935238407469e-06,
+      "loss": 0.6615,
+      "step": 13768
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.2189066436950204,
+      "learning_rate": 1.3314023768364355e-06,
+      "loss": 0.64,
+      "step": 13769
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.9481078446221174,
+      "learning_rate": 1.3313112267391158e-06,
+      "loss": 0.6191,
+      "step": 13770
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.0892762477784763,
+      "learning_rate": 1.3312200735496392e-06,
+      "loss": 0.6374,
+      "step": 13771
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 4.290247508662385,
+      "learning_rate": 1.331128917268857e-06,
+      "loss": 0.6868,
+      "step": 13772
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.0829706474604954,
+      "learning_rate": 1.3310377578976185e-06,
+      "loss": 0.627,
+      "step": 13773
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.9799261330856857,
+      "learning_rate": 1.3309465954367757e-06,
+      "loss": 0.6738,
+      "step": 13774
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.121909938646436,
+      "learning_rate": 1.3308554298871792e-06,
+      "loss": 0.6328,
+      "step": 13775
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.399422132512429,
+      "learning_rate": 1.3307642612496794e-06,
+      "loss": 0.6732,
+      "step": 13776
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.299841216914005,
+      "learning_rate": 1.3306730895251277e-06,
+      "loss": 0.6842,
+      "step": 13777
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.513476286927116,
+      "learning_rate": 1.3305819147143747e-06,
+      "loss": 0.6751,
+      "step": 13778
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.240644648679385,
+      "learning_rate": 1.3304907368182714e-06,
+      "loss": 0.6895,
+      "step": 13779
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.697503004098666,
+      "learning_rate": 1.330399555837669e-06,
+      "loss": 0.6667,
+      "step": 13780
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.731051738076965,
+      "learning_rate": 1.3303083717734183e-06,
+      "loss": 0.6576,
+      "step": 13781
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.917324024311845,
+      "learning_rate": 1.3302171846263703e-06,
+      "loss": 0.6595,
+      "step": 13782
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.499722602297964,
+      "learning_rate": 1.3301259943973762e-06,
+      "loss": 0.6536,
+      "step": 13783
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.9405181293829366,
+      "learning_rate": 1.3300348010872869e-06,
+      "loss": 0.6387,
+      "step": 13784
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.071425028495375,
+      "learning_rate": 1.329943604696954e-06,
+      "loss": 0.64,
+      "step": 13785
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8312392969378817,
+      "learning_rate": 1.3298524052272276e-06,
+      "loss": 0.6868,
+      "step": 13786
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.476623426436841,
+      "learning_rate": 1.3297612026789598e-06,
+      "loss": 0.6309,
+      "step": 13787
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.100767497063008,
+      "learning_rate": 1.3296699970530017e-06,
+      "loss": 0.6478,
+      "step": 13788
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.0264435621436054,
+      "learning_rate": 1.3295787883502042e-06,
+      "loss": 0.6673,
+      "step": 13789
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.0299925017443097,
+      "learning_rate": 1.3294875765714186e-06,
+      "loss": 0.6445,
+      "step": 13790
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.588238879369781,
+      "learning_rate": 1.3293963617174964e-06,
+      "loss": 0.653,
+      "step": 13791
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.703441890570828,
+      "learning_rate": 1.329305143789289e-06,
+      "loss": 0.6699,
+      "step": 13792
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8636878278256956,
+      "learning_rate": 1.3292139227876475e-06,
+      "loss": 0.6634,
+      "step": 13793
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.041864388888658,
+      "learning_rate": 1.3291226987134231e-06,
+      "loss": 0.6517,
+      "step": 13794
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.823710325451892,
+      "learning_rate": 1.3290314715674678e-06,
+      "loss": 0.6927,
+      "step": 13795
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.026709805854554,
+      "learning_rate": 1.3289402413506325e-06,
+      "loss": 0.6549,
+      "step": 13796
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.9846001960809962,
+      "learning_rate": 1.3288490080637689e-06,
+      "loss": 0.6562,
+      "step": 13797
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.2265946752661625,
+      "learning_rate": 1.3287577717077285e-06,
+      "loss": 0.6497,
+      "step": 13798
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8196726306778648,
+      "learning_rate": 1.3286665322833626e-06,
+      "loss": 0.6419,
+      "step": 13799
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.1367386191806674,
+      "learning_rate": 1.328575289791523e-06,
+      "loss": 0.6113,
+      "step": 13800
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.8280847176913406,
+      "learning_rate": 1.328484044233061e-06,
+      "loss": 0.6764,
+      "step": 13801
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.9686094220077714,
+      "learning_rate": 1.3283927956088285e-06,
+      "loss": 0.6641,
+      "step": 13802
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8738439977031984,
+      "learning_rate": 1.328301543919677e-06,
+      "loss": 0.6406,
+      "step": 13803
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.826021459095901,
+      "learning_rate": 1.3282102891664579e-06,
+      "loss": 0.627,
+      "step": 13804
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.643802070019372,
+      "learning_rate": 1.3281190313500234e-06,
+      "loss": 0.6243,
+      "step": 13805
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.2474507843715577,
+      "learning_rate": 1.3280277704712252e-06,
+      "loss": 0.6569,
+      "step": 13806
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.245922710266816,
+      "learning_rate": 1.3279365065309142e-06,
+      "loss": 0.6276,
+      "step": 13807
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.820161772138568,
+      "learning_rate": 1.327845239529943e-06,
+      "loss": 0.6217,
+      "step": 13808
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.422170655922268,
+      "learning_rate": 1.3277539694691635e-06,
+      "loss": 0.6478,
+      "step": 13809
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.1640488414145733,
+      "learning_rate": 1.327662696349427e-06,
+      "loss": 0.6504,
+      "step": 13810
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.1313321753720014,
+      "learning_rate": 1.3275714201715855e-06,
+      "loss": 0.6602,
+      "step": 13811
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.10703351999231,
+      "learning_rate": 1.3274801409364912e-06,
+      "loss": 0.6224,
+      "step": 13812
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.619281030079863,
+      "learning_rate": 1.3273888586449955e-06,
+      "loss": 0.6842,
+      "step": 13813
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.5609035145044676,
+      "learning_rate": 1.3272975732979508e-06,
+      "loss": 0.6628,
+      "step": 13814
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.95612504897589,
+      "learning_rate": 1.3272062848962088e-06,
+      "loss": 0.6393,
+      "step": 13815
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.086118918188606,
+      "learning_rate": 1.327114993440622e-06,
+      "loss": 0.6777,
+      "step": 13816
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.2984152208613704,
+      "learning_rate": 1.3270236989320415e-06,
+      "loss": 0.6263,
+      "step": 13817
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.9660819045383415,
+      "learning_rate": 1.32693240137132e-06,
+      "loss": 0.6569,
+      "step": 13818
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.470414167954572,
+      "learning_rate": 1.32684110075931e-06,
+      "loss": 0.6081,
+      "step": 13819
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.712125890770199,
+      "learning_rate": 1.3267497970968625e-06,
+      "loss": 0.6602,
+      "step": 13820
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.3387909158417144,
+      "learning_rate": 1.3266584903848303e-06,
+      "loss": 0.668,
+      "step": 13821
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.023951173266732,
+      "learning_rate": 1.326567180624066e-06,
+      "loss": 0.6322,
+      "step": 13822
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.3835255837565756,
+      "learning_rate": 1.3264758678154206e-06,
+      "loss": 0.6908,
+      "step": 13823
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.003749118454782,
+      "learning_rate": 1.3263845519597477e-06,
+      "loss": 0.6348,
+      "step": 13824
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8514406330546804,
+      "learning_rate": 1.3262932330578985e-06,
+      "loss": 0.6315,
+      "step": 13825
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.0638575644414043,
+      "learning_rate": 1.3262019111107258e-06,
+      "loss": 0.6628,
+      "step": 13826
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.9840694893571795,
+      "learning_rate": 1.326110586119082e-06,
+      "loss": 0.6497,
+      "step": 13827
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.363503061613652,
+      "learning_rate": 1.3260192580838193e-06,
+      "loss": 0.6654,
+      "step": 13828
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.657327481040882,
+      "learning_rate": 1.3259279270057897e-06,
+      "loss": 0.6126,
+      "step": 13829
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8626317356683109,
+      "learning_rate": 1.3258365928858462e-06,
+      "loss": 0.6113,
+      "step": 13830
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.8281914095946488,
+      "learning_rate": 1.3257452557248408e-06,
+      "loss": 0.6283,
+      "step": 13831
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.571132568312137,
+      "learning_rate": 1.3256539155236261e-06,
+      "loss": 0.6686,
+      "step": 13832
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.9430685978792759,
+      "learning_rate": 1.3255625722830547e-06,
+      "loss": 0.6393,
+      "step": 13833
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8486465898207265,
+      "learning_rate": 1.3254712260039788e-06,
+      "loss": 0.6484,
+      "step": 13834
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.610684571891913,
+      "learning_rate": 1.3253798766872517e-06,
+      "loss": 0.6315,
+      "step": 13835
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.355367263697948,
+      "learning_rate": 1.325288524333725e-06,
+      "loss": 0.6491,
+      "step": 13836
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.1121258125572333,
+      "learning_rate": 1.3251971689442521e-06,
+      "loss": 0.6328,
+      "step": 13837
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.173705341364245,
+      "learning_rate": 1.325105810519685e-06,
+      "loss": 0.6673,
+      "step": 13838
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.1333270549368355,
+      "learning_rate": 1.3250144490608763e-06,
+      "loss": 0.6699,
+      "step": 13839
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 3.240077174160661,
+      "learning_rate": 1.3249230845686796e-06,
+      "loss": 0.6673,
+      "step": 13840
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.586420504256988,
+      "learning_rate": 1.3248317170439468e-06,
+      "loss": 0.6198,
+      "step": 13841
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.6809821466293284,
+      "learning_rate": 1.3247403464875306e-06,
+      "loss": 0.6322,
+      "step": 13842
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.2675152468217723,
+      "learning_rate": 1.3246489729002844e-06,
+      "loss": 0.6465,
+      "step": 13843
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 5.845433471158114,
+      "learning_rate": 1.32455759628306e-06,
+      "loss": 0.6914,
+      "step": 13844
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 4.22733796167694,
+      "learning_rate": 1.3244662166367117e-06,
+      "loss": 0.6855,
+      "step": 13845
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.164664102187886,
+      "learning_rate": 1.324374833962091e-06,
+      "loss": 0.6751,
+      "step": 13846
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.786891213948318,
+      "learning_rate": 1.3242834482600512e-06,
+      "loss": 0.6517,
+      "step": 13847
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1724321227355157,
+      "learning_rate": 1.3241920595314459e-06,
+      "loss": 0.5944,
+      "step": 13848
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 6.186871853334639,
+      "learning_rate": 1.3241006677771268e-06,
+      "loss": 0.7227,
+      "step": 13849
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1905249019928967,
+      "learning_rate": 1.3240092729979477e-06,
+      "loss": 0.6628,
+      "step": 13850
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.2729972750758956,
+      "learning_rate": 1.3239178751947618e-06,
+      "loss": 0.6113,
+      "step": 13851
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1122073263967884,
+      "learning_rate": 1.3238264743684213e-06,
+      "loss": 0.6452,
+      "step": 13852
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.8233345559264589,
+      "learning_rate": 1.32373507051978e-06,
+      "loss": 0.6628,
+      "step": 13853
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.5854310078797895,
+      "learning_rate": 1.3236436636496902e-06,
+      "loss": 0.6006,
+      "step": 13854
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.7619698046934853,
+      "learning_rate": 1.3235522537590055e-06,
+      "loss": 0.6074,
+      "step": 13855
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.0360769978781943,
+      "learning_rate": 1.3234608408485794e-06,
+      "loss": 0.6159,
+      "step": 13856
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.986830065416827,
+      "learning_rate": 1.3233694249192643e-06,
+      "loss": 0.6211,
+      "step": 13857
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.2110357326881602,
+      "learning_rate": 1.323278005971914e-06,
+      "loss": 0.6556,
+      "step": 13858
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1406999014450605,
+      "learning_rate": 1.3231865840073814e-06,
+      "loss": 0.6237,
+      "step": 13859
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.292904872988112,
+      "learning_rate": 1.3230951590265198e-06,
+      "loss": 0.6335,
+      "step": 13860
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.263632119819142,
+      "learning_rate": 1.3230037310301825e-06,
+      "loss": 0.6413,
+      "step": 13861
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.2332337218400644,
+      "learning_rate": 1.3229123000192228e-06,
+      "loss": 0.6914,
+      "step": 13862
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.5803020119125497,
+      "learning_rate": 1.322820865994494e-06,
+      "loss": 0.6556,
+      "step": 13863
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.996497331736874,
+      "learning_rate": 1.3227294289568498e-06,
+      "loss": 0.6549,
+      "step": 13864
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.0012789493108905,
+      "learning_rate": 1.322637988907143e-06,
+      "loss": 0.61,
+      "step": 13865
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.0906364514905875,
+      "learning_rate": 1.3225465458462274e-06,
+      "loss": 0.6361,
+      "step": 13866
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.739086695282214,
+      "learning_rate": 1.3224550997749565e-06,
+      "loss": 0.6549,
+      "step": 13867
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.426177200473305,
+      "learning_rate": 1.3223636506941836e-06,
+      "loss": 0.6497,
+      "step": 13868
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.4478709541698125,
+      "learning_rate": 1.3222721986047622e-06,
+      "loss": 0.6328,
+      "step": 13869
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.6194812652253647,
+      "learning_rate": 1.3221807435075457e-06,
+      "loss": 0.6029,
+      "step": 13870
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.731631464422215,
+      "learning_rate": 1.322089285403388e-06,
+      "loss": 0.6146,
+      "step": 13871
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.4621160817179275,
+      "learning_rate": 1.3219978242931424e-06,
+      "loss": 0.6243,
+      "step": 13872
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.679492820142655,
+      "learning_rate": 1.3219063601776626e-06,
+      "loss": 0.6745,
+      "step": 13873
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.5820715261904446,
+      "learning_rate": 1.3218148930578023e-06,
+      "loss": 0.6751,
+      "step": 13874
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.2263263089993197,
+      "learning_rate": 1.3217234229344154e-06,
+      "loss": 0.6322,
+      "step": 13875
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.6796972363542233,
+      "learning_rate": 1.321631949808355e-06,
+      "loss": 0.6589,
+      "step": 13876
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.5800673949869823,
+      "learning_rate": 1.3215404736804753e-06,
+      "loss": 0.653,
+      "step": 13877
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.094769735470674,
+      "learning_rate": 1.32144899455163e-06,
+      "loss": 0.6393,
+      "step": 13878
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.5231326724629315,
+      "learning_rate": 1.3213575124226728e-06,
+      "loss": 0.6406,
+      "step": 13879
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.6479425738599125,
+      "learning_rate": 1.3212660272944575e-06,
+      "loss": 0.6374,
+      "step": 13880
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.35470676407272,
+      "learning_rate": 1.321174539167838e-06,
+      "loss": 0.6608,
+      "step": 13881
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.442125200242429,
+      "learning_rate": 1.321083048043668e-06,
+      "loss": 0.6361,
+      "step": 13882
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.2987295523058604,
+      "learning_rate": 1.3209915539228016e-06,
+      "loss": 0.6686,
+      "step": 13883
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.624878385789162,
+      "learning_rate": 1.3209000568060926e-06,
+      "loss": 0.679,
+      "step": 13884
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.246032293775682,
+      "learning_rate": 1.3208085566943949e-06,
+      "loss": 0.6393,
+      "step": 13885
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.0829292861205677,
+      "learning_rate": 1.3207170535885628e-06,
+      "loss": 0.6055,
+      "step": 13886
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.2878372459866747,
+      "learning_rate": 1.3206255474894499e-06,
+      "loss": 0.6471,
+      "step": 13887
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1211230745805563,
+      "learning_rate": 1.3205340383979104e-06,
+      "loss": 0.6146,
+      "step": 13888
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.8117874055028693,
+      "learning_rate": 1.3204425263147985e-06,
+      "loss": 0.6211,
+      "step": 13889
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.095051740819382,
+      "learning_rate": 1.3203510112409681e-06,
+      "loss": 0.6341,
+      "step": 13890
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.9934125003734837,
+      "learning_rate": 1.3202594931772734e-06,
+      "loss": 0.6686,
+      "step": 13891
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.260796962801286,
+      "learning_rate": 1.3201679721245684e-06,
+      "loss": 0.6107,
+      "step": 13892
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.426600773537462,
+      "learning_rate": 1.3200764480837076e-06,
+      "loss": 0.6914,
+      "step": 13893
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.1580491585744586,
+      "learning_rate": 1.319984921055545e-06,
+      "loss": 0.6589,
+      "step": 13894
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.19596255621131,
+      "learning_rate": 1.3198933910409345e-06,
+      "loss": 0.6289,
+      "step": 13895
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.698883363107001,
+      "learning_rate": 1.319801858040731e-06,
+      "loss": 0.6445,
+      "step": 13896
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.4031100083153887,
+      "learning_rate": 1.3197103220557883e-06,
+      "loss": 0.6641,
+      "step": 13897
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 5.121335801927704,
+      "learning_rate": 1.3196187830869612e-06,
+      "loss": 0.6504,
+      "step": 13898
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.5598934779303635,
+      "learning_rate": 1.3195272411351035e-06,
+      "loss": 0.6419,
+      "step": 13899
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.2581619127664823,
+      "learning_rate": 1.3194356962010694e-06,
+      "loss": 0.6491,
+      "step": 13900
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 4.720261223657902,
+      "learning_rate": 1.3193441482857147e-06,
+      "loss": 0.6536,
+      "step": 13901
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.7549660760210055,
+      "learning_rate": 1.3192525973898921e-06,
+      "loss": 0.6823,
+      "step": 13902
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.268244135170035,
+      "learning_rate": 1.319161043514457e-06,
+      "loss": 0.6595,
+      "step": 13903
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.177535122617673,
+      "learning_rate": 1.3190694866602637e-06,
+      "loss": 0.6471,
+      "step": 13904
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.0521607067018968,
+      "learning_rate": 1.3189779268281665e-06,
+      "loss": 0.6113,
+      "step": 13905
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.2824886108480213,
+      "learning_rate": 1.31888636401902e-06,
+      "loss": 0.6602,
+      "step": 13906
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.7716311200717945,
+      "learning_rate": 1.3187947982336792e-06,
+      "loss": 0.6263,
+      "step": 13907
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.03745012928237,
+      "learning_rate": 1.3187032294729982e-06,
+      "loss": 0.6367,
+      "step": 13908
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.142427461400779,
+      "learning_rate": 1.3186116577378316e-06,
+      "loss": 0.6562,
+      "step": 13909
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 3.7666614387672155,
+      "learning_rate": 1.3185200830290342e-06,
+      "loss": 0.6465,
+      "step": 13910
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.156466970367543,
+      "learning_rate": 1.3184285053474609e-06,
+      "loss": 0.6562,
+      "step": 13911
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.7310848185976595,
+      "learning_rate": 1.318336924693966e-06,
+      "loss": 0.6536,
+      "step": 13912
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.139611215105596,
+      "learning_rate": 1.3182453410694048e-06,
+      "loss": 0.5964,
+      "step": 13913
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.2537591895778766,
+      "learning_rate": 1.3181537544746318e-06,
+      "loss": 0.6458,
+      "step": 13914
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.4418218719351286,
+      "learning_rate": 1.318062164910501e-06,
+      "loss": 0.6504,
+      "step": 13915
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.5172476201770313,
+      "learning_rate": 1.317970572377868e-06,
+      "loss": 0.6536,
+      "step": 13916
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.829840242069281,
+      "learning_rate": 1.3178789768775883e-06,
+      "loss": 0.6458,
+      "step": 13917
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.0497025476715973,
+      "learning_rate": 1.3177873784105152e-06,
+      "loss": 0.6602,
+      "step": 13918
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.3468141656046346,
+      "learning_rate": 1.3176957769775049e-06,
+      "loss": 0.653,
+      "step": 13919
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.225643438837969,
+      "learning_rate": 1.3176041725794114e-06,
+      "loss": 0.625,
+      "step": 13920
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.3815780893743366,
+      "learning_rate": 1.3175125652170901e-06,
+      "loss": 0.666,
+      "step": 13921
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.634369280778054,
+      "learning_rate": 1.3174209548913962e-06,
+      "loss": 0.6582,
+      "step": 13922
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.320234496456539,
+      "learning_rate": 1.3173293416031843e-06,
+      "loss": 0.6771,
+      "step": 13923
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.9353761168153298,
+      "learning_rate": 1.3172377253533095e-06,
+      "loss": 0.6322,
+      "step": 13924
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 6.007887430121914,
+      "learning_rate": 1.3171461061426272e-06,
+      "loss": 0.6309,
+      "step": 13925
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.124295761312196,
+      "learning_rate": 1.3170544839719921e-06,
+      "loss": 0.6641,
+      "step": 13926
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.5213831473688635,
+      "learning_rate": 1.3169628588422597e-06,
+      "loss": 0.6745,
+      "step": 13927
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.1697825532683384,
+      "learning_rate": 1.3168712307542846e-06,
+      "loss": 0.623,
+      "step": 13928
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.8805287009753586,
+      "learning_rate": 1.3167795997089224e-06,
+      "loss": 0.651,
+      "step": 13929
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.499934393889355,
+      "learning_rate": 1.316687965707028e-06,
+      "loss": 0.6523,
+      "step": 13930
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.5411511974214225,
+      "learning_rate": 1.3165963287494573e-06,
+      "loss": 0.6016,
+      "step": 13931
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.523079308569891,
+      "learning_rate": 1.3165046888370647e-06,
+      "loss": 0.6517,
+      "step": 13932
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.299822112265228,
+      "learning_rate": 1.3164130459707057e-06,
+      "loss": 0.651,
+      "step": 13933
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.171393713373506,
+      "learning_rate": 1.316321400151236e-06,
+      "loss": 0.6738,
+      "step": 13934
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.961354734417333,
+      "learning_rate": 1.3162297513795108e-06,
+      "loss": 0.6204,
+      "step": 13935
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.2773391466726953,
+      "learning_rate": 1.3161380996563851e-06,
+      "loss": 0.6139,
+      "step": 13936
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.137281913783716,
+      "learning_rate": 1.3160464449827146e-06,
+      "loss": 0.6478,
+      "step": 13937
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.6262694954579078,
+      "learning_rate": 1.315954787359355e-06,
+      "loss": 0.612,
+      "step": 13938
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.5868698034784128,
+      "learning_rate": 1.3158631267871609e-06,
+      "loss": 0.6523,
+      "step": 13939
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.3835969247841953,
+      "learning_rate": 1.3157714632669887e-06,
+      "loss": 0.6042,
+      "step": 13940
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.398281167267324,
+      "learning_rate": 1.3156797967996935e-06,
+      "loss": 0.6289,
+      "step": 13941
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.653170762884248,
+      "learning_rate": 1.3155881273861305e-06,
+      "loss": 0.6387,
+      "step": 13942
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.871150616556487,
+      "learning_rate": 1.315496455027156e-06,
+      "loss": 0.6764,
+      "step": 13943
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.228802812608975,
+      "learning_rate": 1.3154047797236251e-06,
+      "loss": 0.6647,
+      "step": 13944
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.0499368753435085,
+      "learning_rate": 1.3153131014763933e-06,
+      "loss": 0.6159,
+      "step": 13945
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.805727978725046,
+      "learning_rate": 1.3152214202863165e-06,
+      "loss": 0.6217,
+      "step": 13946
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.875521331133765,
+      "learning_rate": 1.3151297361542506e-06,
+      "loss": 0.6289,
+      "step": 13947
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.0271440610482916,
+      "learning_rate": 1.3150380490810506e-06,
+      "loss": 0.6413,
+      "step": 13948
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.261271972695732,
+      "learning_rate": 1.314946359067573e-06,
+      "loss": 0.6758,
+      "step": 13949
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.6163689365003804,
+      "learning_rate": 1.3148546661146728e-06,
+      "loss": 0.6517,
+      "step": 13950
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.034742978888434,
+      "learning_rate": 1.3147629702232063e-06,
+      "loss": 0.6335,
+      "step": 13951
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.4329399284284823,
+      "learning_rate": 1.314671271394029e-06,
+      "loss": 0.6784,
+      "step": 13952
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.329848691009587,
+      "learning_rate": 1.3145795696279974e-06,
+      "loss": 0.6719,
+      "step": 13953
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.7476492854396612,
+      "learning_rate": 1.3144878649259666e-06,
+      "loss": 0.6908,
+      "step": 13954
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.259250225462121,
+      "learning_rate": 1.3143961572887925e-06,
+      "loss": 0.6227,
+      "step": 13955
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.50963790125373,
+      "learning_rate": 1.3143044467173313e-06,
+      "loss": 0.5921,
+      "step": 13956
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.1735037641973376,
+      "learning_rate": 1.3142127332124392e-06,
+      "loss": 0.6016,
+      "step": 13957
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.3951246323329234,
+      "learning_rate": 1.3141210167749717e-06,
+      "loss": 0.6908,
+      "step": 13958
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.750884509441376,
+      "learning_rate": 1.314029297405785e-06,
+      "loss": 0.6048,
+      "step": 13959
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.5473447494460344,
+      "learning_rate": 1.3139375751057352e-06,
+      "loss": 0.6536,
+      "step": 13960
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.183841627563956,
+      "learning_rate": 1.313845849875678e-06,
+      "loss": 0.6302,
+      "step": 13961
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.698450138481281,
+      "learning_rate": 1.31375412171647e-06,
+      "loss": 0.6452,
+      "step": 13962
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.74981487023842,
+      "learning_rate": 1.3136623906289668e-06,
+      "loss": 0.6283,
+      "step": 13963
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.7214759727135776,
+      "learning_rate": 1.313570656614025e-06,
+      "loss": 0.6667,
+      "step": 13964
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.190454415569204,
+      "learning_rate": 1.3134789196725002e-06,
+      "loss": 0.6458,
+      "step": 13965
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.01937385682406,
+      "learning_rate": 1.3133871798052492e-06,
+      "loss": 0.6484,
+      "step": 13966
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.3382815276736872,
+      "learning_rate": 1.313295437013128e-06,
+      "loss": 0.627,
+      "step": 13967
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.2419294894325996,
+      "learning_rate": 1.3132036912969925e-06,
+      "loss": 0.6387,
+      "step": 13968
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.7603171262712136,
+      "learning_rate": 1.3131119426576996e-06,
+      "loss": 0.707,
+      "step": 13969
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.2932547189267396,
+      "learning_rate": 1.313020191096105e-06,
+      "loss": 0.6836,
+      "step": 13970
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.388985931429113,
+      "learning_rate": 1.312928436613065e-06,
+      "loss": 0.6097,
+      "step": 13971
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 4.094740509499872,
+      "learning_rate": 1.3128366792094366e-06,
+      "loss": 0.6309,
+      "step": 13972
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.9447041292129965,
+      "learning_rate": 1.312744918886076e-06,
+      "loss": 0.6257,
+      "step": 13973
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.268061025788881,
+      "learning_rate": 1.312653155643839e-06,
+      "loss": 0.6556,
+      "step": 13974
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.417093074979135,
+      "learning_rate": 1.3125613894835827e-06,
+      "loss": 0.6556,
+      "step": 13975
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 3.0539128828025426,
+      "learning_rate": 1.3124696204061631e-06,
+      "loss": 0.6243,
+      "step": 13976
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.4243004868794578,
+      "learning_rate": 1.312377848412437e-06,
+      "loss": 0.6576,
+      "step": 13977
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.359424640918365,
+      "learning_rate": 1.3122860735032611e-06,
+      "loss": 0.6771,
+      "step": 13978
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 5.183208264040546,
+      "learning_rate": 1.3121942956794912e-06,
+      "loss": 0.6426,
+      "step": 13979
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.073700310893715,
+      "learning_rate": 1.3121025149419846e-06,
+      "loss": 0.6523,
+      "step": 13980
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.7735648857073465,
+      "learning_rate": 1.3120107312915976e-06,
+      "loss": 0.6797,
+      "step": 13981
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.5401276831429036,
+      "learning_rate": 1.3119189447291867e-06,
+      "loss": 0.6751,
+      "step": 13982
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.0568415214916542,
+      "learning_rate": 1.3118271552556092e-06,
+      "loss": 0.6204,
+      "step": 13983
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.2405336210666316,
+      "learning_rate": 1.3117353628717204e-06,
+      "loss": 0.6341,
+      "step": 13984
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.037078253561951,
+      "learning_rate": 1.3116435675783785e-06,
+      "loss": 0.6478,
+      "step": 13985
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.040137436267887,
+      "learning_rate": 1.3115517693764396e-06,
+      "loss": 0.6484,
+      "step": 13986
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.8886062097477883,
+      "learning_rate": 1.3114599682667602e-06,
+      "loss": 0.6413,
+      "step": 13987
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.121228160712867,
+      "learning_rate": 1.3113681642501976e-06,
+      "loss": 0.6185,
+      "step": 13988
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.2170440507638,
+      "learning_rate": 1.3112763573276081e-06,
+      "loss": 0.6126,
+      "step": 13989
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.1047758622877755,
+      "learning_rate": 1.3111845474998494e-06,
+      "loss": 0.6309,
+      "step": 13990
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.262592833096236,
+      "learning_rate": 1.3110927347677777e-06,
+      "loss": 0.6686,
+      "step": 13991
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.739262101419198,
+      "learning_rate": 1.3110009191322494e-06,
+      "loss": 0.6016,
+      "step": 13992
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.2922299033587943,
+      "learning_rate": 1.3109091005941229e-06,
+      "loss": 0.6328,
+      "step": 13993
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.7248524919905464,
+      "learning_rate": 1.3108172791542535e-06,
+      "loss": 0.6673,
+      "step": 13994
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.469443790019274,
+      "learning_rate": 1.310725454813499e-06,
+      "loss": 0.6673,
+      "step": 13995
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.693595459040533,
+      "learning_rate": 1.3106336275727166e-06,
+      "loss": 0.6302,
+      "step": 13996
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.1442582709440288,
+      "learning_rate": 1.310541797432763e-06,
+      "loss": 0.6126,
+      "step": 13997
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.416849083846049,
+      "learning_rate": 1.3104499643944956e-06,
+      "loss": 0.6634,
+      "step": 13998
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.464838855124471,
+      "learning_rate": 1.3103581284587708e-06,
+      "loss": 0.6152,
+      "step": 13999
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.375699596506862,
+      "learning_rate": 1.3102662896264465e-06,
+      "loss": 0.6699,
+      "step": 14000
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.4822844126912362,
+      "learning_rate": 1.3101744478983794e-06,
+      "loss": 0.694,
+      "step": 14001
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.691394611113899,
+      "learning_rate": 1.3100826032754264e-06,
+      "loss": 0.653,
+      "step": 14002
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.488353626601446,
+      "learning_rate": 1.3099907557584456e-06,
+      "loss": 0.6367,
+      "step": 14003
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.5978734354816955,
+      "learning_rate": 1.3098989053482935e-06,
+      "loss": 0.64,
+      "step": 14004
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.500591430135355,
+      "learning_rate": 1.3098070520458271e-06,
+      "loss": 0.6882,
+      "step": 14005
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.822545291144264,
+      "learning_rate": 1.3097151958519043e-06,
+      "loss": 0.6986,
+      "step": 14006
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.4063583061105915,
+      "learning_rate": 1.3096233367673825e-06,
+      "loss": 0.681,
+      "step": 14007
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.4655855936286444,
+      "learning_rate": 1.3095314747931184e-06,
+      "loss": 0.6504,
+      "step": 14008
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.80885185826291,
+      "learning_rate": 1.3094396099299699e-06,
+      "loss": 0.6257,
+      "step": 14009
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.022848192232941,
+      "learning_rate": 1.3093477421787937e-06,
+      "loss": 0.6315,
+      "step": 14010
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.116968481379542,
+      "learning_rate": 1.3092558715404481e-06,
+      "loss": 0.6432,
+      "step": 14011
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.323287659845757,
+      "learning_rate": 1.3091639980157897e-06,
+      "loss": 0.623,
+      "step": 14012
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.154082940421053,
+      "learning_rate": 1.3090721216056771e-06,
+      "loss": 0.6172,
+      "step": 14013
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.5110652017014554,
+      "learning_rate": 1.3089802423109663e-06,
+      "loss": 0.6647,
+      "step": 14014
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.8974814878584845,
+      "learning_rate": 1.3088883601325158e-06,
+      "loss": 0.64,
+      "step": 14015
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.084452429268707,
+      "learning_rate": 1.3087964750711833e-06,
+      "loss": 0.6367,
+      "step": 14016
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.9195282091532198,
+      "learning_rate": 1.3087045871278254e-06,
+      "loss": 0.6367,
+      "step": 14017
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.808441817813912,
+      "learning_rate": 1.3086126963033006e-06,
+      "loss": 0.6413,
+      "step": 14018
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.4319982602123904,
+      "learning_rate": 1.3085208025984664e-06,
+      "loss": 0.6732,
+      "step": 14019
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.9675965615831925,
+      "learning_rate": 1.30842890601418e-06,
+      "loss": 0.6178,
+      "step": 14020
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.3824254164431946,
+      "learning_rate": 1.3083370065512992e-06,
+      "loss": 0.6536,
+      "step": 14021
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.167775184016157,
+      "learning_rate": 1.308245104210682e-06,
+      "loss": 0.6178,
+      "step": 14022
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.1109796967201757,
+      "learning_rate": 1.3081531989931862e-06,
+      "loss": 0.6465,
+      "step": 14023
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.275058682693567,
+      "learning_rate": 1.308061290899669e-06,
+      "loss": 0.6654,
+      "step": 14024
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.0556929338961005,
+      "learning_rate": 1.3079693799309887e-06,
+      "loss": 0.6328,
+      "step": 14025
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.131959977707898,
+      "learning_rate": 1.3078774660880031e-06,
+      "loss": 0.6263,
+      "step": 14026
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.526636201691364,
+      "learning_rate": 1.3077855493715694e-06,
+      "loss": 0.6764,
+      "step": 14027
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.2251185585977,
+      "learning_rate": 1.3076936297825465e-06,
+      "loss": 0.638,
+      "step": 14028
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.2495833482747134,
+      "learning_rate": 1.3076017073217913e-06,
+      "loss": 0.6725,
+      "step": 14029
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.5547437453635577,
+      "learning_rate": 1.3075097819901629e-06,
+      "loss": 0.6549,
+      "step": 14030
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.486338022853924,
+      "learning_rate": 1.3074178537885176e-06,
+      "loss": 0.6156,
+      "step": 14031
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.0827616210177915,
+      "learning_rate": 1.3073259227177147e-06,
+      "loss": 0.6296,
+      "step": 14032
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.306761871870497,
+      "learning_rate": 1.307233988778612e-06,
+      "loss": 0.6562,
+      "step": 14033
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.206319354328709,
+      "learning_rate": 1.3071420519720668e-06,
+      "loss": 0.5938,
+      "step": 14034
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.417430392572754,
+      "learning_rate": 1.3070501122989379e-06,
+      "loss": 0.6562,
+      "step": 14035
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.443353306584782,
+      "learning_rate": 1.3069581697600832e-06,
+      "loss": 0.6556,
+      "step": 14036
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.0303924226315146,
+      "learning_rate": 1.3068662243563608e-06,
+      "loss": 0.668,
+      "step": 14037
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.600325101860032,
+      "learning_rate": 1.3067742760886287e-06,
+      "loss": 0.6569,
+      "step": 14038
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.7301262016565557,
+      "learning_rate": 1.306682324957745e-06,
+      "loss": 0.6191,
+      "step": 14039
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.176902282727186,
+      "learning_rate": 1.3065903709645682e-06,
+      "loss": 0.6445,
+      "step": 14040
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 5.6854976763506855,
+      "learning_rate": 1.3064984141099564e-06,
+      "loss": 0.6354,
+      "step": 14041
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 4.156481923885818,
+      "learning_rate": 1.3064064543947678e-06,
+      "loss": 0.6146,
+      "step": 14042
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.71060233768252,
+      "learning_rate": 1.3063144918198608e-06,
+      "loss": 0.651,
+      "step": 14043
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.4686811992623516,
+      "learning_rate": 1.3062225263860934e-06,
+      "loss": 0.6335,
+      "step": 14044
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.337651366334313,
+      "learning_rate": 1.3061305580943239e-06,
+      "loss": 0.6315,
+      "step": 14045
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.2718384355066297,
+      "learning_rate": 1.3060385869454112e-06,
+      "loss": 0.6309,
+      "step": 14046
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.044308490110256,
+      "learning_rate": 1.3059466129402129e-06,
+      "loss": 0.6758,
+      "step": 14047
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.971691230329399,
+      "learning_rate": 1.305854636079588e-06,
+      "loss": 0.6706,
+      "step": 14048
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.94117300107784,
+      "learning_rate": 1.305762656364395e-06,
+      "loss": 0.6536,
+      "step": 14049
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.774729111432184,
+      "learning_rate": 1.3056706737954915e-06,
+      "loss": 0.6855,
+      "step": 14050
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.737299109206128,
+      "learning_rate": 1.305578688373737e-06,
+      "loss": 0.6393,
+      "step": 14051
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.7250640922589744,
+      "learning_rate": 1.3054867000999893e-06,
+      "loss": 0.61,
+      "step": 14052
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.8417168571003937,
+      "learning_rate": 1.3053947089751076e-06,
+      "loss": 0.5983,
+      "step": 14053
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 5.255387159100226,
+      "learning_rate": 1.3053027149999498e-06,
+      "loss": 0.6868,
+      "step": 14054
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.249993399561968,
+      "learning_rate": 1.3052107181753747e-06,
+      "loss": 0.61,
+      "step": 14055
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.856388160987972,
+      "learning_rate": 1.3051187185022411e-06,
+      "loss": 0.6654,
+      "step": 14056
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.8090768885626876,
+      "learning_rate": 1.3050267159814078e-06,
+      "loss": 0.6491,
+      "step": 14057
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.9344212358749284,
+      "learning_rate": 1.3049347106137325e-06,
+      "loss": 0.638,
+      "step": 14058
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.951433676570609,
+      "learning_rate": 1.3048427024000751e-06,
+      "loss": 0.5892,
+      "step": 14059
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.2385815924866503,
+      "learning_rate": 1.3047506913412939e-06,
+      "loss": 0.6862,
+      "step": 14060
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.910010095312655,
+      "learning_rate": 1.3046586774382471e-06,
+      "loss": 0.638,
+      "step": 14061
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.5112305039958107,
+      "learning_rate": 1.304566660691794e-06,
+      "loss": 0.6693,
+      "step": 14062
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.3597517513737114,
+      "learning_rate": 1.3044746411027936e-06,
+      "loss": 0.6732,
+      "step": 14063
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.2122101410538795,
+      "learning_rate": 1.3043826186721046e-06,
+      "loss": 0.6003,
+      "step": 14064
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.092791303375667,
+      "learning_rate": 1.304290593400585e-06,
+      "loss": 0.6087,
+      "step": 14065
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.2381723088342453,
+      "learning_rate": 1.304198565289095e-06,
+      "loss": 0.6341,
+      "step": 14066
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.070267983657381,
+      "learning_rate": 1.3041065343384927e-06,
+      "loss": 0.6745,
+      "step": 14067
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.771548117302386,
+      "learning_rate": 1.304014500549637e-06,
+      "loss": 0.6621,
+      "step": 14068
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.5159612280532517,
+      "learning_rate": 1.3039224639233874e-06,
+      "loss": 0.6393,
+      "step": 14069
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.9132652450151522,
+      "learning_rate": 1.3038304244606025e-06,
+      "loss": 0.626,
+      "step": 14070
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.586694195338884,
+      "learning_rate": 1.303738382162141e-06,
+      "loss": 0.6315,
+      "step": 14071
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.5757391657987405,
+      "learning_rate": 1.3036463370288628e-06,
+      "loss": 0.6602,
+      "step": 14072
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.9385034943765898,
+      "learning_rate": 1.3035542890616265e-06,
+      "loss": 0.5938,
+      "step": 14073
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.7053528879559665,
+      "learning_rate": 1.3034622382612907e-06,
+      "loss": 0.6315,
+      "step": 14074
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.795385600553459,
+      "learning_rate": 1.3033701846287154e-06,
+      "loss": 0.6882,
+      "step": 14075
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.501130722329823,
+      "learning_rate": 1.3032781281647593e-06,
+      "loss": 0.6374,
+      "step": 14076
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.187685699420716,
+      "learning_rate": 1.3031860688702816e-06,
+      "loss": 0.6517,
+      "step": 14077
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.8744000499192777,
+      "learning_rate": 1.3030940067461415e-06,
+      "loss": 0.6465,
+      "step": 14078
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.557733867006154,
+      "learning_rate": 1.303001941793198e-06,
+      "loss": 0.6628,
+      "step": 14079
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.604929412950054,
+      "learning_rate": 1.302909874012311e-06,
+      "loss": 0.6667,
+      "step": 14080
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.3708179818901396,
+      "learning_rate": 1.3028178034043392e-06,
+      "loss": 0.6367,
+      "step": 14081
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.1394646197510756,
+      "learning_rate": 1.3027257299701417e-06,
+      "loss": 0.6113,
+      "step": 14082
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.7190368606398545,
+      "learning_rate": 1.3026336537105787e-06,
+      "loss": 0.6224,
+      "step": 14083
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.9685606006715632,
+      "learning_rate": 1.3025415746265086e-06,
+      "loss": 0.6745,
+      "step": 14084
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.203667737948935,
+      "learning_rate": 1.3024494927187917e-06,
+      "loss": 0.6719,
+      "step": 14085
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.8492192658552753,
+      "learning_rate": 1.3023574079882867e-06,
+      "loss": 0.6484,
+      "step": 14086
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.5768891337493143,
+      "learning_rate": 1.302265320435853e-06,
+      "loss": 0.6719,
+      "step": 14087
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.6683041071269606,
+      "learning_rate": 1.3021732300623506e-06,
+      "loss": 0.6439,
+      "step": 14088
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.820167355749166,
+      "learning_rate": 1.3020811368686388e-06,
+      "loss": 0.61,
+      "step": 14089
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.470225836298866,
+      "learning_rate": 1.301989040855577e-06,
+      "loss": 0.6732,
+      "step": 14090
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.378496110739665,
+      "learning_rate": 1.3018969420240247e-06,
+      "loss": 0.6081,
+      "step": 14091
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.2692245109534652,
+      "learning_rate": 1.3018048403748417e-06,
+      "loss": 0.6419,
+      "step": 14092
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 7.006027562161865,
+      "learning_rate": 1.301712735908887e-06,
+      "loss": 0.6803,
+      "step": 14093
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.8360887231726086,
+      "learning_rate": 1.301620628627021e-06,
+      "loss": 0.6224,
+      "step": 14094
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 6.093665365850763,
+      "learning_rate": 1.3015285185301027e-06,
+      "loss": 0.6882,
+      "step": 14095
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.539727768168298,
+      "learning_rate": 1.3014364056189928e-06,
+      "loss": 0.6523,
+      "step": 14096
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.4654370336483176,
+      "learning_rate": 1.3013442898945496e-06,
+      "loss": 0.612,
+      "step": 14097
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.7465308263291215,
+      "learning_rate": 1.3012521713576336e-06,
+      "loss": 0.6517,
+      "step": 14098
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.620337385560622,
+      "learning_rate": 1.3011600500091046e-06,
+      "loss": 0.6732,
+      "step": 14099
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.914166636144451,
+      "learning_rate": 1.3010679258498222e-06,
+      "loss": 0.6185,
+      "step": 14100
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.703713925853438,
+      "learning_rate": 1.3009757988806462e-06,
+      "loss": 0.679,
+      "step": 14101
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.3833504759478026,
+      "learning_rate": 1.3008836691024366e-06,
+      "loss": 0.6341,
+      "step": 14102
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.8632875020444097,
+      "learning_rate": 1.3007915365160526e-06,
+      "loss": 0.6361,
+      "step": 14103
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.5687863348768656,
+      "learning_rate": 1.3006994011223551e-06,
+      "loss": 0.6628,
+      "step": 14104
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.4083656437542245,
+      "learning_rate": 1.300607262922203e-06,
+      "loss": 0.6618,
+      "step": 14105
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.1201571393189735,
+      "learning_rate": 1.3005151219164572e-06,
+      "loss": 0.6803,
+      "step": 14106
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.22237632338309,
+      "learning_rate": 1.3004229781059772e-06,
+      "loss": 0.6556,
+      "step": 14107
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.276319301582747,
+      "learning_rate": 1.3003308314916226e-06,
+      "loss": 0.6556,
+      "step": 14108
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 4.423779224030957,
+      "learning_rate": 1.300238682074254e-06,
+      "loss": 0.6322,
+      "step": 14109
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.466807904842791,
+      "learning_rate": 1.3001465298547314e-06,
+      "loss": 0.6992,
+      "step": 14110
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.9482735873453,
+      "learning_rate": 1.3000543748339142e-06,
+      "loss": 0.6191,
+      "step": 14111
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.628426146281082,
+      "learning_rate": 1.2999622170126637e-06,
+      "loss": 0.627,
+      "step": 14112
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.4339106750187347,
+      "learning_rate": 1.2998700563918388e-06,
+      "loss": 0.6328,
+      "step": 14113
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.2663672721891306,
+      "learning_rate": 1.2997778929723001e-06,
+      "loss": 0.6263,
+      "step": 14114
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.611042518793047,
+      "learning_rate": 1.2996857267549079e-06,
+      "loss": 0.6517,
+      "step": 14115
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.9037147137041246,
+      "learning_rate": 1.2995935577405226e-06,
+      "loss": 0.6667,
+      "step": 14116
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.9664680705476814,
+      "learning_rate": 1.299501385930004e-06,
+      "loss": 0.6582,
+      "step": 14117
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.3334257069849236,
+      "learning_rate": 1.299409211324212e-06,
+      "loss": 0.6393,
+      "step": 14118
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.4894538862619955,
+      "learning_rate": 1.299317033924008e-06,
+      "loss": 0.6628,
+      "step": 14119
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.178699479924641,
+      "learning_rate": 1.2992248537302515e-06,
+      "loss": 0.6771,
+      "step": 14120
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.385723397043175,
+      "learning_rate": 1.2991326707438026e-06,
+      "loss": 0.653,
+      "step": 14121
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.128421486785411,
+      "learning_rate": 1.2990404849655226e-06,
+      "loss": 0.627,
+      "step": 14122
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.9743708127595503,
+      "learning_rate": 1.2989482963962708e-06,
+      "loss": 0.6309,
+      "step": 14123
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.5848445045858304,
+      "learning_rate": 1.2988561050369083e-06,
+      "loss": 0.612,
+      "step": 14124
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.600555033201738,
+      "learning_rate": 1.2987639108882954e-06,
+      "loss": 0.6634,
+      "step": 14125
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.811280089407511,
+      "learning_rate": 1.2986717139512927e-06,
+      "loss": 0.6615,
+      "step": 14126
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.106999222367392,
+      "learning_rate": 1.29857951422676e-06,
+      "loss": 0.6556,
+      "step": 14127
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.449405506058259,
+      "learning_rate": 1.2984873117155586e-06,
+      "loss": 0.6452,
+      "step": 14128
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.2958834442817753,
+      "learning_rate": 1.2983951064185488e-06,
+      "loss": 0.6595,
+      "step": 14129
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.8194586750902495,
+      "learning_rate": 1.2983028983365908e-06,
+      "loss": 0.6465,
+      "step": 14130
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.063885449609387,
+      "learning_rate": 1.2982106874705455e-06,
+      "loss": 0.6855,
+      "step": 14131
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.2930392660471535,
+      "learning_rate": 1.2981184738212737e-06,
+      "loss": 0.6497,
+      "step": 14132
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.319144119959071,
+      "learning_rate": 1.298026257389636e-06,
+      "loss": 0.6276,
+      "step": 14133
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.799210475473489,
+      "learning_rate": 1.2979340381764924e-06,
+      "loss": 0.6621,
+      "step": 14134
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.3140307313196025,
+      "learning_rate": 1.2978418161827041e-06,
+      "loss": 0.6393,
+      "step": 14135
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.29522423140808,
+      "learning_rate": 1.297749591409132e-06,
+      "loss": 0.6862,
+      "step": 14136
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.6231189302161573,
+      "learning_rate": 1.2976573638566363e-06,
+      "loss": 0.6478,
+      "step": 14137
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.2387995956487283,
+      "learning_rate": 1.2975651335260787e-06,
+      "loss": 0.6758,
+      "step": 14138
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.11049505201342,
+      "learning_rate": 1.297472900418319e-06,
+      "loss": 0.6816,
+      "step": 14139
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.154490572615039,
+      "learning_rate": 1.2973806645342184e-06,
+      "loss": 0.6549,
+      "step": 14140
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.0700131611069863,
+      "learning_rate": 1.2972884258746379e-06,
+      "loss": 0.6426,
+      "step": 14141
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.9402659034724845,
+      "learning_rate": 1.2971961844404385e-06,
+      "loss": 0.6452,
+      "step": 14142
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.579711600120295,
+      "learning_rate": 1.2971039402324802e-06,
+      "loss": 0.6979,
+      "step": 14143
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.365519061547131,
+      "learning_rate": 1.2970116932516249e-06,
+      "loss": 0.638,
+      "step": 14144
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.3469763580551004,
+      "learning_rate": 1.2969194434987331e-06,
+      "loss": 0.6165,
+      "step": 14145
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.251678334221605,
+      "learning_rate": 1.2968271909746662e-06,
+      "loss": 0.6341,
+      "step": 14146
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.530739074266839,
+      "learning_rate": 1.2967349356802843e-06,
+      "loss": 0.6081,
+      "step": 14147
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.5859497474541078,
+      "learning_rate": 1.296642677616449e-06,
+      "loss": 0.6517,
+      "step": 14148
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.8956210026771703,
+      "learning_rate": 1.2965504167840223e-06,
+      "loss": 0.6452,
+      "step": 14149
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.5483358936311893,
+      "learning_rate": 1.2964581531838635e-06,
+      "loss": 0.6243,
+      "step": 14150
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.1869443433983866,
+      "learning_rate": 1.2963658868168349e-06,
+      "loss": 0.6289,
+      "step": 14151
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.287760929624699,
+      "learning_rate": 1.296273617683797e-06,
+      "loss": 0.6758,
+      "step": 14152
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.870035833824922,
+      "learning_rate": 1.2961813457856112e-06,
+      "loss": 0.6693,
+      "step": 14153
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.2550391506576615,
+      "learning_rate": 1.296089071123139e-06,
+      "loss": 0.668,
+      "step": 14154
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.709832969136977,
+      "learning_rate": 1.2959967936972412e-06,
+      "loss": 0.6647,
+      "step": 14155
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.8292315684874003,
+      "learning_rate": 1.295904513508779e-06,
+      "loss": 0.6257,
+      "step": 14156
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.269125529686266,
+      "learning_rate": 1.2958122305586138e-06,
+      "loss": 0.6452,
+      "step": 14157
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.4841219430862505,
+      "learning_rate": 1.2957199448476073e-06,
+      "loss": 0.6771,
+      "step": 14158
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 7.544661988243387,
+      "learning_rate": 1.29562765637662e-06,
+      "loss": 0.6556,
+      "step": 14159
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.3121561511369406,
+      "learning_rate": 1.2955353651465138e-06,
+      "loss": 0.5931,
+      "step": 14160
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.1336300997227102,
+      "learning_rate": 1.2954430711581497e-06,
+      "loss": 0.6133,
+      "step": 14161
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.288089332354023,
+      "learning_rate": 1.2953507744123897e-06,
+      "loss": 0.6738,
+      "step": 14162
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 5.114090132392826,
+      "learning_rate": 1.2952584749100944e-06,
+      "loss": 0.6908,
+      "step": 14163
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.173109856215186,
+      "learning_rate": 1.2951661726521254e-06,
+      "loss": 0.6035,
+      "step": 14164
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.190770960843817,
+      "learning_rate": 1.2950738676393452e-06,
+      "loss": 0.6074,
+      "step": 14165
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.4607397654961156,
+      "learning_rate": 1.294981559872614e-06,
+      "loss": 0.6712,
+      "step": 14166
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.263420246330259,
+      "learning_rate": 1.2948892493527938e-06,
+      "loss": 0.6523,
+      "step": 14167
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.9453013101818044,
+      "learning_rate": 1.2947969360807466e-06,
+      "loss": 0.7057,
+      "step": 14168
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 4.299932922842319,
+      "learning_rate": 1.2947046200573328e-06,
+      "loss": 0.651,
+      "step": 14169
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 6.973879415199715,
+      "learning_rate": 1.2946123012834154e-06,
+      "loss": 0.6999,
+      "step": 14170
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.512649430870439,
+      "learning_rate": 1.2945199797598546e-06,
+      "loss": 0.582,
+      "step": 14171
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.344328374636671,
+      "learning_rate": 1.2944276554875135e-06,
+      "loss": 0.6569,
+      "step": 14172
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.2717060174148482,
+      "learning_rate": 1.2943353284672529e-06,
+      "loss": 0.6842,
+      "step": 14173
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.645514328740593,
+      "learning_rate": 1.2942429986999344e-06,
+      "loss": 0.6348,
+      "step": 14174
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.898737669057107,
+      "learning_rate": 1.29415066618642e-06,
+      "loss": 0.5938,
+      "step": 14175
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.3881782004755845,
+      "learning_rate": 1.2940583309275717e-06,
+      "loss": 0.6139,
+      "step": 14176
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.6452918881899383,
+      "learning_rate": 1.2939659929242507e-06,
+      "loss": 0.6035,
+      "step": 14177
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.8292026298984094,
+      "learning_rate": 1.2938736521773195e-06,
+      "loss": 0.6413,
+      "step": 14178
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.9872427926155305,
+      "learning_rate": 1.293781308687639e-06,
+      "loss": 0.64,
+      "step": 14179
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.76703983762408,
+      "learning_rate": 1.2936889624560715e-06,
+      "loss": 0.6237,
+      "step": 14180
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.154027078920593,
+      "learning_rate": 1.2935966134834795e-06,
+      "loss": 0.6211,
+      "step": 14181
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.5613064202913836,
+      "learning_rate": 1.293504261770724e-06,
+      "loss": 0.6432,
+      "step": 14182
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.6726155176976185,
+      "learning_rate": 1.2934119073186676e-06,
+      "loss": 0.6549,
+      "step": 14183
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.1969458536606976,
+      "learning_rate": 1.2933195501281715e-06,
+      "loss": 0.6549,
+      "step": 14184
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.861665870480649,
+      "learning_rate": 1.293227190200098e-06,
+      "loss": 0.6777,
+      "step": 14185
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.6691881576076595,
+      "learning_rate": 1.2931348275353098e-06,
+      "loss": 0.5749,
+      "step": 14186
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.316532576764467,
+      "learning_rate": 1.2930424621346679e-06,
+      "loss": 0.6198,
+      "step": 14187
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.9873436551332455,
+      "learning_rate": 1.292950093999035e-06,
+      "loss": 0.6842,
+      "step": 14188
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.013300316069063,
+      "learning_rate": 1.2928577231292727e-06,
+      "loss": 0.6068,
+      "step": 14189
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.06598914481342,
+      "learning_rate": 1.2927653495262435e-06,
+      "loss": 0.6914,
+      "step": 14190
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.303683302837477,
+      "learning_rate": 1.2926729731908094e-06,
+      "loss": 0.6113,
+      "step": 14191
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.007297435125526,
+      "learning_rate": 1.2925805941238329e-06,
+      "loss": 0.5931,
+      "step": 14192
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.7600639051327462,
+      "learning_rate": 1.2924882123261752e-06,
+      "loss": 0.6406,
+      "step": 14193
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.5194556413804783,
+      "learning_rate": 1.2923958277986995e-06,
+      "loss": 0.6419,
+      "step": 14194
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.721017061128091,
+      "learning_rate": 1.2923034405422676e-06,
+      "loss": 0.6504,
+      "step": 14195
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.843090211072589,
+      "learning_rate": 1.292211050557742e-06,
+      "loss": 0.6484,
+      "step": 14196
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.6405173673094224,
+      "learning_rate": 1.2921186578459846e-06,
+      "loss": 0.6439,
+      "step": 14197
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.459100637853179,
+      "learning_rate": 1.292026262407858e-06,
+      "loss": 0.651,
+      "step": 14198
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.5806381937404703,
+      "learning_rate": 1.2919338642442249e-06,
+      "loss": 0.6491,
+      "step": 14199
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.8877044305816457,
+      "learning_rate": 1.2918414633559466e-06,
+      "loss": 0.6523,
+      "step": 14200
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.632024311093966,
+      "learning_rate": 1.2917490597438862e-06,
+      "loss": 0.612,
+      "step": 14201
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.381741443242542,
+      "learning_rate": 1.2916566534089063e-06,
+      "loss": 0.6146,
+      "step": 14202
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.486213447857525,
+      "learning_rate": 1.2915642443518689e-06,
+      "loss": 0.6634,
+      "step": 14203
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.8531931606461636,
+      "learning_rate": 1.2914718325736367e-06,
+      "loss": 0.5983,
+      "step": 14204
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.781318769990147,
+      "learning_rate": 1.291379418075072e-06,
+      "loss": 0.6315,
+      "step": 14205
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.9641048060511532,
+      "learning_rate": 1.2912870008570373e-06,
+      "loss": 0.64,
+      "step": 14206
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.545763389979362,
+      "learning_rate": 1.2911945809203956e-06,
+      "loss": 0.6523,
+      "step": 14207
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.4752746493342883,
+      "learning_rate": 1.291102158266009e-06,
+      "loss": 0.6471,
+      "step": 14208
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.7098848843739973,
+      "learning_rate": 1.2910097328947401e-06,
+      "loss": 0.6582,
+      "step": 14209
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.3501934418377624,
+      "learning_rate": 1.2909173048074518e-06,
+      "loss": 0.6289,
+      "step": 14210
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.948924904755176,
+      "learning_rate": 1.2908248740050064e-06,
+      "loss": 0.6107,
+      "step": 14211
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.736144310745111,
+      "learning_rate": 1.290732440488267e-06,
+      "loss": 0.6419,
+      "step": 14212
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.498231726121505,
+      "learning_rate": 1.2906400042580956e-06,
+      "loss": 0.6374,
+      "step": 14213
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.5212761917497883,
+      "learning_rate": 1.2905475653153555e-06,
+      "loss": 0.6452,
+      "step": 14214
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.8445995816505056,
+      "learning_rate": 1.2904551236609097e-06,
+      "loss": 0.6191,
+      "step": 14215
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.822887155951058,
+      "learning_rate": 1.29036267929562e-06,
+      "loss": 0.6693,
+      "step": 14216
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.165324356500572,
+      "learning_rate": 1.29027023222035e-06,
+      "loss": 0.6126,
+      "step": 14217
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.372909854863169,
+      "learning_rate": 1.2901777824359624e-06,
+      "loss": 0.623,
+      "step": 14218
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.406781662752044,
+      "learning_rate": 1.2900853299433194e-06,
+      "loss": 0.6003,
+      "step": 14219
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.5002492603683,
+      "learning_rate": 1.2899928747432847e-06,
+      "loss": 0.6999,
+      "step": 14220
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.7907898038855277,
+      "learning_rate": 1.289900416836721e-06,
+      "loss": 0.6562,
+      "step": 14221
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.569029733907951,
+      "learning_rate": 1.2898079562244906e-06,
+      "loss": 0.6328,
+      "step": 14222
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.5432863679096265,
+      "learning_rate": 1.289715492907457e-06,
+      "loss": 0.651,
+      "step": 14223
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.7127503692187585,
+      "learning_rate": 1.2896230268864833e-06,
+      "loss": 0.6107,
+      "step": 14224
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.1953941947588236,
+      "learning_rate": 1.2895305581624323e-06,
+      "loss": 0.6107,
+      "step": 14225
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.9450372374245024,
+      "learning_rate": 1.2894380867361671e-06,
+      "loss": 0.6439,
+      "step": 14226
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 6.516629433760076,
+      "learning_rate": 1.2893456126085501e-06,
+      "loss": 0.6126,
+      "step": 14227
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.4285842288597173,
+      "learning_rate": 1.2892531357804458e-06,
+      "loss": 0.6322,
+      "step": 14228
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.3066988472832364,
+      "learning_rate": 1.2891606562527158e-06,
+      "loss": 0.612,
+      "step": 14229
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.609044468965294,
+      "learning_rate": 1.2890681740262238e-06,
+      "loss": 0.6022,
+      "step": 14230
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.4759232555753816,
+      "learning_rate": 1.2889756891018331e-06,
+      "loss": 0.599,
+      "step": 14231
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.422722603652132,
+      "learning_rate": 1.2888832014804068e-06,
+      "loss": 0.6797,
+      "step": 14232
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.5543668344124426,
+      "learning_rate": 1.2887907111628077e-06,
+      "loss": 0.6901,
+      "step": 14233
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.3013853421568156,
+      "learning_rate": 1.2886982181499e-06,
+      "loss": 0.6576,
+      "step": 14234
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.424036085599298,
+      "learning_rate": 1.2886057224425457e-06,
+      "loss": 0.6302,
+      "step": 14235
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.4464148699978128,
+      "learning_rate": 1.2885132240416087e-06,
+      "loss": 0.6452,
+      "step": 14236
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.6404311913313347,
+      "learning_rate": 1.2884207229479525e-06,
+      "loss": 0.6237,
+      "step": 14237
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.477922447035,
+      "learning_rate": 1.2883282191624402e-06,
+      "loss": 0.6211,
+      "step": 14238
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.6578360460769446,
+      "learning_rate": 1.288235712685935e-06,
+      "loss": 0.6556,
+      "step": 14239
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.540707871565778,
+      "learning_rate": 1.2881432035193004e-06,
+      "loss": 0.6302,
+      "step": 14240
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 4.996362633030808,
+      "learning_rate": 1.2880506916634e-06,
+      "loss": 0.6387,
+      "step": 14241
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.1215405223165233,
+      "learning_rate": 1.2879581771190969e-06,
+      "loss": 0.6387,
+      "step": 14242
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.332994571008652,
+      "learning_rate": 1.2878656598872546e-06,
+      "loss": 0.6204,
+      "step": 14243
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.4254568659712925,
+      "learning_rate": 1.2877731399687368e-06,
+      "loss": 0.6374,
+      "step": 14244
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.497190711584487,
+      "learning_rate": 1.2876806173644066e-06,
+      "loss": 0.6322,
+      "step": 14245
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.889793440456418,
+      "learning_rate": 1.2875880920751278e-06,
+      "loss": 0.6595,
+      "step": 14246
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 5.638137896224925,
+      "learning_rate": 1.287495564101764e-06,
+      "loss": 0.6263,
+      "step": 14247
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.7169274993799535,
+      "learning_rate": 1.2874030334451785e-06,
+      "loss": 0.6576,
+      "step": 14248
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.1942860381299503,
+      "learning_rate": 1.2873105001062356e-06,
+      "loss": 0.6217,
+      "step": 14249
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.814712346551674,
+      "learning_rate": 1.287217964085798e-06,
+      "loss": 0.5931,
+      "step": 14250
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.4002932004562116,
+      "learning_rate": 1.2871254253847297e-06,
+      "loss": 0.6706,
+      "step": 14251
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.499048063815453,
+      "learning_rate": 1.2870328840038944e-06,
+      "loss": 0.6328,
+      "step": 14252
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.595477252909791,
+      "learning_rate": 1.2869403399441559e-06,
+      "loss": 0.6647,
+      "step": 14253
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.233040992603795,
+      "learning_rate": 1.2868477932063776e-06,
+      "loss": 0.6471,
+      "step": 14254
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.0531559316593126,
+      "learning_rate": 1.286755243791424e-06,
+      "loss": 0.6634,
+      "step": 14255
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.2949213151850705,
+      "learning_rate": 1.2866626917001578e-06,
+      "loss": 0.6562,
+      "step": 14256
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.010484363737371,
+      "learning_rate": 1.2865701369334438e-06,
+      "loss": 0.6641,
+      "step": 14257
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.6827157845380496,
+      "learning_rate": 1.2864775794921452e-06,
+      "loss": 0.6367,
+      "step": 14258
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.7653155787772454,
+      "learning_rate": 1.2863850193771258e-06,
+      "loss": 0.6823,
+      "step": 14259
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.386076277297341,
+      "learning_rate": 1.28629245658925e-06,
+      "loss": 0.6823,
+      "step": 14260
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.8781943189564645,
+      "learning_rate": 1.2861998911293813e-06,
+      "loss": 0.6895,
+      "step": 14261
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.7578657001424207,
+      "learning_rate": 1.2861073229983837e-06,
+      "loss": 0.6706,
+      "step": 14262
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.0070267931272494,
+      "learning_rate": 1.2860147521971209e-06,
+      "loss": 0.6595,
+      "step": 14263
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.276155493323584,
+      "learning_rate": 1.2859221787264574e-06,
+      "loss": 0.6725,
+      "step": 14264
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.0670065242420548,
+      "learning_rate": 1.2858296025872571e-06,
+      "loss": 0.6699,
+      "step": 14265
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.5307223277848236,
+      "learning_rate": 1.2857370237803833e-06,
+      "loss": 0.6341,
+      "step": 14266
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.039436515518532,
+      "learning_rate": 1.285644442306701e-06,
+      "loss": 0.6602,
+      "step": 14267
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.669518602234941,
+      "learning_rate": 1.2855518581670737e-06,
+      "loss": 0.6602,
+      "step": 14268
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.623271422215557,
+      "learning_rate": 1.2854592713623658e-06,
+      "loss": 0.6465,
+      "step": 14269
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.8803644781489632,
+      "learning_rate": 1.285366681893441e-06,
+      "loss": 0.6224,
+      "step": 14270
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.617719435607542,
+      "learning_rate": 1.285274089761164e-06,
+      "loss": 0.6361,
+      "step": 14271
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.4910442785080247,
+      "learning_rate": 1.2851814949663985e-06,
+      "loss": 0.6543,
+      "step": 14272
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.678281625371637,
+      "learning_rate": 1.285088897510009e-06,
+      "loss": 0.6686,
+      "step": 14273
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.8369771651619597,
+      "learning_rate": 1.2849962973928596e-06,
+      "loss": 0.6797,
+      "step": 14274
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.8408461909355114,
+      "learning_rate": 1.2849036946158146e-06,
+      "loss": 0.64,
+      "step": 14275
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.1551327810082053,
+      "learning_rate": 1.2848110891797381e-06,
+      "loss": 0.6654,
+      "step": 14276
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.8257579840828795,
+      "learning_rate": 1.2847184810854948e-06,
+      "loss": 0.6751,
+      "step": 14277
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.0637996934821228,
+      "learning_rate": 1.2846258703339482e-06,
+      "loss": 0.6322,
+      "step": 14278
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.9832806225663684,
+      "learning_rate": 1.2845332569259636e-06,
+      "loss": 0.6784,
+      "step": 14279
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.3855638370068135,
+      "learning_rate": 1.2844406408624047e-06,
+      "loss": 0.6374,
+      "step": 14280
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.7922970367700075,
+      "learning_rate": 1.2843480221441365e-06,
+      "loss": 0.653,
+      "step": 14281
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.006888172654611,
+      "learning_rate": 1.2842554007720228e-06,
+      "loss": 0.6478,
+      "step": 14282
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.026645154923924,
+      "learning_rate": 1.2841627767469283e-06,
+      "loss": 0.6777,
+      "step": 14283
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.025777475349999,
+      "learning_rate": 1.2840701500697179e-06,
+      "loss": 0.6439,
+      "step": 14284
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6929191376105182,
+      "learning_rate": 1.283977520741255e-06,
+      "loss": 0.6517,
+      "step": 14285
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.234800299314572,
+      "learning_rate": 1.283884888762405e-06,
+      "loss": 0.6289,
+      "step": 14286
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.4549225937194357,
+      "learning_rate": 1.2837922541340325e-06,
+      "loss": 0.6569,
+      "step": 14287
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.0494549540886258,
+      "learning_rate": 1.2836996168570014e-06,
+      "loss": 0.6921,
+      "step": 14288
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.9421055233707631,
+      "learning_rate": 1.2836069769321769e-06,
+      "loss": 0.638,
+      "step": 14289
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.9844820657753575,
+      "learning_rate": 1.2835143343604233e-06,
+      "loss": 0.6576,
+      "step": 14290
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.984119118711319,
+      "learning_rate": 1.2834216891426055e-06,
+      "loss": 0.6842,
+      "step": 14291
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.2825198409061582,
+      "learning_rate": 1.283329041279588e-06,
+      "loss": 0.6543,
+      "step": 14292
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.1868631937260066,
+      "learning_rate": 1.2832363907722352e-06,
+      "loss": 0.6445,
+      "step": 14293
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.6831025075360704,
+      "learning_rate": 1.2831437376214128e-06,
+      "loss": 0.6484,
+      "step": 14294
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.020379261891402,
+      "learning_rate": 1.2830510818279842e-06,
+      "loss": 0.6732,
+      "step": 14295
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.3608749981122834,
+      "learning_rate": 1.282958423392815e-06,
+      "loss": 0.6523,
+      "step": 14296
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.405406161712065,
+      "learning_rate": 1.28286576231677e-06,
+      "loss": 0.6771,
+      "step": 14297
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 6.172391307465908,
+      "learning_rate": 1.2827730986007135e-06,
+      "loss": 0.6777,
+      "step": 14298
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.499650469583569,
+      "learning_rate": 1.282680432245511e-06,
+      "loss": 0.6569,
+      "step": 14299
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.998835586604439,
+      "learning_rate": 1.2825877632520268e-06,
+      "loss": 0.6348,
+      "step": 14300
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.9788671263122677,
+      "learning_rate": 1.2824950916211262e-06,
+      "loss": 0.7044,
+      "step": 14301
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.626719543601096,
+      "learning_rate": 1.282402417353674e-06,
+      "loss": 0.6133,
+      "step": 14302
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.9175972043008314,
+      "learning_rate": 1.2823097404505345e-06,
+      "loss": 0.6406,
+      "step": 14303
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 5.019320880609217,
+      "learning_rate": 1.2822170609125738e-06,
+      "loss": 0.6367,
+      "step": 14304
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.2001161074841797,
+      "learning_rate": 1.282124378740656e-06,
+      "loss": 0.6452,
+      "step": 14305
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.072416876831584,
+      "learning_rate": 1.2820316939356466e-06,
+      "loss": 0.6842,
+      "step": 14306
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.494274285247469,
+      "learning_rate": 1.2819390064984103e-06,
+      "loss": 0.6354,
+      "step": 14307
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 4.340434477748056,
+      "learning_rate": 1.2818463164298125e-06,
+      "loss": 0.6465,
+      "step": 14308
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.1939010364296494,
+      "learning_rate": 1.281753623730718e-06,
+      "loss": 0.6309,
+      "step": 14309
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.402360289214628,
+      "learning_rate": 1.2816609284019922e-06,
+      "loss": 0.6777,
+      "step": 14310
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.2340465976949515,
+      "learning_rate": 1.2815682304445e-06,
+      "loss": 0.6094,
+      "step": 14311
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.221595434777645,
+      "learning_rate": 1.2814755298591063e-06,
+      "loss": 0.6562,
+      "step": 14312
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.0932173424131295,
+      "learning_rate": 1.2813828266466767e-06,
+      "loss": 0.6276,
+      "step": 14313
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.593299486411927,
+      "learning_rate": 1.2812901208080765e-06,
+      "loss": 0.6159,
+      "step": 14314
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 3.393834064830222,
+      "learning_rate": 1.2811974123441705e-06,
+      "loss": 0.638,
+      "step": 14315
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.5714438294728437,
+      "learning_rate": 1.2811047012558242e-06,
+      "loss": 0.6003,
+      "step": 14316
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.289644798164945,
+      "learning_rate": 1.281011987543903e-06,
+      "loss": 0.6087,
+      "step": 14317
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.802676995689498,
+      "learning_rate": 1.280919271209272e-06,
+      "loss": 0.6673,
+      "step": 14318
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.488485525546758,
+      "learning_rate": 1.2808265522527966e-06,
+      "loss": 0.6367,
+      "step": 14319
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.705454420558888,
+      "learning_rate": 1.280733830675342e-06,
+      "loss": 0.6816,
+      "step": 14320
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.6789555061308685,
+      "learning_rate": 1.2806411064777743e-06,
+      "loss": 0.6602,
+      "step": 14321
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.5663880453318777,
+      "learning_rate": 1.2805483796609577e-06,
+      "loss": 0.623,
+      "step": 14322
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.4674946626370664,
+      "learning_rate": 1.2804556502257583e-06,
+      "loss": 0.6823,
+      "step": 14323
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.470307992325265,
+      "learning_rate": 1.280362918173042e-06,
+      "loss": 0.6556,
+      "step": 14324
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.9172016111525236,
+      "learning_rate": 1.2802701835036733e-06,
+      "loss": 0.6764,
+      "step": 14325
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.793268690498671,
+      "learning_rate": 1.2801774462185184e-06,
+      "loss": 0.6283,
+      "step": 14326
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.307567009692997,
+      "learning_rate": 1.2800847063184427e-06,
+      "loss": 0.6361,
+      "step": 14327
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.960820689362724,
+      "learning_rate": 1.2799919638043118e-06,
+      "loss": 0.6335,
+      "step": 14328
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.7078754545633577,
+      "learning_rate": 1.2798992186769906e-06,
+      "loss": 0.6133,
+      "step": 14329
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.263709850727529,
+      "learning_rate": 1.2798064709373457e-06,
+      "loss": 0.6289,
+      "step": 14330
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.7858303880939297,
+      "learning_rate": 1.279713720586242e-06,
+      "loss": 0.6322,
+      "step": 14331
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.031680381571265,
+      "learning_rate": 1.2796209676245455e-06,
+      "loss": 0.6289,
+      "step": 14332
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.4534847809535987,
+      "learning_rate": 1.279528212053122e-06,
+      "loss": 0.651,
+      "step": 14333
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 7.112505322801568,
+      "learning_rate": 1.2794354538728367e-06,
+      "loss": 0.6576,
+      "step": 14334
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.3089221367182007,
+      "learning_rate": 1.2793426930845556e-06,
+      "loss": 0.6569,
+      "step": 14335
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.461935751213244,
+      "learning_rate": 1.2792499296891447e-06,
+      "loss": 0.6426,
+      "step": 14336
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.912685517050995,
+      "learning_rate": 1.2791571636874692e-06,
+      "loss": 0.6475,
+      "step": 14337
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.5157231982517976,
+      "learning_rate": 1.2790643950803952e-06,
+      "loss": 0.6126,
+      "step": 14338
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.8064392758964507,
+      "learning_rate": 1.2789716238687888e-06,
+      "loss": 0.6393,
+      "step": 14339
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.7813202587127854,
+      "learning_rate": 1.2788788500535154e-06,
+      "loss": 0.6263,
+      "step": 14340
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.6623238105691276,
+      "learning_rate": 1.278786073635441e-06,
+      "loss": 0.6497,
+      "step": 14341
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.899417201225184,
+      "learning_rate": 1.2786932946154316e-06,
+      "loss": 0.6204,
+      "step": 14342
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.3542829662186495,
+      "learning_rate": 1.2786005129943528e-06,
+      "loss": 0.6315,
+      "step": 14343
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.424016077603569,
+      "learning_rate": 1.2785077287730713e-06,
+      "loss": 0.6999,
+      "step": 14344
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.7892051878276924,
+      "learning_rate": 1.2784149419524521e-06,
+      "loss": 0.6582,
+      "step": 14345
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.7233057763425847,
+      "learning_rate": 1.2783221525333616e-06,
+      "loss": 0.6595,
+      "step": 14346
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.3153040746332842,
+      "learning_rate": 1.2782293605166663e-06,
+      "loss": 0.5996,
+      "step": 14347
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.6179941095525687,
+      "learning_rate": 1.2781365659032313e-06,
+      "loss": 0.6419,
+      "step": 14348
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 6.501462159709191,
+      "learning_rate": 1.278043768693923e-06,
+      "loss": 0.64,
+      "step": 14349
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.6329805341081,
+      "learning_rate": 1.277950968889608e-06,
+      "loss": 0.6725,
+      "step": 14350
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.712645441570771,
+      "learning_rate": 1.277858166491152e-06,
+      "loss": 0.6738,
+      "step": 14351
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.074296320236339,
+      "learning_rate": 1.277765361499421e-06,
+      "loss": 0.64,
+      "step": 14352
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.9324032721471185,
+      "learning_rate": 1.2776725539152815e-06,
+      "loss": 0.6361,
+      "step": 14353
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.1611995074447448,
+      "learning_rate": 1.2775797437395995e-06,
+      "loss": 0.6094,
+      "step": 14354
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.9260297102470934,
+      "learning_rate": 1.277486930973241e-06,
+      "loss": 0.6999,
+      "step": 14355
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.085119094622811,
+      "learning_rate": 1.2773941156170724e-06,
+      "loss": 0.6328,
+      "step": 14356
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.158862401662736,
+      "learning_rate": 1.2773012976719601e-06,
+      "loss": 0.681,
+      "step": 14357
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.8925685686149105,
+      "learning_rate": 1.2772084771387706e-06,
+      "loss": 0.6589,
+      "step": 14358
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.284447914515811,
+      "learning_rate": 1.2771156540183694e-06,
+      "loss": 0.6914,
+      "step": 14359
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.821471959172181,
+      "learning_rate": 1.2770228283116237e-06,
+      "loss": 0.6693,
+      "step": 14360
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.9839987442460854,
+      "learning_rate": 1.276930000019399e-06,
+      "loss": 0.6243,
+      "step": 14361
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.4903849818090724,
+      "learning_rate": 1.276837169142562e-06,
+      "loss": 0.61,
+      "step": 14362
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.835560270918959,
+      "learning_rate": 1.27674433568198e-06,
+      "loss": 0.6257,
+      "step": 14363
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.094130972801763,
+      "learning_rate": 1.2766514996385181e-06,
+      "loss": 0.6367,
+      "step": 14364
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.394987424844345,
+      "learning_rate": 1.2765586610130435e-06,
+      "loss": 0.6784,
+      "step": 14365
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.537434250716388,
+      "learning_rate": 1.276465819806422e-06,
+      "loss": 0.6224,
+      "step": 14366
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.9923943985317631,
+      "learning_rate": 1.276372976019521e-06,
+      "loss": 0.6055,
+      "step": 14367
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.657861690783009,
+      "learning_rate": 1.2762801296532064e-06,
+      "loss": 0.6641,
+      "step": 14368
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 5.700944465535807,
+      "learning_rate": 1.2761872807083448e-06,
+      "loss": 0.6458,
+      "step": 14369
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.1955545928657707,
+      "learning_rate": 1.2760944291858034e-06,
+      "loss": 0.6383,
+      "step": 14370
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.593337073744588,
+      "learning_rate": 1.2760015750864476e-06,
+      "loss": 0.6504,
+      "step": 14371
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.8906843705986134,
+      "learning_rate": 1.275908718411145e-06,
+      "loss": 0.6432,
+      "step": 14372
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.601587415580353,
+      "learning_rate": 1.2758158591607622e-06,
+      "loss": 0.6621,
+      "step": 14373
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.258031416253772,
+      "learning_rate": 1.2757229973361654e-06,
+      "loss": 0.6322,
+      "step": 14374
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.450460395290824,
+      "learning_rate": 1.2756301329382213e-06,
+      "loss": 0.6523,
+      "step": 14375
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.4686313162256295,
+      "learning_rate": 1.275537265967797e-06,
+      "loss": 0.6569,
+      "step": 14376
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.093842948327486,
+      "learning_rate": 1.2754443964257588e-06,
+      "loss": 0.6289,
+      "step": 14377
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.510602367215658,
+      "learning_rate": 1.2753515243129738e-06,
+      "loss": 0.6582,
+      "step": 14378
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.0817118768381233,
+      "learning_rate": 1.275258649630309e-06,
+      "loss": 0.6139,
+      "step": 14379
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.5130947520151334,
+      "learning_rate": 1.2751657723786305e-06,
+      "loss": 0.6562,
+      "step": 14380
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.5641773682579507,
+      "learning_rate": 1.2750728925588056e-06,
+      "loss": 0.6491,
+      "step": 14381
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 4.193486192375764,
+      "learning_rate": 1.274980010171701e-06,
+      "loss": 0.6615,
+      "step": 14382
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.647080078004876,
+      "learning_rate": 1.2748871252181839e-06,
+      "loss": 0.6582,
+      "step": 14383
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.954678888875232,
+      "learning_rate": 1.2747942376991207e-06,
+      "loss": 0.64,
+      "step": 14384
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.6508762270224975,
+      "learning_rate": 1.2747013476153785e-06,
+      "loss": 0.5964,
+      "step": 14385
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.4649682713536794,
+      "learning_rate": 1.2746084549678245e-06,
+      "loss": 0.6224,
+      "step": 14386
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.1033686456659857,
+      "learning_rate": 1.2745155597573256e-06,
+      "loss": 0.6452,
+      "step": 14387
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.639178012896704,
+      "learning_rate": 1.2744226619847484e-06,
+      "loss": 0.6634,
+      "step": 14388
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.9772914282375433,
+      "learning_rate": 1.2743297616509606e-06,
+      "loss": 0.6745,
+      "step": 14389
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.821960758209532,
+      "learning_rate": 1.2742368587568287e-06,
+      "loss": 0.6094,
+      "step": 14390
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.730566824450621,
+      "learning_rate": 1.2741439533032197e-06,
+      "loss": 0.6315,
+      "step": 14391
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.11575495239181,
+      "learning_rate": 1.2740510452910011e-06,
+      "loss": 0.6842,
+      "step": 14392
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.0142758231327185,
+      "learning_rate": 1.2739581347210401e-06,
+      "loss": 0.6315,
+      "step": 14393
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 7.648492209854851,
+      "learning_rate": 1.2738652215942037e-06,
+      "loss": 0.679,
+      "step": 14394
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.621379422590176,
+      "learning_rate": 1.2737723059113585e-06,
+      "loss": 0.651,
+      "step": 14395
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.9291080816226547,
+      "learning_rate": 1.2736793876733724e-06,
+      "loss": 0.6374,
+      "step": 14396
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.1364650544627133,
+      "learning_rate": 1.2735864668811123e-06,
+      "loss": 0.611,
+      "step": 14397
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.615692533352473,
+      "learning_rate": 1.2734935435354455e-06,
+      "loss": 0.6217,
+      "step": 14398
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.3649202363850907,
+      "learning_rate": 1.2734006176372394e-06,
+      "loss": 0.6621,
+      "step": 14399
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.707241234235636,
+      "learning_rate": 1.2733076891873611e-06,
+      "loss": 0.6914,
+      "step": 14400
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.4714930846212275,
+      "learning_rate": 1.2732147581866777e-06,
+      "loss": 0.6042,
+      "step": 14401
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.3757434045235866,
+      "learning_rate": 1.273121824636057e-06,
+      "loss": 0.6426,
+      "step": 14402
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.550017694337806,
+      "learning_rate": 1.2730288885363666e-06,
+      "loss": 0.6055,
+      "step": 14403
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.266206507222662,
+      "learning_rate": 1.2729359498884725e-06,
+      "loss": 0.6452,
+      "step": 14404
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.476219437175364,
+      "learning_rate": 1.2728430086932438e-06,
+      "loss": 0.735,
+      "step": 14405
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.062788548505642,
+      "learning_rate": 1.272750064951547e-06,
+      "loss": 0.6673,
+      "step": 14406
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.595104521344208,
+      "learning_rate": 1.2726571186642496e-06,
+      "loss": 0.6387,
+      "step": 14407
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.8994452548909324,
+      "learning_rate": 1.2725641698322194e-06,
+      "loss": 0.6621,
+      "step": 14408
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.3741363252483665,
+      "learning_rate": 1.2724712184563234e-06,
+      "loss": 0.6432,
+      "step": 14409
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.9171320374930754,
+      "learning_rate": 1.2723782645374297e-06,
+      "loss": 0.6576,
+      "step": 14410
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.457912589496075,
+      "learning_rate": 1.2722853080764053e-06,
+      "loss": 0.6523,
+      "step": 14411
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 6.560604298965864,
+      "learning_rate": 1.2721923490741183e-06,
+      "loss": 0.6419,
+      "step": 14412
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.2654571459035227,
+      "learning_rate": 1.2720993875314362e-06,
+      "loss": 0.6354,
+      "step": 14413
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.59633079265889,
+      "learning_rate": 1.2720064234492258e-06,
+      "loss": 0.6348,
+      "step": 14414
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.2927384491091125,
+      "learning_rate": 1.271913456828356e-06,
+      "loss": 0.6908,
+      "step": 14415
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.3687353475372874,
+      "learning_rate": 1.2718204876696936e-06,
+      "loss": 0.6133,
+      "step": 14416
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.3393939258442886,
+      "learning_rate": 1.2717275159741065e-06,
+      "loss": 0.6217,
+      "step": 14417
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.4143154080653613,
+      "learning_rate": 1.2716345417424626e-06,
+      "loss": 0.6152,
+      "step": 14418
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.5625752648916786,
+      "learning_rate": 1.2715415649756293e-06,
+      "loss": 0.6595,
+      "step": 14419
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.404956669186006,
+      "learning_rate": 1.2714485856744747e-06,
+      "loss": 0.6797,
+      "step": 14420
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.914165886686815,
+      "learning_rate": 1.2713556038398664e-06,
+      "loss": 0.6465,
+      "step": 14421
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.2889820360867863,
+      "learning_rate": 1.2712626194726724e-06,
+      "loss": 0.6211,
+      "step": 14422
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.3911438870459385,
+      "learning_rate": 1.2711696325737604e-06,
+      "loss": 0.6393,
+      "step": 14423
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.131110842740263,
+      "learning_rate": 1.271076643143998e-06,
+      "loss": 0.6224,
+      "step": 14424
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.6866194940910786,
+      "learning_rate": 1.2709836511842532e-06,
+      "loss": 0.6296,
+      "step": 14425
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.8148885776111596,
+      "learning_rate": 1.2708906566953947e-06,
+      "loss": 0.5736,
+      "step": 14426
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.7668089709143127,
+      "learning_rate": 1.2707976596782892e-06,
+      "loss": 0.6335,
+      "step": 14427
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.5786962853744866,
+      "learning_rate": 1.270704660133805e-06,
+      "loss": 0.6289,
+      "step": 14428
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.9787995316608042,
+      "learning_rate": 1.270611658062811e-06,
+      "loss": 0.6706,
+      "step": 14429
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.069642190483013,
+      "learning_rate": 1.270518653466174e-06,
+      "loss": 0.6367,
+      "step": 14430
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.7050478577073878,
+      "learning_rate": 1.2704256463447625e-06,
+      "loss": 0.6374,
+      "step": 14431
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.2952539013150837,
+      "learning_rate": 1.2703326366994444e-06,
+      "loss": 0.6452,
+      "step": 14432
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.4640732897530104,
+      "learning_rate": 1.270239624531088e-06,
+      "loss": 0.6296,
+      "step": 14433
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.129986741701044,
+      "learning_rate": 1.2701466098405615e-06,
+      "loss": 0.6784,
+      "step": 14434
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.830750926101436,
+      "learning_rate": 1.2700535926287326e-06,
+      "loss": 0.6523,
+      "step": 14435
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.714835502998659,
+      "learning_rate": 1.2699605728964694e-06,
+      "loss": 0.6465,
+      "step": 14436
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.172082500070012,
+      "learning_rate": 1.2698675506446407e-06,
+      "loss": 0.638,
+      "step": 14437
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.1554492061731487,
+      "learning_rate": 1.2697745258741142e-06,
+      "loss": 0.6491,
+      "step": 14438
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.222977408635115,
+      "learning_rate": 1.2696814985857583e-06,
+      "loss": 0.5905,
+      "step": 14439
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.7597916122370045,
+      "learning_rate": 1.2695884687804409e-06,
+      "loss": 0.6419,
+      "step": 14440
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.403424634638367,
+      "learning_rate": 1.2694954364590303e-06,
+      "loss": 0.6237,
+      "step": 14441
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.6970650666333884,
+      "learning_rate": 1.2694024016223952e-06,
+      "loss": 0.6706,
+      "step": 14442
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.742674173456901,
+      "learning_rate": 1.269309364271404e-06,
+      "loss": 0.6406,
+      "step": 14443
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.406069818907037,
+      "learning_rate": 1.2692163244069241e-06,
+      "loss": 0.6439,
+      "step": 14444
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.543322116355035,
+      "learning_rate": 1.2691232820298247e-06,
+      "loss": 0.5749,
+      "step": 14445
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.211133659820332,
+      "learning_rate": 1.269030237140974e-06,
+      "loss": 0.6816,
+      "step": 14446
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 5.269369256220564,
+      "learning_rate": 1.2689371897412402e-06,
+      "loss": 0.6549,
+      "step": 14447
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.565025406643107,
+      "learning_rate": 1.2688441398314917e-06,
+      "loss": 0.6654,
+      "step": 14448
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 4.100804352126534,
+      "learning_rate": 1.2687510874125973e-06,
+      "loss": 0.6191,
+      "step": 14449
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.5254833494352797,
+      "learning_rate": 1.2686580324854253e-06,
+      "loss": 0.6458,
+      "step": 14450
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.280874958637613,
+      "learning_rate": 1.2685649750508435e-06,
+      "loss": 0.6406,
+      "step": 14451
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.597100240295592,
+      "learning_rate": 1.2684719151097216e-06,
+      "loss": 0.6354,
+      "step": 14452
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.4290086838347986,
+      "learning_rate": 1.2683788526629275e-06,
+      "loss": 0.6217,
+      "step": 14453
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.568690761624818,
+      "learning_rate": 1.2682857877113298e-06,
+      "loss": 0.6471,
+      "step": 14454
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.041428782628014,
+      "learning_rate": 1.2681927202557972e-06,
+      "loss": 0.6217,
+      "step": 14455
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.2858229794898675,
+      "learning_rate": 1.2680996502971982e-06,
+      "loss": 0.6224,
+      "step": 14456
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.1141700056602963,
+      "learning_rate": 1.268006577836401e-06,
+      "loss": 0.6296,
+      "step": 14457
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.5204941302195083,
+      "learning_rate": 1.2679135028742752e-06,
+      "loss": 0.6432,
+      "step": 14458
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.601079799855409,
+      "learning_rate": 1.2678204254116892e-06,
+      "loss": 0.5866,
+      "step": 14459
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.709101582757441,
+      "learning_rate": 1.2677273454495112e-06,
+      "loss": 0.6634,
+      "step": 14460
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.859956969336383,
+      "learning_rate": 1.2676342629886102e-06,
+      "loss": 0.6914,
+      "step": 14461
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 7.06459072095258,
+      "learning_rate": 1.2675411780298548e-06,
+      "loss": 0.6204,
+      "step": 14462
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.6902061818020337,
+      "learning_rate": 1.2674480905741142e-06,
+      "loss": 0.6367,
+      "step": 14463
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.3431014545800637,
+      "learning_rate": 1.2673550006222566e-06,
+      "loss": 0.6133,
+      "step": 14464
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.5520782768679595,
+      "learning_rate": 1.2672619081751514e-06,
+      "loss": 0.5898,
+      "step": 14465
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.9816436962633235,
+      "learning_rate": 1.2671688132336671e-06,
+      "loss": 0.6348,
+      "step": 14466
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.581020798668728,
+      "learning_rate": 1.2670757157986725e-06,
+      "loss": 0.6139,
+      "step": 14467
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.216797118385344,
+      "learning_rate": 1.2669826158710368e-06,
+      "loss": 0.6139,
+      "step": 14468
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.7908854441613373,
+      "learning_rate": 1.2668895134516287e-06,
+      "loss": 0.6647,
+      "step": 14469
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.731089144510179,
+      "learning_rate": 1.266796408541317e-06,
+      "loss": 0.6198,
+      "step": 14470
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.7325316666295563,
+      "learning_rate": 1.2667033011409712e-06,
+      "loss": 0.612,
+      "step": 14471
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.060595743391856,
+      "learning_rate": 1.2666101912514596e-06,
+      "loss": 0.6647,
+      "step": 14472
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.8604561183000845,
+      "learning_rate": 1.2665170788736514e-06,
+      "loss": 0.6484,
+      "step": 14473
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.864300900504003,
+      "learning_rate": 1.266423964008416e-06,
+      "loss": 0.6348,
+      "step": 14474
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.719892112545786,
+      "learning_rate": 1.266330846656622e-06,
+      "loss": 0.7077,
+      "step": 14475
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.1684530794123753,
+      "learning_rate": 1.2662377268191392e-06,
+      "loss": 0.6146,
+      "step": 14476
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.8116449940842534,
+      "learning_rate": 1.2661446044968353e-06,
+      "loss": 0.6471,
+      "step": 14477
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.259382118786745,
+      "learning_rate": 1.2660514796905806e-06,
+      "loss": 0.6445,
+      "step": 14478
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.9010711192691923,
+      "learning_rate": 1.2659583524012443e-06,
+      "loss": 0.64,
+      "step": 14479
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.258239978402526,
+      "learning_rate": 1.265865222629695e-06,
+      "loss": 0.6986,
+      "step": 14480
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.003017973722865,
+      "learning_rate": 1.2657720903768018e-06,
+      "loss": 0.6576,
+      "step": 14481
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.016927435053214,
+      "learning_rate": 1.2656789556434344e-06,
+      "loss": 0.6159,
+      "step": 14482
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.988906147768949,
+      "learning_rate": 1.2655858184304615e-06,
+      "loss": 0.6471,
+      "step": 14483
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.7560144425257036,
+      "learning_rate": 1.265492678738753e-06,
+      "loss": 0.6458,
+      "step": 14484
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.767478158866083,
+      "learning_rate": 1.2653995365691773e-06,
+      "loss": 0.7005,
+      "step": 14485
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.3421273639230007,
+      "learning_rate": 1.265306391922605e-06,
+      "loss": 0.6686,
+      "step": 14486
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.825319510339591,
+      "learning_rate": 1.2652132447999044e-06,
+      "loss": 0.6439,
+      "step": 14487
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.879262210178548,
+      "learning_rate": 1.265120095201945e-06,
+      "loss": 0.6562,
+      "step": 14488
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.221670088735159,
+      "learning_rate": 1.2650269431295966e-06,
+      "loss": 0.651,
+      "step": 14489
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.1578355850904423,
+      "learning_rate": 1.2649337885837282e-06,
+      "loss": 0.696,
+      "step": 14490
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.016580826400459,
+      "learning_rate": 1.2648406315652088e-06,
+      "loss": 0.6458,
+      "step": 14491
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.8669520683008813,
+      "learning_rate": 1.2647474720749092e-06,
+      "loss": 0.6562,
+      "step": 14492
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.287115737806639,
+      "learning_rate": 1.2646543101136974e-06,
+      "loss": 0.6458,
+      "step": 14493
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.1810277826121123,
+      "learning_rate": 1.264561145682444e-06,
+      "loss": 0.6484,
+      "step": 14494
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.1816191457572813,
+      "learning_rate": 1.2644679787820177e-06,
+      "loss": 0.6335,
+      "step": 14495
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.0109490560717376,
+      "learning_rate": 1.2643748094132885e-06,
+      "loss": 0.651,
+      "step": 14496
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.277472537921852,
+      "learning_rate": 1.2642816375771258e-06,
+      "loss": 0.6569,
+      "step": 14497
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.120056008189543,
+      "learning_rate": 1.2641884632743992e-06,
+      "loss": 0.6374,
+      "step": 14498
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.919581930650713,
+      "learning_rate": 1.2640952865059783e-06,
+      "loss": 0.6647,
+      "step": 14499
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.2263323786125544,
+      "learning_rate": 1.2640021072727328e-06,
+      "loss": 0.6204,
+      "step": 14500
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.284057434751612,
+      "learning_rate": 1.263908925575532e-06,
+      "loss": 0.6296,
+      "step": 14501
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.478605838019839,
+      "learning_rate": 1.2638157414152462e-06,
+      "loss": 0.6764,
+      "step": 14502
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.6398134916668994,
+      "learning_rate": 1.2637225547927446e-06,
+      "loss": 0.7031,
+      "step": 14503
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.176241861548069,
+      "learning_rate": 1.2636293657088969e-06,
+      "loss": 0.6491,
+      "step": 14504
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.356650420015605,
+      "learning_rate": 1.2635361741645732e-06,
+      "loss": 0.668,
+      "step": 14505
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.9587402116109762,
+      "learning_rate": 1.2634429801606433e-06,
+      "loss": 0.6465,
+      "step": 14506
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.6957132981359404,
+      "learning_rate": 1.2633497836979762e-06,
+      "loss": 0.6426,
+      "step": 14507
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.142334001384778,
+      "learning_rate": 1.2632565847774427e-06,
+      "loss": 0.666,
+      "step": 14508
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 3.2215271822612195,
+      "learning_rate": 1.2631633833999123e-06,
+      "loss": 0.6439,
+      "step": 14509
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.177123675263066,
+      "learning_rate": 1.2630701795662546e-06,
+      "loss": 0.6543,
+      "step": 14510
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.011321859478851,
+      "learning_rate": 1.2629769732773396e-06,
+      "loss": 0.6497,
+      "step": 14511
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 5.353041814346228,
+      "learning_rate": 1.2628837645340373e-06,
+      "loss": 0.6458,
+      "step": 14512
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 4.338584747943553,
+      "learning_rate": 1.262790553337218e-06,
+      "loss": 0.6589,
+      "step": 14513
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.308227353754241,
+      "learning_rate": 1.2626973396877505e-06,
+      "loss": 0.6719,
+      "step": 14514
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.273624704337359,
+      "learning_rate": 1.262604123586506e-06,
+      "loss": 0.6387,
+      "step": 14515
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.6894128738156433,
+      "learning_rate": 1.2625109050343538e-06,
+      "loss": 0.679,
+      "step": 14516
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.4721233907982905,
+      "learning_rate": 1.262417684032164e-06,
+      "loss": 0.6471,
+      "step": 14517
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.112770796741894,
+      "learning_rate": 1.2623244605808069e-06,
+      "loss": 0.6452,
+      "step": 14518
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.4730432613846065,
+      "learning_rate": 1.2622312346811525e-06,
+      "loss": 0.6341,
+      "step": 14519
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.014010869772636,
+      "learning_rate": 1.2621380063340706e-06,
+      "loss": 0.623,
+      "step": 14520
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.7473389318574934,
+      "learning_rate": 1.2620447755404316e-06,
+      "loss": 0.6497,
+      "step": 14521
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.344040158254235,
+      "learning_rate": 1.2619515423011055e-06,
+      "loss": 0.6406,
+      "step": 14522
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.2272084480232235,
+      "learning_rate": 1.2618583066169624e-06,
+      "loss": 0.6667,
+      "step": 14523
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7797863329992984,
+      "learning_rate": 1.2617650684888727e-06,
+      "loss": 0.6191,
+      "step": 14524
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.515614713036336,
+      "learning_rate": 1.2616718279177065e-06,
+      "loss": 0.6549,
+      "step": 14525
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.042485813110608,
+      "learning_rate": 1.2615785849043338e-06,
+      "loss": 0.6419,
+      "step": 14526
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.0231458132105,
+      "learning_rate": 1.2614853394496252e-06,
+      "loss": 0.6569,
+      "step": 14527
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.5157461206543923,
+      "learning_rate": 1.2613920915544507e-06,
+      "loss": 0.6348,
+      "step": 14528
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.132689407819257,
+      "learning_rate": 1.261298841219681e-06,
+      "loss": 0.6536,
+      "step": 14529
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.5693073545419587,
+      "learning_rate": 1.2612055884461857e-06,
+      "loss": 0.6706,
+      "step": 14530
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.203635362786755,
+      "learning_rate": 1.261112333234836e-06,
+      "loss": 0.5983,
+      "step": 14531
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.048477988186719,
+      "learning_rate": 1.2610190755865014e-06,
+      "loss": 0.6647,
+      "step": 14532
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.076702619007778,
+      "learning_rate": 1.2609258155020525e-06,
+      "loss": 0.6543,
+      "step": 14533
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.6323020329490063,
+      "learning_rate": 1.2608325529823604e-06,
+      "loss": 0.6478,
+      "step": 14534
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.6078735406023945,
+      "learning_rate": 1.2607392880282948e-06,
+      "loss": 0.61,
+      "step": 14535
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.518486780900882,
+      "learning_rate": 1.2606460206407263e-06,
+      "loss": 0.6322,
+      "step": 14536
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.833706273283816,
+      "learning_rate": 1.2605527508205257e-06,
+      "loss": 0.6823,
+      "step": 14537
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.6793307501800325,
+      "learning_rate": 1.260459478568563e-06,
+      "loss": 0.6146,
+      "step": 14538
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7222719319706234,
+      "learning_rate": 1.2603662038857088e-06,
+      "loss": 0.6335,
+      "step": 14539
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7545775865957487,
+      "learning_rate": 1.260272926772834e-06,
+      "loss": 0.6341,
+      "step": 14540
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.8254422099017997,
+      "learning_rate": 1.2601796472308088e-06,
+      "loss": 0.6719,
+      "step": 14541
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.8541435605704666,
+      "learning_rate": 1.2600863652605043e-06,
+      "loss": 0.6322,
+      "step": 14542
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.29314151082272,
+      "learning_rate": 1.2599930808627906e-06,
+      "loss": 0.6243,
+      "step": 14543
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.868447585531573,
+      "learning_rate": 1.2598997940385384e-06,
+      "loss": 0.6074,
+      "step": 14544
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.8981291681427126,
+      "learning_rate": 1.2598065047886187e-06,
+      "loss": 0.6348,
+      "step": 14545
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.269635996557185,
+      "learning_rate": 1.2597132131139013e-06,
+      "loss": 0.6543,
+      "step": 14546
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7142968082098373,
+      "learning_rate": 1.259619919015258e-06,
+      "loss": 0.6309,
+      "step": 14547
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.759400511940406,
+      "learning_rate": 1.259526622493559e-06,
+      "loss": 0.6139,
+      "step": 14548
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.503872018852784,
+      "learning_rate": 1.2594333235496749e-06,
+      "loss": 0.6224,
+      "step": 14549
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.5712471672228663,
+      "learning_rate": 1.2593400221844769e-06,
+      "loss": 0.6764,
+      "step": 14550
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.0182980354770077,
+      "learning_rate": 1.2592467183988352e-06,
+      "loss": 0.6628,
+      "step": 14551
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.786350096329054,
+      "learning_rate": 1.259153412193621e-06,
+      "loss": 0.6615,
+      "step": 14552
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.9798793062278355,
+      "learning_rate": 1.2590601035697054e-06,
+      "loss": 0.6641,
+      "step": 14553
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.630609588919435,
+      "learning_rate": 1.2589667925279588e-06,
+      "loss": 0.6188,
+      "step": 14554
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.308108597640222,
+      "learning_rate": 1.2588734790692521e-06,
+      "loss": 0.6094,
+      "step": 14555
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 9.639249588744361,
+      "learning_rate": 1.2587801631944566e-06,
+      "loss": 0.6699,
+      "step": 14556
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.964704889616016,
+      "learning_rate": 1.2586868449044426e-06,
+      "loss": 0.6686,
+      "step": 14557
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.5794570891025215,
+      "learning_rate": 1.2585935242000818e-06,
+      "loss": 0.6432,
+      "step": 14558
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.338536335358407,
+      "learning_rate": 1.2585002010822442e-06,
+      "loss": 0.6354,
+      "step": 14559
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.0971322240679116,
+      "learning_rate": 1.2584068755518019e-06,
+      "loss": 0.6322,
+      "step": 14560
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.3326740222087063,
+      "learning_rate": 1.258313547609625e-06,
+      "loss": 0.623,
+      "step": 14561
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.845494048309921,
+      "learning_rate": 1.2582202172565852e-06,
+      "loss": 0.6419,
+      "step": 14562
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.520682117302017,
+      "learning_rate": 1.2581268844935534e-06,
+      "loss": 0.6315,
+      "step": 14563
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7841404492849215,
+      "learning_rate": 1.2580335493214e-06,
+      "loss": 0.6452,
+      "step": 14564
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.3936689346323496,
+      "learning_rate": 1.2579402117409973e-06,
+      "loss": 0.597,
+      "step": 14565
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.794125518775692,
+      "learning_rate": 1.2578468717532153e-06,
+      "loss": 0.6458,
+      "step": 14566
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.135329857233327,
+      "learning_rate": 1.257753529358926e-06,
+      "loss": 0.6582,
+      "step": 14567
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.197817248713362,
+      "learning_rate": 1.257660184559e-06,
+      "loss": 0.6751,
+      "step": 14568
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7773563444805474,
+      "learning_rate": 1.2575668373543087e-06,
+      "loss": 0.6413,
+      "step": 14569
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 5.319798076191076,
+      "learning_rate": 1.2574734877457235e-06,
+      "loss": 0.6484,
+      "step": 14570
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.8569268724685863,
+      "learning_rate": 1.2573801357341153e-06,
+      "loss": 0.6068,
+      "step": 14571
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 6.469928324341321,
+      "learning_rate": 1.2572867813203555e-06,
+      "loss": 0.6706,
+      "step": 14572
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7514497174934034,
+      "learning_rate": 1.2571934245053156e-06,
+      "loss": 0.6803,
+      "step": 14573
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.8091171424595935,
+      "learning_rate": 1.2571000652898667e-06,
+      "loss": 0.6419,
+      "step": 14574
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.9591351134065214,
+      "learning_rate": 1.2570067036748801e-06,
+      "loss": 0.6491,
+      "step": 14575
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.8523186679441794,
+      "learning_rate": 1.256913339661227e-06,
+      "loss": 0.6413,
+      "step": 14576
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.4223269863097134,
+      "learning_rate": 1.2568199732497793e-06,
+      "loss": 0.6549,
+      "step": 14577
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 4.023907464062921,
+      "learning_rate": 1.256726604441408e-06,
+      "loss": 0.6693,
+      "step": 14578
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.649868445734953,
+      "learning_rate": 1.2566332332369847e-06,
+      "loss": 0.6432,
+      "step": 14579
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.7545168923880694,
+      "learning_rate": 1.2565398596373805e-06,
+      "loss": 0.6361,
+      "step": 14580
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.911444894034895,
+      "learning_rate": 1.256446483643467e-06,
+      "loss": 0.6263,
+      "step": 14581
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.371808384242374,
+      "learning_rate": 1.256353105256116e-06,
+      "loss": 0.6126,
+      "step": 14582
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 3.9264271337501553,
+      "learning_rate": 1.2562597244761987e-06,
+      "loss": 0.6497,
+      "step": 14583
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.5643461589246557,
+      "learning_rate": 1.2561663413045868e-06,
+      "loss": 0.651,
+      "step": 14584
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.214547147731528,
+      "learning_rate": 1.2560729557421518e-06,
+      "loss": 0.6712,
+      "step": 14585
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.4923566164699773,
+      "learning_rate": 1.2559795677897653e-06,
+      "loss": 0.6296,
+      "step": 14586
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.2426427562513442,
+      "learning_rate": 1.255886177448299e-06,
+      "loss": 0.597,
+      "step": 14587
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.856288166849874,
+      "learning_rate": 1.255792784718624e-06,
+      "loss": 0.6673,
+      "step": 14588
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.9386484459066704,
+      "learning_rate": 1.2556993896016125e-06,
+      "loss": 0.6979,
+      "step": 14589
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.2470756283027034,
+      "learning_rate": 1.2556059920981362e-06,
+      "loss": 0.6706,
+      "step": 14590
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.8644965084792324,
+      "learning_rate": 1.2555125922090662e-06,
+      "loss": 0.6367,
+      "step": 14591
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.5911271255615134,
+      "learning_rate": 1.2554191899352749e-06,
+      "loss": 0.6966,
+      "step": 14592
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.3363668196658147,
+      "learning_rate": 1.2553257852776334e-06,
+      "loss": 0.6348,
+      "step": 14593
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.431099418293222,
+      "learning_rate": 1.2552323782370136e-06,
+      "loss": 0.6419,
+      "step": 14594
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.299832268004123,
+      "learning_rate": 1.255138968814288e-06,
+      "loss": 0.696,
+      "step": 14595
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.3353919694392165,
+      "learning_rate": 1.2550455570103275e-06,
+      "loss": 0.6628,
+      "step": 14596
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.3553417597803357,
+      "learning_rate": 1.2549521428260046e-06,
+      "loss": 0.638,
+      "step": 14597
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.1853300194436414,
+      "learning_rate": 1.2548587262621906e-06,
+      "loss": 0.6536,
+      "step": 14598
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.336300132271448,
+      "learning_rate": 1.2547653073197573e-06,
+      "loss": 0.6296,
+      "step": 14599
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.880594938325326,
+      "learning_rate": 1.2546718859995772e-06,
+      "loss": 0.638,
+      "step": 14600
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.41822312758735,
+      "learning_rate": 1.2545784623025217e-06,
+      "loss": 0.6328,
+      "step": 14601
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.799964168484688,
+      "learning_rate": 1.2544850362294629e-06,
+      "loss": 0.6315,
+      "step": 14602
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.7077027258564734,
+      "learning_rate": 1.2543916077812729e-06,
+      "loss": 0.6335,
+      "step": 14603
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.843176456450894,
+      "learning_rate": 1.2542981769588232e-06,
+      "loss": 0.6211,
+      "step": 14604
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.471131521938686,
+      "learning_rate": 1.2542047437629863e-06,
+      "loss": 0.6758,
+      "step": 14605
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.4457284396174366,
+      "learning_rate": 1.2541113081946343e-06,
+      "loss": 0.627,
+      "step": 14606
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.3165601666790057,
+      "learning_rate": 1.2540178702546386e-06,
+      "loss": 0.6074,
+      "step": 14607
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.0723660844167715,
+      "learning_rate": 1.2539244299438723e-06,
+      "loss": 0.61,
+      "step": 14608
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.1837114403688336,
+      "learning_rate": 1.253830987263206e-06,
+      "loss": 0.6289,
+      "step": 14609
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.5349721048863096,
+      "learning_rate": 1.253737542213513e-06,
+      "loss": 0.6478,
+      "step": 14610
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.692753150916202,
+      "learning_rate": 1.2536440947956655e-06,
+      "loss": 0.6315,
+      "step": 14611
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.951354894136671,
+      "learning_rate": 1.2535506450105347e-06,
+      "loss": 0.6556,
+      "step": 14612
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.165333250186095,
+      "learning_rate": 1.2534571928589934e-06,
+      "loss": 0.651,
+      "step": 14613
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.7917696941613643,
+      "learning_rate": 1.2533637383419139e-06,
+      "loss": 0.6517,
+      "step": 14614
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.4291150256909386,
+      "learning_rate": 1.2532702814601678e-06,
+      "loss": 0.6556,
+      "step": 14615
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.352227834801144,
+      "learning_rate": 1.253176822214628e-06,
+      "loss": 0.6198,
+      "step": 14616
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.528940403400527,
+      "learning_rate": 1.2530833606061665e-06,
+      "loss": 0.6908,
+      "step": 14617
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.5074853009977005,
+      "learning_rate": 1.2529898966356558e-06,
+      "loss": 0.6602,
+      "step": 14618
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.3951608355712635,
+      "learning_rate": 1.252896430303968e-06,
+      "loss": 0.6556,
+      "step": 14619
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.388416381509622,
+      "learning_rate": 1.2528029616119753e-06,
+      "loss": 0.6087,
+      "step": 14620
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.169312903696001,
+      "learning_rate": 1.25270949056055e-06,
+      "loss": 0.6536,
+      "step": 14621
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.6056455556474476,
+      "learning_rate": 1.2526160171505652e-06,
+      "loss": 0.6576,
+      "step": 14622
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.7127279150765924,
+      "learning_rate": 1.2525225413828924e-06,
+      "loss": 0.6263,
+      "step": 14623
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.7436536521580352,
+      "learning_rate": 1.252429063258405e-06,
+      "loss": 0.6589,
+      "step": 14624
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 6.689216017235354,
+      "learning_rate": 1.252335582777974e-06,
+      "loss": 0.6556,
+      "step": 14625
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.229573516969056,
+      "learning_rate": 1.2522420999424728e-06,
+      "loss": 0.5905,
+      "step": 14626
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.503612380415172,
+      "learning_rate": 1.252148614752774e-06,
+      "loss": 0.6185,
+      "step": 14627
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.302786611302436,
+      "learning_rate": 1.2520551272097501e-06,
+      "loss": 0.6771,
+      "step": 14628
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.7587563480510204,
+      "learning_rate": 1.2519616373142733e-06,
+      "loss": 0.6751,
+      "step": 14629
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.136179110277189,
+      "learning_rate": 1.2518681450672159e-06,
+      "loss": 0.6315,
+      "step": 14630
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.9278830816557075,
+      "learning_rate": 1.251774650469451e-06,
+      "loss": 0.6719,
+      "step": 14631
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.937453408590218,
+      "learning_rate": 1.2516811535218512e-06,
+      "loss": 0.6569,
+      "step": 14632
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.759597490278777,
+      "learning_rate": 1.2515876542252889e-06,
+      "loss": 0.627,
+      "step": 14633
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.760331692006826,
+      "learning_rate": 1.2514941525806366e-06,
+      "loss": 0.6901,
+      "step": 14634
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.2546604141825703,
+      "learning_rate": 1.2514006485887673e-06,
+      "loss": 0.6348,
+      "step": 14635
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.451880368453681,
+      "learning_rate": 1.2513071422505533e-06,
+      "loss": 0.64,
+      "step": 14636
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.5930482161919532,
+      "learning_rate": 1.2512136335668677e-06,
+      "loss": 0.6504,
+      "step": 14637
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.4940940024511327,
+      "learning_rate": 1.251120122538583e-06,
+      "loss": 0.6621,
+      "step": 14638
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.5297059910459776,
+      "learning_rate": 1.251026609166572e-06,
+      "loss": 0.6816,
+      "step": 14639
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.11210716471469,
+      "learning_rate": 1.2509330934517075e-06,
+      "loss": 0.6296,
+      "step": 14640
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.184086864698617,
+      "learning_rate": 1.2508395753948622e-06,
+      "loss": 0.6738,
+      "step": 14641
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.221171097438364,
+      "learning_rate": 1.2507460549969086e-06,
+      "loss": 0.6589,
+      "step": 14642
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.2854126569059936,
+      "learning_rate": 1.2506525322587204e-06,
+      "loss": 0.6302,
+      "step": 14643
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 5.112814690562303,
+      "learning_rate": 1.25055900718117e-06,
+      "loss": 0.6719,
+      "step": 14644
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 4.587917208557352,
+      "learning_rate": 1.25046547976513e-06,
+      "loss": 0.6634,
+      "step": 14645
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.2226896628729786,
+      "learning_rate": 1.2503719500114733e-06,
+      "loss": 0.6107,
+      "step": 14646
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.2572762076658996,
+      "learning_rate": 1.2502784179210734e-06,
+      "loss": 0.6393,
+      "step": 14647
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.350974690717858,
+      "learning_rate": 1.250184883494803e-06,
+      "loss": 0.6582,
+      "step": 14648
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.232343327159912,
+      "learning_rate": 1.2500913467335346e-06,
+      "loss": 0.6217,
+      "step": 14649
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.9395367706616784,
+      "learning_rate": 1.2499978076381416e-06,
+      "loss": 0.6348,
+      "step": 14650
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.564379565911669,
+      "learning_rate": 1.2499042662094972e-06,
+      "loss": 0.6335,
+      "step": 14651
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.8167435505991785,
+      "learning_rate": 1.2498107224484739e-06,
+      "loss": 0.6621,
+      "step": 14652
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.7776599445138364,
+      "learning_rate": 1.2497171763559454e-06,
+      "loss": 0.6296,
+      "step": 14653
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.19837682001825,
+      "learning_rate": 1.2496236279327841e-06,
+      "loss": 0.6107,
+      "step": 14654
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.107777715599288,
+      "learning_rate": 1.2495300771798634e-06,
+      "loss": 0.6549,
+      "step": 14655
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.1503420432138523,
+      "learning_rate": 1.2494365240980565e-06,
+      "loss": 0.6523,
+      "step": 14656
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.7356279536314925,
+      "learning_rate": 1.2493429686882366e-06,
+      "loss": 0.627,
+      "step": 14657
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.6752695208027464,
+      "learning_rate": 1.2492494109512766e-06,
+      "loss": 0.7246,
+      "step": 14658
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.359622986391555,
+      "learning_rate": 1.2491558508880494e-06,
+      "loss": 0.6139,
+      "step": 14659
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.5634969375508208,
+      "learning_rate": 1.249062288499429e-06,
+      "loss": 0.6706,
+      "step": 14660
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.202708167612195,
+      "learning_rate": 1.2489687237862888e-06,
+      "loss": 0.6178,
+      "step": 14661
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.9225731863131816,
+      "learning_rate": 1.2488751567495006e-06,
+      "loss": 0.6419,
+      "step": 14662
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.111798023384258,
+      "learning_rate": 1.248781587389939e-06,
+      "loss": 0.6458,
+      "step": 14663
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.682422899978719,
+      "learning_rate": 1.2486880157084766e-06,
+      "loss": 0.6471,
+      "step": 14664
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.7331953107953564,
+      "learning_rate": 1.248594441705987e-06,
+      "loss": 0.6413,
+      "step": 14665
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.611901823531557,
+      "learning_rate": 1.2485008653833435e-06,
+      "loss": 0.6543,
+      "step": 14666
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.21010753134833,
+      "learning_rate": 1.2484072867414197e-06,
+      "loss": 0.6621,
+      "step": 14667
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.312758570666702,
+      "learning_rate": 1.2483137057810885e-06,
+      "loss": 0.651,
+      "step": 14668
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.42522761890329,
+      "learning_rate": 1.2482201225032234e-06,
+      "loss": 0.638,
+      "step": 14669
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.504476195483101,
+      "learning_rate": 1.2481265369086979e-06,
+      "loss": 0.6484,
+      "step": 14670
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.4655470242208946,
+      "learning_rate": 1.248032948998386e-06,
+      "loss": 0.6595,
+      "step": 14671
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.8734851270958095,
+      "learning_rate": 1.2479393587731604e-06,
+      "loss": 0.6009,
+      "step": 14672
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.7139608098169594,
+      "learning_rate": 1.2478457662338945e-06,
+      "loss": 0.6751,
+      "step": 14673
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.4269437901248905,
+      "learning_rate": 1.2477521713814627e-06,
+      "loss": 0.6322,
+      "step": 14674
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.8722685893191215,
+      "learning_rate": 1.2476585742167375e-06,
+      "loss": 0.6361,
+      "step": 14675
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.4324407314069107,
+      "learning_rate": 1.2475649747405932e-06,
+      "loss": 0.6569,
+      "step": 14676
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.066282701007285,
+      "learning_rate": 1.2474713729539033e-06,
+      "loss": 0.6178,
+      "step": 14677
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.998330758483331,
+      "learning_rate": 1.2473777688575405e-06,
+      "loss": 0.6504,
+      "step": 14678
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.8541500152241435,
+      "learning_rate": 1.2472841624523798e-06,
+      "loss": 0.6393,
+      "step": 14679
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.0913355504190227,
+      "learning_rate": 1.2471905537392938e-06,
+      "loss": 0.6478,
+      "step": 14680
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.3814173242758225,
+      "learning_rate": 1.2470969427191564e-06,
+      "loss": 0.6296,
+      "step": 14681
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.902816549233596,
+      "learning_rate": 1.2470033293928418e-06,
+      "loss": 0.6367,
+      "step": 14682
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.803616533084308,
+      "learning_rate": 1.2469097137612228e-06,
+      "loss": 0.6204,
+      "step": 14683
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 5.379392815147804,
+      "learning_rate": 1.246816095825174e-06,
+      "loss": 0.6803,
+      "step": 14684
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.1041559695454435,
+      "learning_rate": 1.2467224755855684e-06,
+      "loss": 0.6426,
+      "step": 14685
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.816739285757766,
+      "learning_rate": 1.2466288530432802e-06,
+      "loss": 0.6608,
+      "step": 14686
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.4535542451272145,
+      "learning_rate": 1.2465352281991835e-06,
+      "loss": 0.6217,
+      "step": 14687
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.849249757859827,
+      "learning_rate": 1.2464416010541513e-06,
+      "loss": 0.6536,
+      "step": 14688
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.520691598253275,
+      "learning_rate": 1.2463479716090579e-06,
+      "loss": 0.6178,
+      "step": 14689
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.321419839887923,
+      "learning_rate": 1.2462543398647773e-06,
+      "loss": 0.6803,
+      "step": 14690
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.3988052993939655,
+      "learning_rate": 1.246160705822183e-06,
+      "loss": 0.6686,
+      "step": 14691
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.7877558379746143,
+      "learning_rate": 1.2460670694821488e-06,
+      "loss": 0.6647,
+      "step": 14692
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.2743605971954066,
+      "learning_rate": 1.2459734308455494e-06,
+      "loss": 0.6361,
+      "step": 14693
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.633594533935641,
+      "learning_rate": 1.245879789913258e-06,
+      "loss": 0.6237,
+      "step": 14694
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.736365215424283,
+      "learning_rate": 1.2457861466861487e-06,
+      "loss": 0.6426,
+      "step": 14695
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.4290630649729503,
+      "learning_rate": 1.2456925011650956e-06,
+      "loss": 0.6139,
+      "step": 14696
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.0708634298294792,
+      "learning_rate": 1.2455988533509727e-06,
+      "loss": 0.6432,
+      "step": 14697
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.6313574720246846,
+      "learning_rate": 1.245505203244654e-06,
+      "loss": 0.6146,
+      "step": 14698
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.8367838900048667,
+      "learning_rate": 1.2454115508470134e-06,
+      "loss": 0.6224,
+      "step": 14699
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.554370736361144,
+      "learning_rate": 1.2453178961589253e-06,
+      "loss": 0.6426,
+      "step": 14700
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.457866552882189,
+      "learning_rate": 1.2452242391812635e-06,
+      "loss": 0.6439,
+      "step": 14701
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.3365222074714382,
+      "learning_rate": 1.245130579914902e-06,
+      "loss": 0.6478,
+      "step": 14702
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.9262336354177596,
+      "learning_rate": 1.2450369183607151e-06,
+      "loss": 0.6523,
+      "step": 14703
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.8864998287619086,
+      "learning_rate": 1.2449432545195776e-06,
+      "loss": 0.6504,
+      "step": 14704
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.4217379021163166,
+      "learning_rate": 1.2448495883923622e-06,
+      "loss": 0.6449,
+      "step": 14705
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.7652360895415726,
+      "learning_rate": 1.2447559199799445e-06,
+      "loss": 0.6908,
+      "step": 14706
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.389021390756673,
+      "learning_rate": 1.244662249283198e-06,
+      "loss": 0.6413,
+      "step": 14707
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.794135913927578,
+      "learning_rate": 1.2445685763029969e-06,
+      "loss": 0.625,
+      "step": 14708
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.573859606662216,
+      "learning_rate": 1.2444749010402159e-06,
+      "loss": 0.6133,
+      "step": 14709
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.780435328625068,
+      "learning_rate": 1.2443812234957289e-06,
+      "loss": 0.6608,
+      "step": 14710
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.8849967158301344,
+      "learning_rate": 1.2442875436704106e-06,
+      "loss": 0.6139,
+      "step": 14711
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 4.4455759558085495,
+      "learning_rate": 1.2441938615651346e-06,
+      "loss": 0.6283,
+      "step": 14712
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.510391507132702,
+      "learning_rate": 1.2441001771807757e-06,
+      "loss": 0.6771,
+      "step": 14713
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.3365341892730354,
+      "learning_rate": 1.2440064905182087e-06,
+      "loss": 0.5885,
+      "step": 14714
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.458591662256535,
+      "learning_rate": 1.2439128015783073e-06,
+      "loss": 0.6621,
+      "step": 14715
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.8166402315542842,
+      "learning_rate": 1.2438191103619462e-06,
+      "loss": 0.6589,
+      "step": 14716
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.2878889102735798,
+      "learning_rate": 1.2437254168699997e-06,
+      "loss": 0.6315,
+      "step": 14717
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.9377541099462547,
+      "learning_rate": 1.2436317211033425e-06,
+      "loss": 0.6374,
+      "step": 14718
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.4894959958351017,
+      "learning_rate": 1.2435380230628488e-06,
+      "loss": 0.6628,
+      "step": 14719
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.8847467939333895,
+      "learning_rate": 1.2434443227493932e-06,
+      "loss": 0.6048,
+      "step": 14720
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.6043526188821375,
+      "learning_rate": 1.2433506201638499e-06,
+      "loss": 0.6361,
+      "step": 14721
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.6777449735935,
+      "learning_rate": 1.2432569153070942e-06,
+      "loss": 0.6426,
+      "step": 14722
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.709302886595163,
+      "learning_rate": 1.2431632081800002e-06,
+      "loss": 0.6012,
+      "step": 14723
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.452569445101003,
+      "learning_rate": 1.2430694987834423e-06,
+      "loss": 0.6589,
+      "step": 14724
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.487191092441426,
+      "learning_rate": 1.242975787118295e-06,
+      "loss": 0.6562,
+      "step": 14725
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.377635671123433,
+      "learning_rate": 1.2428820731854334e-06,
+      "loss": 0.6061,
+      "step": 14726
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.588884428865809,
+      "learning_rate": 1.2427883569857324e-06,
+      "loss": 0.666,
+      "step": 14727
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.8390366686781627,
+      "learning_rate": 1.2426946385200655e-06,
+      "loss": 0.6751,
+      "step": 14728
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.332145768957236,
+      "learning_rate": 1.2426009177893085e-06,
+      "loss": 0.6302,
+      "step": 14729
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.3588595165747144,
+      "learning_rate": 1.2425071947943354e-06,
+      "loss": 0.666,
+      "step": 14730
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.5884214924807982,
+      "learning_rate": 1.2424134695360213e-06,
+      "loss": 0.638,
+      "step": 14731
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.3232764389162304,
+      "learning_rate": 1.2423197420152409e-06,
+      "loss": 0.6042,
+      "step": 14732
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.743762453960156,
+      "learning_rate": 1.2422260122328691e-06,
+      "loss": 0.6009,
+      "step": 14733
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.1106063811786497,
+      "learning_rate": 1.2421322801897802e-06,
+      "loss": 0.6986,
+      "step": 14734
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.299120801009346,
+      "learning_rate": 1.2420385458868495e-06,
+      "loss": 0.6452,
+      "step": 14735
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.5863045082107727,
+      "learning_rate": 1.2419448093249514e-06,
+      "loss": 0.6504,
+      "step": 14736
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.072587279682176,
+      "learning_rate": 1.2418510705049611e-06,
+      "loss": 0.6562,
+      "step": 14737
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.604733425006412,
+      "learning_rate": 1.2417573294277538e-06,
+      "loss": 0.6367,
+      "step": 14738
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.885852368397556,
+      "learning_rate": 1.2416635860942034e-06,
+      "loss": 0.6914,
+      "step": 14739
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.159893497211562,
+      "learning_rate": 1.2415698405051857e-06,
+      "loss": 0.6296,
+      "step": 14740
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.013986188465755,
+      "learning_rate": 1.2414760926615753e-06,
+      "loss": 0.5918,
+      "step": 14741
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.2156240240649443,
+      "learning_rate": 1.241382342564247e-06,
+      "loss": 0.6803,
+      "step": 14742
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.472248953810297,
+      "learning_rate": 1.2412885902140761e-06,
+      "loss": 0.6003,
+      "step": 14743
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.103928229333382,
+      "learning_rate": 1.2411948356119373e-06,
+      "loss": 0.597,
+      "step": 14744
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.847831338147369,
+      "learning_rate": 1.241101078758706e-06,
+      "loss": 0.6543,
+      "step": 14745
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.5385649213528834,
+      "learning_rate": 1.2410073196552567e-06,
+      "loss": 0.6315,
+      "step": 14746
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.9784197805706163,
+      "learning_rate": 1.240913558302465e-06,
+      "loss": 0.6504,
+      "step": 14747
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.2580358602483765,
+      "learning_rate": 1.2408197947012057e-06,
+      "loss": 0.6634,
+      "step": 14748
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.956629512043636,
+      "learning_rate": 1.2407260288523538e-06,
+      "loss": 0.6712,
+      "step": 14749
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.689894718185501,
+      "learning_rate": 1.2406322607567847e-06,
+      "loss": 0.6263,
+      "step": 14750
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.0008793302385453,
+      "learning_rate": 1.2405384904153734e-06,
+      "loss": 0.6979,
+      "step": 14751
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.979517997308809,
+      "learning_rate": 1.240444717828995e-06,
+      "loss": 0.6602,
+      "step": 14752
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.740260573350636,
+      "learning_rate": 1.2403509429985254e-06,
+      "loss": 0.653,
+      "step": 14753
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.6439374793196917,
+      "learning_rate": 1.2402571659248386e-06,
+      "loss": 0.61,
+      "step": 14754
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.4531797578617613,
+      "learning_rate": 1.2401633866088104e-06,
+      "loss": 0.6361,
+      "step": 14755
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.787904358794559,
+      "learning_rate": 1.2400696050513164e-06,
+      "loss": 0.6198,
+      "step": 14756
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 4.974870520465567,
+      "learning_rate": 1.2399758212532314e-06,
+      "loss": 0.6953,
+      "step": 14757
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.109050023111121,
+      "learning_rate": 1.2398820352154307e-06,
+      "loss": 0.6569,
+      "step": 14758
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.6289292609840293,
+      "learning_rate": 1.23978824693879e-06,
+      "loss": 0.6341,
+      "step": 14759
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.451226012434306,
+      "learning_rate": 1.2396944564241843e-06,
+      "loss": 0.6432,
+      "step": 14760
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.092808442103815,
+      "learning_rate": 1.239600663672489e-06,
+      "loss": 0.6576,
+      "step": 14761
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.6135795195356546,
+      "learning_rate": 1.2395068686845793e-06,
+      "loss": 0.6367,
+      "step": 14762
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.662308994868519,
+      "learning_rate": 1.239413071461331e-06,
+      "loss": 0.6445,
+      "step": 14763
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.961363668558205,
+      "learning_rate": 1.2393192720036195e-06,
+      "loss": 0.6191,
+      "step": 14764
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.5838260744026327,
+      "learning_rate": 1.2392254703123199e-06,
+      "loss": 0.6165,
+      "step": 14765
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.882577160194331,
+      "learning_rate": 1.2391316663883079e-06,
+      "loss": 0.6628,
+      "step": 14766
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.976856289688177,
+      "learning_rate": 1.239037860232459e-06,
+      "loss": 0.6074,
+      "step": 14767
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.076721820937446,
+      "learning_rate": 1.2389440518456482e-06,
+      "loss": 0.6725,
+      "step": 14768
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.593635464313873,
+      "learning_rate": 1.2388502412287518e-06,
+      "loss": 0.6348,
+      "step": 14769
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.478413813947112,
+      "learning_rate": 1.238756428382645e-06,
+      "loss": 0.6745,
+      "step": 14770
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 5.461980124696165,
+      "learning_rate": 1.238662613308203e-06,
+      "loss": 0.6361,
+      "step": 14771
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.4719066405254524,
+      "learning_rate": 1.2385687960063021e-06,
+      "loss": 0.6745,
+      "step": 14772
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.0188974094749694,
+      "learning_rate": 1.2384749764778174e-06,
+      "loss": 0.6302,
+      "step": 14773
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.4706611818783886,
+      "learning_rate": 1.2383811547236244e-06,
+      "loss": 0.6393,
+      "step": 14774
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.68575083286521,
+      "learning_rate": 1.2382873307445994e-06,
+      "loss": 0.6081,
+      "step": 14775
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.0487996224705243,
+      "learning_rate": 1.2381935045416176e-06,
+      "loss": 0.6732,
+      "step": 14776
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.658606452413135,
+      "learning_rate": 1.2380996761155547e-06,
+      "loss": 0.6413,
+      "step": 14777
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.7820217284006232,
+      "learning_rate": 1.2380058454672862e-06,
+      "loss": 0.6458,
+      "step": 14778
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.594375981731638,
+      "learning_rate": 1.2379120125976883e-06,
+      "loss": 0.6289,
+      "step": 14779
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.3407505565120017,
+      "learning_rate": 1.2378181775076368e-06,
+      "loss": 0.6445,
+      "step": 14780
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 6.726943744047224,
+      "learning_rate": 1.2377243401980069e-06,
+      "loss": 0.6901,
+      "step": 14781
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.960526394337498,
+      "learning_rate": 1.2376305006696747e-06,
+      "loss": 0.6257,
+      "step": 14782
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.7339232434185767,
+      "learning_rate": 1.2375366589235164e-06,
+      "loss": 0.6224,
+      "step": 14783
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.413240897083263,
+      "learning_rate": 1.237442814960407e-06,
+      "loss": 0.6393,
+      "step": 14784
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.7207783689327183,
+      "learning_rate": 1.2373489687812231e-06,
+      "loss": 0.6152,
+      "step": 14785
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.3665250753538642,
+      "learning_rate": 1.2372551203868402e-06,
+      "loss": 0.6328,
+      "step": 14786
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.8568302509737236,
+      "learning_rate": 1.237161269778134e-06,
+      "loss": 0.6257,
+      "step": 14787
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.4463838250896246,
+      "learning_rate": 1.237067416955981e-06,
+      "loss": 0.6374,
+      "step": 14788
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.208359644764564,
+      "learning_rate": 1.2369735619212566e-06,
+      "loss": 0.6686,
+      "step": 14789
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 5.996356960015916,
+      "learning_rate": 1.2368797046748375e-06,
+      "loss": 0.6771,
+      "step": 14790
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.823979717800155,
+      "learning_rate": 1.2367858452175987e-06,
+      "loss": 0.627,
+      "step": 14791
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.4938357800831237,
+      "learning_rate": 1.2366919835504166e-06,
+      "loss": 0.6283,
+      "step": 14792
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 7.030114281023856,
+      "learning_rate": 1.2365981196741677e-06,
+      "loss": 0.6471,
+      "step": 14793
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.736929676538827,
+      "learning_rate": 1.2365042535897274e-06,
+      "loss": 0.6634,
+      "step": 14794
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.7351079879295814,
+      "learning_rate": 1.236410385297972e-06,
+      "loss": 0.638,
+      "step": 14795
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.900997154367461,
+      "learning_rate": 1.2363165147997775e-06,
+      "loss": 0.6536,
+      "step": 14796
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.31783540552272,
+      "learning_rate": 1.2362226420960202e-06,
+      "loss": 0.6393,
+      "step": 14797
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.2111053582699056,
+      "learning_rate": 1.236128767187576e-06,
+      "loss": 0.6289,
+      "step": 14798
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.582089960481317,
+      "learning_rate": 1.2360348900753212e-06,
+      "loss": 0.6576,
+      "step": 14799
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.5291896561414458,
+      "learning_rate": 1.2359410107601318e-06,
+      "loss": 0.64,
+      "step": 14800
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.4526284951376214,
+      "learning_rate": 1.2358471292428842e-06,
+      "loss": 0.5846,
+      "step": 14801
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.072752847299142,
+      "learning_rate": 1.2357532455244542e-06,
+      "loss": 0.6012,
+      "step": 14802
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.6557484217857406,
+      "learning_rate": 1.2356593596057185e-06,
+      "loss": 0.651,
+      "step": 14803
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.3271912755740196,
+      "learning_rate": 1.2355654714875535e-06,
+      "loss": 0.6198,
+      "step": 14804
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.6936992015699435,
+      "learning_rate": 1.2354715811708345e-06,
+      "loss": 0.6895,
+      "step": 14805
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.877768840597768,
+      "learning_rate": 1.2353776886564391e-06,
+      "loss": 0.6283,
+      "step": 14806
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.541046510028128,
+      "learning_rate": 1.2352837939452422e-06,
+      "loss": 0.6406,
+      "step": 14807
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.622660629356291,
+      "learning_rate": 1.235189897038121e-06,
+      "loss": 0.6641,
+      "step": 14808
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.841491098654051,
+      "learning_rate": 1.2350959979359524e-06,
+      "loss": 0.6654,
+      "step": 14809
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.693098633278268,
+      "learning_rate": 1.2350020966396113e-06,
+      "loss": 0.6615,
+      "step": 14810
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.823927934351698,
+      "learning_rate": 1.2349081931499752e-06,
+      "loss": 0.6068,
+      "step": 14811
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.562723309788467,
+      "learning_rate": 1.23481428746792e-06,
+      "loss": 0.6237,
+      "step": 14812
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.605832299036598,
+      "learning_rate": 1.234720379594322e-06,
+      "loss": 0.6322,
+      "step": 14813
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.5461699826522763,
+      "learning_rate": 1.2346264695300586e-06,
+      "loss": 0.6348,
+      "step": 14814
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.5813146651440033,
+      "learning_rate": 1.234532557276005e-06,
+      "loss": 0.6504,
+      "step": 14815
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.0826744084380855,
+      "learning_rate": 1.2344386428330387e-06,
+      "loss": 0.6458,
+      "step": 14816
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.7731174058009977,
+      "learning_rate": 1.2343447262020358e-06,
+      "loss": 0.582,
+      "step": 14817
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.7819992280576855,
+      "learning_rate": 1.2342508073838727e-06,
+      "loss": 0.651,
+      "step": 14818
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.013918809893278,
+      "learning_rate": 1.2341568863794261e-06,
+      "loss": 0.6204,
+      "step": 14819
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.738801467432079,
+      "learning_rate": 1.2340629631895727e-06,
+      "loss": 0.653,
+      "step": 14820
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.308449070283841,
+      "learning_rate": 1.2339690378151887e-06,
+      "loss": 0.6243,
+      "step": 14821
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.6624027273223665,
+      "learning_rate": 1.2338751102571512e-06,
+      "loss": 0.627,
+      "step": 14822
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.519826510986403,
+      "learning_rate": 1.2337811805163367e-06,
+      "loss": 0.5996,
+      "step": 14823
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.1054751957245403,
+      "learning_rate": 1.2336872485936213e-06,
+      "loss": 0.6283,
+      "step": 14824
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.753850603055007,
+      "learning_rate": 1.2335933144898826e-06,
+      "loss": 0.6439,
+      "step": 14825
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.164202757722794,
+      "learning_rate": 1.2334993782059967e-06,
+      "loss": 0.6966,
+      "step": 14826
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.981315660371182,
+      "learning_rate": 1.2334054397428408e-06,
+      "loss": 0.6374,
+      "step": 14827
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.2097403360572856,
+      "learning_rate": 1.2333114991012907e-06,
+      "loss": 0.6719,
+      "step": 14828
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.292577672159467,
+      "learning_rate": 1.233217556282224e-06,
+      "loss": 0.6484,
+      "step": 14829
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.6494334091228446,
+      "learning_rate": 1.2331236112865172e-06,
+      "loss": 0.5924,
+      "step": 14830
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.308286544939772,
+      "learning_rate": 1.2330296641150473e-06,
+      "loss": 0.6491,
+      "step": 14831
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.106058137482465,
+      "learning_rate": 1.2329357147686907e-06,
+      "loss": 0.5905,
+      "step": 14832
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.01613127263927,
+      "learning_rate": 1.2328417632483247e-06,
+      "loss": 0.6406,
+      "step": 14833
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.8023610577514018,
+      "learning_rate": 1.2327478095548256e-06,
+      "loss": 0.6419,
+      "step": 14834
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.144593604731493,
+      "learning_rate": 1.232653853689071e-06,
+      "loss": 0.6764,
+      "step": 14835
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.5922845387629248,
+      "learning_rate": 1.2325598956519373e-06,
+      "loss": 0.6152,
+      "step": 14836
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.9750722736724797,
+      "learning_rate": 1.2324659354443012e-06,
+      "loss": 0.6693,
+      "step": 14837
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.9916342504003155,
+      "learning_rate": 1.2323719730670402e-06,
+      "loss": 0.6504,
+      "step": 14838
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.6166650345255507,
+      "learning_rate": 1.2322780085210312e-06,
+      "loss": 0.6198,
+      "step": 14839
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.7277831528119125,
+      "learning_rate": 1.2321840418071507e-06,
+      "loss": 0.6803,
+      "step": 14840
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.0697596794634454,
+      "learning_rate": 1.2320900729262763e-06,
+      "loss": 0.64,
+      "step": 14841
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.265729850087329,
+      "learning_rate": 1.2319961018792843e-06,
+      "loss": 0.6478,
+      "step": 14842
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.7430929320664013,
+      "learning_rate": 1.2319021286670525e-06,
+      "loss": 0.6582,
+      "step": 14843
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.9651861325575575,
+      "learning_rate": 1.2318081532904576e-06,
+      "loss": 0.6751,
+      "step": 14844
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.430620740314163,
+      "learning_rate": 1.2317141757503764e-06,
+      "loss": 0.6439,
+      "step": 14845
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.3819521406122304,
+      "learning_rate": 1.2316201960476869e-06,
+      "loss": 0.666,
+      "step": 14846
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.090212104553938,
+      "learning_rate": 1.231526214183265e-06,
+      "loss": 0.6921,
+      "step": 14847
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.599457884910559,
+      "learning_rate": 1.231432230157989e-06,
+      "loss": 0.638,
+      "step": 14848
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.3530235877272507,
+      "learning_rate": 1.2313382439727353e-06,
+      "loss": 0.6602,
+      "step": 14849
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.4412711006640615,
+      "learning_rate": 1.2312442556283812e-06,
+      "loss": 0.6439,
+      "step": 14850
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 4.312765496758351,
+      "learning_rate": 1.2311502651258042e-06,
+      "loss": 0.6243,
+      "step": 14851
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.515934661630445,
+      "learning_rate": 1.2310562724658812e-06,
+      "loss": 0.6406,
+      "step": 14852
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.3233082939797147,
+      "learning_rate": 1.2309622776494894e-06,
+      "loss": 0.6204,
+      "step": 14853
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.456097350573825,
+      "learning_rate": 1.2308682806775066e-06,
+      "loss": 0.6608,
+      "step": 14854
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.274975439594404,
+      "learning_rate": 1.2307742815508096e-06,
+      "loss": 0.6641,
+      "step": 14855
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.3031433274100146,
+      "learning_rate": 1.2306802802702761e-06,
+      "loss": 0.6764,
+      "step": 14856
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.088899302239602,
+      "learning_rate": 1.2305862768367825e-06,
+      "loss": 0.6419,
+      "step": 14857
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.0208671871928323,
+      "learning_rate": 1.230492271251207e-06,
+      "loss": 0.6602,
+      "step": 14858
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.0694375717846825,
+      "learning_rate": 1.2303982635144273e-06,
+      "loss": 0.6693,
+      "step": 14859
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.5838115236571224,
+      "learning_rate": 1.2303042536273196e-06,
+      "loss": 0.6517,
+      "step": 14860
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.075132142652196,
+      "learning_rate": 1.2302102415907622e-06,
+      "loss": 0.6146,
+      "step": 14861
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.345147673171704,
+      "learning_rate": 1.2301162274056322e-06,
+      "loss": 0.681,
+      "step": 14862
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.493609918073523,
+      "learning_rate": 1.230022211072807e-06,
+      "loss": 0.6302,
+      "step": 14863
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.180331236032236,
+      "learning_rate": 1.2299281925931643e-06,
+      "loss": 0.6888,
+      "step": 14864
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.570147226418743,
+      "learning_rate": 1.2298341719675812e-06,
+      "loss": 0.6354,
+      "step": 14865
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.413159904516105,
+      "learning_rate": 1.2297401491969357e-06,
+      "loss": 0.6152,
+      "step": 14866
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.503808785046484,
+      "learning_rate": 1.2296461242821052e-06,
+      "loss": 0.6673,
+      "step": 14867
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.096795527644905,
+      "learning_rate": 1.2295520972239668e-06,
+      "loss": 0.6426,
+      "step": 14868
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.013433398498144,
+      "learning_rate": 1.2294580680233986e-06,
+      "loss": 0.6439,
+      "step": 14869
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.646327776712175,
+      "learning_rate": 1.229364036681278e-06,
+      "loss": 0.6491,
+      "step": 14870
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.175248865550721,
+      "learning_rate": 1.2292700031984822e-06,
+      "loss": 0.6628,
+      "step": 14871
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.2321038679773157,
+      "learning_rate": 1.2291759675758896e-06,
+      "loss": 0.6445,
+      "step": 14872
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.2803920530034665,
+      "learning_rate": 1.229081929814377e-06,
+      "loss": 0.6309,
+      "step": 14873
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.531280975075025,
+      "learning_rate": 1.2289878899148227e-06,
+      "loss": 0.6875,
+      "step": 14874
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.294244646093791,
+      "learning_rate": 1.2288938478781047e-06,
+      "loss": 0.6263,
+      "step": 14875
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.455831984070948,
+      "learning_rate": 1.2287998037050993e-06,
+      "loss": 0.6517,
+      "step": 14876
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.5022888977262827,
+      "learning_rate": 1.2287057573966857e-06,
+      "loss": 0.6348,
+      "step": 14877
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.6177095800203323,
+      "learning_rate": 1.2286117089537408e-06,
+      "loss": 0.6595,
+      "step": 14878
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.0769677484291593,
+      "learning_rate": 1.2285176583771423e-06,
+      "loss": 0.7025,
+      "step": 14879
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.8522028310047416,
+      "learning_rate": 1.2284236056677687e-06,
+      "loss": 0.6615,
+      "step": 14880
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.848876919097087,
+      "learning_rate": 1.2283295508264973e-06,
+      "loss": 0.6576,
+      "step": 14881
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.43828338751755,
+      "learning_rate": 1.228235493854206e-06,
+      "loss": 0.6458,
+      "step": 14882
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.8538484193100135,
+      "learning_rate": 1.2281414347517728e-06,
+      "loss": 0.6745,
+      "step": 14883
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.1384431597981566,
+      "learning_rate": 1.2280473735200752e-06,
+      "loss": 0.6634,
+      "step": 14884
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.98300003134596,
+      "learning_rate": 1.2279533101599914e-06,
+      "loss": 0.6439,
+      "step": 14885
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.323841448161686,
+      "learning_rate": 1.2278592446723992e-06,
+      "loss": 0.6621,
+      "step": 14886
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.1685562515543615,
+      "learning_rate": 1.2277651770581763e-06,
+      "loss": 0.6471,
+      "step": 14887
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.3217006867579033,
+      "learning_rate": 1.227671107318201e-06,
+      "loss": 0.625,
+      "step": 14888
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.7711284102937617,
+      "learning_rate": 1.2275770354533514e-06,
+      "loss": 0.6693,
+      "step": 14889
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.213852650089714,
+      "learning_rate": 1.2274829614645048e-06,
+      "loss": 0.6198,
+      "step": 14890
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.869401874099204,
+      "learning_rate": 1.2273888853525399e-06,
+      "loss": 0.7031,
+      "step": 14891
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.856540106657301,
+      "learning_rate": 1.2272948071183343e-06,
+      "loss": 0.6283,
+      "step": 14892
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.4894298566449455,
+      "learning_rate": 1.2272007267627663e-06,
+      "loss": 0.6452,
+      "step": 14893
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.991350565724439,
+      "learning_rate": 1.2271066442867135e-06,
+      "loss": 0.6224,
+      "step": 14894
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 5.1380605296612325,
+      "learning_rate": 1.2270125596910546e-06,
+      "loss": 0.6764,
+      "step": 14895
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.6698565170813993,
+      "learning_rate": 1.2269184729766673e-06,
+      "loss": 0.6458,
+      "step": 14896
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.9263413756641183,
+      "learning_rate": 1.2268243841444297e-06,
+      "loss": 0.666,
+      "step": 14897
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.155871220180058,
+      "learning_rate": 1.22673029319522e-06,
+      "loss": 0.6048,
+      "step": 14898
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.510760038976961,
+      "learning_rate": 1.226636200129917e-06,
+      "loss": 0.653,
+      "step": 14899
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.9540869318581455,
+      "learning_rate": 1.2265421049493976e-06,
+      "loss": 0.6491,
+      "step": 14900
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.7250633855498694,
+      "learning_rate": 1.226448007654541e-06,
+      "loss": 0.6725,
+      "step": 14901
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.5473033160154563,
+      "learning_rate": 1.2263539082462254e-06,
+      "loss": 0.6214,
+      "step": 14902
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.778867707562717,
+      "learning_rate": 1.226259806725328e-06,
+      "loss": 0.6406,
+      "step": 14903
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.246283289064335,
+      "learning_rate": 1.2261657030927285e-06,
+      "loss": 0.6335,
+      "step": 14904
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.2668131600026054,
+      "learning_rate": 1.2260715973493043e-06,
+      "loss": 0.6484,
+      "step": 14905
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.714356091875881,
+      "learning_rate": 1.2259774894959337e-06,
+      "loss": 0.6322,
+      "step": 14906
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.534243776794797,
+      "learning_rate": 1.2258833795334956e-06,
+      "loss": 0.6732,
+      "step": 14907
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.7684426282374814,
+      "learning_rate": 1.2257892674628676e-06,
+      "loss": 0.6224,
+      "step": 14908
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.610251339114198,
+      "learning_rate": 1.2256951532849286e-06,
+      "loss": 0.6758,
+      "step": 14909
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.172692988674572,
+      "learning_rate": 1.2256010370005567e-06,
+      "loss": 0.6419,
+      "step": 14910
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.677320735609074,
+      "learning_rate": 1.2255069186106301e-06,
+      "loss": 0.6543,
+      "step": 14911
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.684540955842827,
+      "learning_rate": 1.225412798116028e-06,
+      "loss": 0.653,
+      "step": 14912
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.587167399768182,
+      "learning_rate": 1.2253186755176282e-06,
+      "loss": 0.6341,
+      "step": 14913
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.124523532697952,
+      "learning_rate": 1.225224550816309e-06,
+      "loss": 0.6302,
+      "step": 14914
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.1508168961145455,
+      "learning_rate": 1.2251304240129494e-06,
+      "loss": 0.6243,
+      "step": 14915
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.053458626520905,
+      "learning_rate": 1.2250362951084274e-06,
+      "loss": 0.6901,
+      "step": 14916
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 4.055842363481635,
+      "learning_rate": 1.224942164103622e-06,
+      "loss": 0.6549,
+      "step": 14917
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.7516058228412934,
+      "learning_rate": 1.2248480309994113e-06,
+      "loss": 0.6439,
+      "step": 14918
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.289368819467012,
+      "learning_rate": 1.2247538957966741e-06,
+      "loss": 0.7064,
+      "step": 14919
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.1426867878235774,
+      "learning_rate": 1.224659758496289e-06,
+      "loss": 0.6146,
+      "step": 14920
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.3626716976959874,
+      "learning_rate": 1.2245656190991342e-06,
+      "loss": 0.6413,
+      "step": 14921
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.95641206486636,
+      "learning_rate": 1.2244714776060892e-06,
+      "loss": 0.6582,
+      "step": 14922
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.607742843067064,
+      "learning_rate": 1.2243773340180317e-06,
+      "loss": 0.6003,
+      "step": 14923
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.8435956784432843,
+      "learning_rate": 1.2242831883358404e-06,
+      "loss": 0.6445,
+      "step": 14924
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.543851144382488,
+      "learning_rate": 1.224189040560395e-06,
+      "loss": 0.6634,
+      "step": 14925
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.483720972547578,
+      "learning_rate": 1.2240948906925728e-06,
+      "loss": 0.5892,
+      "step": 14926
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.8795077345881475,
+      "learning_rate": 1.2240007387332535e-06,
+      "loss": 0.6133,
+      "step": 14927
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.3220531316406703,
+      "learning_rate": 1.2239065846833155e-06,
+      "loss": 0.5951,
+      "step": 14928
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.172593049376459,
+      "learning_rate": 1.2238124285436374e-06,
+      "loss": 0.6361,
+      "step": 14929
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.647991696928895,
+      "learning_rate": 1.2237182703150982e-06,
+      "loss": 0.6309,
+      "step": 14930
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.993210288698324,
+      "learning_rate": 1.2236241099985762e-06,
+      "loss": 0.6491,
+      "step": 14931
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.7252812742997676,
+      "learning_rate": 1.2235299475949512e-06,
+      "loss": 0.6745,
+      "step": 14932
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.6342025689664204,
+      "learning_rate": 1.2234357831051012e-06,
+      "loss": 0.6455,
+      "step": 14933
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.824185006894598,
+      "learning_rate": 1.223341616529905e-06,
+      "loss": 0.6419,
+      "step": 14934
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.837842234573174,
+      "learning_rate": 1.2232474478702421e-06,
+      "loss": 0.638,
+      "step": 14935
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.323179965534723,
+      "learning_rate": 1.223153277126991e-06,
+      "loss": 0.6374,
+      "step": 14936
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.7729106928715757,
+      "learning_rate": 1.2230591043010305e-06,
+      "loss": 0.6803,
+      "step": 14937
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.6106876092854376,
+      "learning_rate": 1.22296492939324e-06,
+      "loss": 0.6439,
+      "step": 14938
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.510332468272036,
+      "learning_rate": 1.2228707524044976e-06,
+      "loss": 0.6296,
+      "step": 14939
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.773020544297159,
+      "learning_rate": 1.222776573335683e-06,
+      "loss": 0.6738,
+      "step": 14940
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.1652797704998936,
+      "learning_rate": 1.222682392187675e-06,
+      "loss": 0.6947,
+      "step": 14941
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.9317107553164434,
+      "learning_rate": 1.2225882089613525e-06,
+      "loss": 0.6517,
+      "step": 14942
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.8328082198448095,
+      "learning_rate": 1.2224940236575944e-06,
+      "loss": 0.6471,
+      "step": 14943
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.3897386014666124,
+      "learning_rate": 1.22239983627728e-06,
+      "loss": 0.6016,
+      "step": 14944
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.4774104379214066,
+      "learning_rate": 1.2223056468212885e-06,
+      "loss": 0.6758,
+      "step": 14945
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.664944867372338,
+      "learning_rate": 1.2222114552904985e-06,
+      "loss": 0.6732,
+      "step": 14946
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.3164986910823915,
+      "learning_rate": 1.2221172616857892e-06,
+      "loss": 0.6341,
+      "step": 14947
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.4254899606884726,
+      "learning_rate": 1.22202306600804e-06,
+      "loss": 0.6419,
+      "step": 14948
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.5633120394412354,
+      "learning_rate": 1.22192886825813e-06,
+      "loss": 0.6556,
+      "step": 14949
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.21862264091901,
+      "learning_rate": 1.221834668436938e-06,
+      "loss": 0.6445,
+      "step": 14950
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.531473709553242,
+      "learning_rate": 1.2217404665453437e-06,
+      "loss": 0.6354,
+      "step": 14951
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.6803760463346173,
+      "learning_rate": 1.221646262584226e-06,
+      "loss": 0.6491,
+      "step": 14952
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.5633049511178525,
+      "learning_rate": 1.2215520565544638e-06,
+      "loss": 0.6641,
+      "step": 14953
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.497661977545487,
+      "learning_rate": 1.2214578484569369e-06,
+      "loss": 0.666,
+      "step": 14954
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.264893433678513,
+      "learning_rate": 1.2213636382925244e-06,
+      "loss": 0.6458,
+      "step": 14955
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.0599054643543706,
+      "learning_rate": 1.221269426062105e-06,
+      "loss": 0.6113,
+      "step": 14956
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 6.355018986244844,
+      "learning_rate": 1.2211752117665591e-06,
+      "loss": 0.6257,
+      "step": 14957
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.374034231123579,
+      "learning_rate": 1.2210809954067652e-06,
+      "loss": 0.6432,
+      "step": 14958
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.1554283268316823,
+      "learning_rate": 1.220986776983603e-06,
+      "loss": 0.6139,
+      "step": 14959
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.12471833072646,
+      "learning_rate": 1.2208925564979511e-06,
+      "loss": 0.6419,
+      "step": 14960
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.795324425519524,
+      "learning_rate": 1.2207983339506897e-06,
+      "loss": 0.6628,
+      "step": 14961
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.4362275418740866,
+      "learning_rate": 1.220704109342698e-06,
+      "loss": 0.6445,
+      "step": 14962
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.149436097997367,
+      "learning_rate": 1.2206098826748554e-06,
+      "loss": 0.653,
+      "step": 14963
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.83277888811337,
+      "learning_rate": 1.2205156539480411e-06,
+      "loss": 0.6504,
+      "step": 14964
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.5502171479313014,
+      "learning_rate": 1.2204214231631348e-06,
+      "loss": 0.6367,
+      "step": 14965
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.675823510031592,
+      "learning_rate": 1.2203271903210157e-06,
+      "loss": 0.6387,
+      "step": 14966
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.3904408514651063,
+      "learning_rate": 1.2202329554225638e-06,
+      "loss": 0.6198,
+      "step": 14967
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.4920719919736274,
+      "learning_rate": 1.220138718468658e-06,
+      "loss": 0.6419,
+      "step": 14968
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.530068143195515,
+      "learning_rate": 1.220044479460178e-06,
+      "loss": 0.6445,
+      "step": 14969
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.0963784266639167,
+      "learning_rate": 1.2199502383980035e-06,
+      "loss": 0.6641,
+      "step": 14970
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.649001340568826,
+      "learning_rate": 1.2198559952830141e-06,
+      "loss": 0.6211,
+      "step": 14971
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.447840727029424,
+      "learning_rate": 1.2197617501160892e-06,
+      "loss": 0.6836,
+      "step": 14972
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.316059849865935,
+      "learning_rate": 1.2196675028981082e-06,
+      "loss": 0.6634,
+      "step": 14973
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.869552214653787,
+      "learning_rate": 1.2195732536299513e-06,
+      "loss": 0.6296,
+      "step": 14974
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.580098592464276,
+      "learning_rate": 1.219479002312498e-06,
+      "loss": 0.6523,
+      "step": 14975
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.4182591593321416,
+      "learning_rate": 1.2193847489466272e-06,
+      "loss": 0.6159,
+      "step": 14976
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.5161664991709123,
+      "learning_rate": 1.2192904935332191e-06,
+      "loss": 0.6426,
+      "step": 14977
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 5.877126831836021,
+      "learning_rate": 1.219196236073154e-06,
+      "loss": 0.6315,
+      "step": 14978
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.9453893885318174,
+      "learning_rate": 1.2191019765673106e-06,
+      "loss": 0.6452,
+      "step": 14979
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.7614027663970897,
+      "learning_rate": 1.2190077150165695e-06,
+      "loss": 0.6966,
+      "step": 14980
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.001378536077536,
+      "learning_rate": 1.2189134514218096e-06,
+      "loss": 0.624,
+      "step": 14981
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 3.9931089189550066,
+      "learning_rate": 1.2188191857839111e-06,
+      "loss": 0.6999,
+      "step": 14982
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.5448994535146676,
+      "learning_rate": 1.218724918103754e-06,
+      "loss": 0.6413,
+      "step": 14983
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 4.607153672166351,
+      "learning_rate": 1.2186306483822178e-06,
+      "loss": 0.6901,
+      "step": 14984
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.7585352468960913,
+      "learning_rate": 1.2185363766201824e-06,
+      "loss": 0.6921,
+      "step": 14985
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.3367307116628933,
+      "learning_rate": 1.2184421028185279e-06,
+      "loss": 0.6481,
+      "step": 14986
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.708620407341763,
+      "learning_rate": 1.2183478269781336e-06,
+      "loss": 0.6204,
+      "step": 14987
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.728547828368543,
+      "learning_rate": 1.21825354909988e-06,
+      "loss": 0.6621,
+      "step": 14988
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.6624579379113342,
+      "learning_rate": 1.2181592691846466e-06,
+      "loss": 0.6302,
+      "step": 14989
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.0657032102416255,
+      "learning_rate": 1.2180649872333133e-06,
+      "loss": 0.6523,
+      "step": 14990
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.7035802503469695,
+      "learning_rate": 1.2179707032467606e-06,
+      "loss": 0.6497,
+      "step": 14991
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.321554282542098,
+      "learning_rate": 1.2178764172258676e-06,
+      "loss": 0.6471,
+      "step": 14992
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.79410583250377,
+      "learning_rate": 1.2177821291715152e-06,
+      "loss": 0.5957,
+      "step": 14993
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.760228973143944,
+      "learning_rate": 1.2176878390845826e-06,
+      "loss": 0.6523,
+      "step": 14994
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.432944746515558,
+      "learning_rate": 1.21759354696595e-06,
+      "loss": 0.6439,
+      "step": 14995
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.8848740252334455,
+      "learning_rate": 1.2174992528164979e-06,
+      "loss": 0.6458,
+      "step": 14996
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.3680375322692377,
+      "learning_rate": 1.217404956637106e-06,
+      "loss": 0.6107,
+      "step": 14997
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.3640605017818785,
+      "learning_rate": 1.217310658428654e-06,
+      "loss": 0.6224,
+      "step": 14998
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.525248347534493,
+      "learning_rate": 1.217216358192023e-06,
+      "loss": 0.653,
+      "step": 14999
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.8875420722615726,
+      "learning_rate": 1.2171220559280921e-06,
+      "loss": 0.6497,
+      "step": 15000
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.03616289860513,
+      "learning_rate": 1.217027751637742e-06,
+      "loss": 0.5762,
+      "step": 15001
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.102572751788112,
+      "learning_rate": 1.2169334453218529e-06,
+      "loss": 0.651,
+      "step": 15002
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.518145156927488,
+      "learning_rate": 1.2168391369813044e-06,
+      "loss": 0.64,
+      "step": 15003
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.093670368646742,
+      "learning_rate": 1.2167448266169774e-06,
+      "loss": 0.6777,
+      "step": 15004
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.8596982515376883,
+      "learning_rate": 1.2166505142297515e-06,
+      "loss": 0.6361,
+      "step": 15005
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.063526027147953,
+      "learning_rate": 1.2165561998205072e-06,
+      "loss": 0.6348,
+      "step": 15006
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.6595119604960433,
+      "learning_rate": 1.216461883390125e-06,
+      "loss": 0.6191,
+      "step": 15007
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.343829012477665,
+      "learning_rate": 1.2163675649394847e-06,
+      "loss": 0.7012,
+      "step": 15008
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 6.246109810999246,
+      "learning_rate": 1.216273244469467e-06,
+      "loss": 0.612,
+      "step": 15009
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.363215338696499,
+      "learning_rate": 1.2161789219809516e-06,
+      "loss": 0.6178,
+      "step": 15010
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.2951142196895864,
+      "learning_rate": 1.2160845974748196e-06,
+      "loss": 0.5947,
+      "step": 15011
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.3618862627780937,
+      "learning_rate": 1.2159902709519506e-06,
+      "loss": 0.6439,
+      "step": 15012
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 5.118430282521023,
+      "learning_rate": 1.2158959424132254e-06,
+      "loss": 0.6641,
+      "step": 15013
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.529673637924808,
+      "learning_rate": 1.2158016118595244e-06,
+      "loss": 0.6699,
+      "step": 15014
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.2205623175682225,
+      "learning_rate": 1.2157072792917278e-06,
+      "loss": 0.6836,
+      "step": 15015
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 10.23939689993248,
+      "learning_rate": 1.215612944710716e-06,
+      "loss": 0.6576,
+      "step": 15016
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.857040956093604,
+      "learning_rate": 1.2155186081173695e-06,
+      "loss": 0.6387,
+      "step": 15017
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.604307041734988,
+      "learning_rate": 1.2154242695125692e-06,
+      "loss": 0.6465,
+      "step": 15018
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.255893456449706,
+      "learning_rate": 1.2153299288971946e-06,
+      "loss": 0.6484,
+      "step": 15019
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.6727235952049186,
+      "learning_rate": 1.2152355862721269e-06,
+      "loss": 0.6341,
+      "step": 15020
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.5896473005311016,
+      "learning_rate": 1.2151412416382468e-06,
+      "loss": 0.6419,
+      "step": 15021
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.72416030142104,
+      "learning_rate": 1.215046894996434e-06,
+      "loss": 0.6413,
+      "step": 15022
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.9051715289127675,
+      "learning_rate": 1.2149525463475696e-06,
+      "loss": 0.666,
+      "step": 15023
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.3913526392672417,
+      "learning_rate": 1.2148581956925342e-06,
+      "loss": 0.6569,
+      "step": 15024
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.8184617728691013,
+      "learning_rate": 1.2147638430322082e-06,
+      "loss": 0.623,
+      "step": 15025
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.60749768572689,
+      "learning_rate": 1.2146694883674722e-06,
+      "loss": 0.6504,
+      "step": 15026
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.0690636579991097,
+      "learning_rate": 1.214575131699207e-06,
+      "loss": 0.6751,
+      "step": 15027
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.29910572576604,
+      "learning_rate": 1.214480773028293e-06,
+      "loss": 0.6576,
+      "step": 15028
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.608633893889012,
+      "learning_rate": 1.214386412355611e-06,
+      "loss": 0.6745,
+      "step": 15029
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.1127673160513174,
+      "learning_rate": 1.2142920496820418e-06,
+      "loss": 0.6003,
+      "step": 15030
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.1758989972116005,
+      "learning_rate": 1.2141976850084658e-06,
+      "loss": 0.6608,
+      "step": 15031
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.731216472850241,
+      "learning_rate": 1.2141033183357636e-06,
+      "loss": 0.7038,
+      "step": 15032
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.256449953309859,
+      "learning_rate": 1.2140089496648168e-06,
+      "loss": 0.6146,
+      "step": 15033
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.7360217253552213,
+      "learning_rate": 1.2139145789965053e-06,
+      "loss": 0.6029,
+      "step": 15034
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.4094210641449703,
+      "learning_rate": 1.2138202063317097e-06,
+      "loss": 0.6393,
+      "step": 15035
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.264536405618316,
+      "learning_rate": 1.2137258316713116e-06,
+      "loss": 0.668,
+      "step": 15036
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.1497696821511036,
+      "learning_rate": 1.2136314550161916e-06,
+      "loss": 0.6266,
+      "step": 15037
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.6370606489808033,
+      "learning_rate": 1.21353707636723e-06,
+      "loss": 0.5931,
+      "step": 15038
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.1966777168120992,
+      "learning_rate": 1.213442695725308e-06,
+      "loss": 0.625,
+      "step": 15039
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.520722819473551,
+      "learning_rate": 1.2133483130913065e-06,
+      "loss": 0.6204,
+      "step": 15040
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.475536002947585,
+      "learning_rate": 1.2132539284661066e-06,
+      "loss": 0.6582,
+      "step": 15041
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.0621103388771393,
+      "learning_rate": 1.2131595418505888e-06,
+      "loss": 0.6549,
+      "step": 15042
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.6106867528996367,
+      "learning_rate": 1.213065153245634e-06,
+      "loss": 0.6471,
+      "step": 15043
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 3.7338902617244307,
+      "learning_rate": 1.2129707626521234e-06,
+      "loss": 0.6276,
+      "step": 15044
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.5474802957420213,
+      "learning_rate": 1.2128763700709374e-06,
+      "loss": 0.6458,
+      "step": 15045
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.5230745558124283,
+      "learning_rate": 1.212781975502958e-06,
+      "loss": 0.6257,
+      "step": 15046
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.7422119740197313,
+      "learning_rate": 1.2126875789490654e-06,
+      "loss": 0.6491,
+      "step": 15047
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.423415466690318,
+      "learning_rate": 1.2125931804101406e-06,
+      "loss": 0.6055,
+      "step": 15048
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.2381158305589106,
+      "learning_rate": 1.2124987798870652e-06,
+      "loss": 0.6257,
+      "step": 15049
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 7.699796551411144,
+      "learning_rate": 1.2124043773807196e-06,
+      "loss": 0.6335,
+      "step": 15050
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.544073588982971,
+      "learning_rate": 1.2123099728919852e-06,
+      "loss": 0.6335,
+      "step": 15051
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 4.0365466833429675,
+      "learning_rate": 1.2122155664217434e-06,
+      "loss": 0.6126,
+      "step": 15052
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.537505181023195,
+      "learning_rate": 1.2121211579708742e-06,
+      "loss": 0.6419,
+      "step": 15053
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.547026386290593,
+      "learning_rate": 1.2120267475402606e-06,
+      "loss": 0.6025,
+      "step": 15054
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.722324600625174,
+      "learning_rate": 1.2119323351307818e-06,
+      "loss": 0.6367,
+      "step": 15055
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.8611535191027,
+      "learning_rate": 1.2118379207433198e-06,
+      "loss": 0.6608,
+      "step": 15056
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.479727212343739,
+      "learning_rate": 1.211743504378756e-06,
+      "loss": 0.6328,
+      "step": 15057
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.061803401763941,
+      "learning_rate": 1.2116490860379712e-06,
+      "loss": 0.6497,
+      "step": 15058
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.512784468948868,
+      "learning_rate": 1.211554665721847e-06,
+      "loss": 0.6621,
+      "step": 15059
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.7782489283976917,
+      "learning_rate": 1.211460243431264e-06,
+      "loss": 0.6641,
+      "step": 15060
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.735835953791786,
+      "learning_rate": 1.211365819167104e-06,
+      "loss": 0.6608,
+      "step": 15061
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.35304992055977,
+      "learning_rate": 1.2112713929302484e-06,
+      "loss": 0.6556,
+      "step": 15062
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.3414170957899096,
+      "learning_rate": 1.2111769647215776e-06,
+      "loss": 0.6471,
+      "step": 15063
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.528642240150095,
+      "learning_rate": 1.2110825345419741e-06,
+      "loss": 0.6706,
+      "step": 15064
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.803666811359874,
+      "learning_rate": 1.2109881023923185e-06,
+      "loss": 0.5931,
+      "step": 15065
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.1713731309368858,
+      "learning_rate": 1.2108936682734921e-06,
+      "loss": 0.6413,
+      "step": 15066
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.9596762083419645,
+      "learning_rate": 1.2107992321863768e-06,
+      "loss": 0.6165,
+      "step": 15067
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.0899678470749627,
+      "learning_rate": 1.2107047941318533e-06,
+      "loss": 0.6191,
+      "step": 15068
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.0685181224407745,
+      "learning_rate": 1.2106103541108033e-06,
+      "loss": 0.6217,
+      "step": 15069
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.585081409390481,
+      "learning_rate": 1.2105159121241085e-06,
+      "loss": 0.6003,
+      "step": 15070
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.4200041025332815,
+      "learning_rate": 1.21042146817265e-06,
+      "loss": 0.6452,
+      "step": 15071
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.5216861570850573,
+      "learning_rate": 1.210327022257309e-06,
+      "loss": 0.6589,
+      "step": 15072
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.777960295058669,
+      "learning_rate": 1.2102325743789677e-06,
+      "loss": 0.6777,
+      "step": 15073
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.2475203956530305,
+      "learning_rate": 1.210138124538507e-06,
+      "loss": 0.6595,
+      "step": 15074
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.1232894406592804,
+      "learning_rate": 1.2100436727368086e-06,
+      "loss": 0.5938,
+      "step": 15075
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.0015511787064795,
+      "learning_rate": 1.209949218974754e-06,
+      "loss": 0.5983,
+      "step": 15076
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.7237979303136726,
+      "learning_rate": 1.209854763253225e-06,
+      "loss": 0.627,
+      "step": 15077
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.823351636568076,
+      "learning_rate": 1.2097603055731026e-06,
+      "loss": 0.5964,
+      "step": 15078
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.9598444373059785,
+      "learning_rate": 1.209665845935269e-06,
+      "loss": 0.6953,
+      "step": 15079
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.6722904146969704,
+      "learning_rate": 1.2095713843406055e-06,
+      "loss": 0.6452,
+      "step": 15080
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.8875882345861346,
+      "learning_rate": 1.2094769207899935e-06,
+      "loss": 0.6374,
+      "step": 15081
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.523861831495266,
+      "learning_rate": 1.209382455284315e-06,
+      "loss": 0.6182,
+      "step": 15082
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.0769881698506403,
+      "learning_rate": 1.2092879878244514e-06,
+      "loss": 0.612,
+      "step": 15083
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.1712611656905687,
+      "learning_rate": 1.2091935184112849e-06,
+      "loss": 0.6328,
+      "step": 15084
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.507145712798861,
+      "learning_rate": 1.2090990470456964e-06,
+      "loss": 0.6335,
+      "step": 15085
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.49799498547592,
+      "learning_rate": 1.2090045737285684e-06,
+      "loss": 0.6517,
+      "step": 15086
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.448965582536913,
+      "learning_rate": 1.2089100984607819e-06,
+      "loss": 0.6315,
+      "step": 15087
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.672911594131762,
+      "learning_rate": 1.208815621243219e-06,
+      "loss": 0.6927,
+      "step": 15088
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.922601877340478,
+      "learning_rate": 1.2087211420767615e-06,
+      "loss": 0.6979,
+      "step": 15089
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 7.108385608399602,
+      "learning_rate": 1.208626660962291e-06,
+      "loss": 0.6087,
+      "step": 15090
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.6617348925826225,
+      "learning_rate": 1.2085321779006898e-06,
+      "loss": 0.6569,
+      "step": 15091
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.4721828454098094,
+      "learning_rate": 1.2084376928928389e-06,
+      "loss": 0.6608,
+      "step": 15092
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.63303006716086,
+      "learning_rate": 1.2083432059396205e-06,
+      "loss": 0.6159,
+      "step": 15093
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.148290368562641,
+      "learning_rate": 1.208248717041917e-06,
+      "loss": 0.638,
+      "step": 15094
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 5.849541786378522,
+      "learning_rate": 1.2081542262006097e-06,
+      "loss": 0.6276,
+      "step": 15095
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.8888342139101506,
+      "learning_rate": 1.2080597334165806e-06,
+      "loss": 0.6406,
+      "step": 15096
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.58373012682329,
+      "learning_rate": 1.2079652386907116e-06,
+      "loss": 0.6517,
+      "step": 15097
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.8559842937201396,
+      "learning_rate": 1.2078707420238843e-06,
+      "loss": 0.6393,
+      "step": 15098
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.9873154121088126,
+      "learning_rate": 1.2077762434169812e-06,
+      "loss": 0.7012,
+      "step": 15099
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.600160841799293,
+      "learning_rate": 1.2076817428708843e-06,
+      "loss": 0.6582,
+      "step": 15100
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.509358379731099,
+      "learning_rate": 1.2075872403864748e-06,
+      "loss": 0.6497,
+      "step": 15101
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.4333546378933466,
+      "learning_rate": 1.2074927359646357e-06,
+      "loss": 0.6816,
+      "step": 15102
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.3388580166444086,
+      "learning_rate": 1.2073982296062483e-06,
+      "loss": 0.6185,
+      "step": 15103
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.336551505698871,
+      "learning_rate": 1.2073037213121949e-06,
+      "loss": 0.6009,
+      "step": 15104
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.5504695410776907,
+      "learning_rate": 1.2072092110833577e-06,
+      "loss": 0.6628,
+      "step": 15105
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.032671048424238,
+      "learning_rate": 1.2071146989206183e-06,
+      "loss": 0.6673,
+      "step": 15106
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.944895019273526,
+      "learning_rate": 1.2070201848248595e-06,
+      "loss": 0.6367,
+      "step": 15107
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.3584723579269147,
+      "learning_rate": 1.2069256687969627e-06,
+      "loss": 0.5872,
+      "step": 15108
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.59033337056881,
+      "learning_rate": 1.2068311508378106e-06,
+      "loss": 0.7057,
+      "step": 15109
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.4889789164594425,
+      "learning_rate": 1.2067366309482849e-06,
+      "loss": 0.6543,
+      "step": 15110
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 6.302509877913746,
+      "learning_rate": 1.2066421091292678e-06,
+      "loss": 0.6849,
+      "step": 15111
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 4.67338875794966,
+      "learning_rate": 1.2065475853816419e-06,
+      "loss": 0.6159,
+      "step": 15112
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.347277271143919,
+      "learning_rate": 1.206453059706289e-06,
+      "loss": 0.6602,
+      "step": 15113
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.7989030840733515,
+      "learning_rate": 1.2063585321040912e-06,
+      "loss": 0.6328,
+      "step": 15114
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.5576213702762836,
+      "learning_rate": 1.2062640025759314e-06,
+      "loss": 0.6387,
+      "step": 15115
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.218218080224127,
+      "learning_rate": 1.2061694711226911e-06,
+      "loss": 0.668,
+      "step": 15116
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.250028664265338,
+      "learning_rate": 1.206074937745253e-06,
+      "loss": 0.6654,
+      "step": 15117
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 3.556951177552612,
+      "learning_rate": 1.2059804024444994e-06,
+      "loss": 0.6445,
+      "step": 15118
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.5672000250052824,
+      "learning_rate": 1.2058858652213122e-06,
+      "loss": 0.6706,
+      "step": 15119
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.4906204009002777,
+      "learning_rate": 1.2057913260765746e-06,
+      "loss": 0.6393,
+      "step": 15120
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.5000364893373255,
+      "learning_rate": 1.2056967850111677e-06,
+      "loss": 0.6341,
+      "step": 15121
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.342898420813274,
+      "learning_rate": 1.2056022420259745e-06,
+      "loss": 0.6224,
+      "step": 15122
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.9706535827929224,
+      "learning_rate": 1.205507697121878e-06,
+      "loss": 0.6152,
+      "step": 15123
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.2011328268790575,
+      "learning_rate": 1.2054131502997594e-06,
+      "loss": 0.6302,
+      "step": 15124
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.3700089992531224,
+      "learning_rate": 1.2053186015605022e-06,
+      "loss": 0.6445,
+      "step": 15125
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.2934850168867618,
+      "learning_rate": 1.2052240509049881e-06,
+      "loss": 0.6497,
+      "step": 15126
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.8687652652694746,
+      "learning_rate": 1.2051294983340996e-06,
+      "loss": 0.7038,
+      "step": 15127
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.4494667654233857,
+      "learning_rate": 1.2050349438487197e-06,
+      "loss": 0.6296,
+      "step": 15128
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.0371616048299526,
+      "learning_rate": 1.2049403874497302e-06,
+      "loss": 0.6393,
+      "step": 15129
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.3851142047639686,
+      "learning_rate": 1.204845829138014e-06,
+      "loss": 0.6198,
+      "step": 15130
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.191245658249985,
+      "learning_rate": 1.204751268914454e-06,
+      "loss": 0.6491,
+      "step": 15131
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.777278005899018,
+      "learning_rate": 1.2046567067799318e-06,
+      "loss": 0.6699,
+      "step": 15132
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.2347708928449515,
+      "learning_rate": 1.2045621427353304e-06,
+      "loss": 0.6562,
+      "step": 15133
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.416769624750579,
+      "learning_rate": 1.204467576781533e-06,
+      "loss": 0.6523,
+      "step": 15134
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.533225379488638,
+      "learning_rate": 1.204373008919421e-06,
+      "loss": 0.6367,
+      "step": 15135
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.774542803338995,
+      "learning_rate": 1.204278439149878e-06,
+      "loss": 0.6048,
+      "step": 15136
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.816185493407762,
+      "learning_rate": 1.2041838674737863e-06,
+      "loss": 0.6667,
+      "step": 15137
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.955298537853003,
+      "learning_rate": 1.2040892938920282e-06,
+      "loss": 0.6621,
+      "step": 15138
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 5.668101987462419,
+      "learning_rate": 1.2039947184054869e-06,
+      "loss": 0.7025,
+      "step": 15139
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.7770436319663503,
+      "learning_rate": 1.2039001410150448e-06,
+      "loss": 0.638,
+      "step": 15140
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.9881887600324584,
+      "learning_rate": 1.2038055617215848e-06,
+      "loss": 0.5684,
+      "step": 15141
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.183360883977672,
+      "learning_rate": 1.203710980525989e-06,
+      "loss": 0.7129,
+      "step": 15142
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.6669415121890663,
+      "learning_rate": 1.203616397429141e-06,
+      "loss": 0.6504,
+      "step": 15143
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.580710219442675,
+      "learning_rate": 1.2035218124319232e-06,
+      "loss": 0.5999,
+      "step": 15144
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.989388249141368,
+      "learning_rate": 1.203427225535218e-06,
+      "loss": 0.6569,
+      "step": 15145
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.6657226902934403,
+      "learning_rate": 1.2033326367399091e-06,
+      "loss": 0.6296,
+      "step": 15146
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.5345888972480615,
+      "learning_rate": 1.2032380460468783e-06,
+      "loss": 0.6094,
+      "step": 15147
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.971758221440026,
+      "learning_rate": 1.2031434534570087e-06,
+      "loss": 0.6374,
+      "step": 15148
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.667938181558141,
+      "learning_rate": 1.2030488589711836e-06,
+      "loss": 0.5996,
+      "step": 15149
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.6796399201350534,
+      "learning_rate": 1.2029542625902853e-06,
+      "loss": 0.6465,
+      "step": 15150
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.9427865648481615,
+      "learning_rate": 1.202859664315197e-06,
+      "loss": 0.6211,
+      "step": 15151
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.201029261700578,
+      "learning_rate": 1.2027650641468015e-06,
+      "loss": 0.6686,
+      "step": 15152
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.248434669256806,
+      "learning_rate": 1.202670462085982e-06,
+      "loss": 0.6322,
+      "step": 15153
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.865049141399605,
+      "learning_rate": 1.2025758581336207e-06,
+      "loss": 0.696,
+      "step": 15154
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.6018616166421764,
+      "learning_rate": 1.2024812522906014e-06,
+      "loss": 0.6562,
+      "step": 15155
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.5372697222374803,
+      "learning_rate": 1.2023866445578063e-06,
+      "loss": 0.6497,
+      "step": 15156
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.6494908802933503,
+      "learning_rate": 1.2022920349361192e-06,
+      "loss": 0.6322,
+      "step": 15157
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.8532895470162365,
+      "learning_rate": 1.2021974234264222e-06,
+      "loss": 0.6367,
+      "step": 15158
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.191991024240357,
+      "learning_rate": 1.2021028100295986e-06,
+      "loss": 0.681,
+      "step": 15159
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.739667840381369,
+      "learning_rate": 1.2020081947465324e-06,
+      "loss": 0.6419,
+      "step": 15160
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.447999529817779,
+      "learning_rate": 1.2019135775781054e-06,
+      "loss": 0.6673,
+      "step": 15161
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.5425809092687053,
+      "learning_rate": 1.201818958525201e-06,
+      "loss": 0.6263,
+      "step": 15162
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 4.267043324280155,
+      "learning_rate": 1.2017243375887027e-06,
+      "loss": 0.7031,
+      "step": 15163
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.0674214197403358,
+      "learning_rate": 1.2016297147694929e-06,
+      "loss": 0.6471,
+      "step": 15164
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.3601632237901673,
+      "learning_rate": 1.2015350900684556e-06,
+      "loss": 0.6087,
+      "step": 15165
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.4387887529311,
+      "learning_rate": 1.2014404634864732e-06,
+      "loss": 0.6016,
+      "step": 15166
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.5912735531603746,
+      "learning_rate": 1.2013458350244293e-06,
+      "loss": 0.6452,
+      "step": 15167
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.3120231009802854,
+      "learning_rate": 1.2012512046832069e-06,
+      "loss": 0.5999,
+      "step": 15168
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.406743235835118,
+      "learning_rate": 1.201156572463689e-06,
+      "loss": 0.6712,
+      "step": 15169
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.6288820498464758,
+      "learning_rate": 1.2010619383667592e-06,
+      "loss": 0.668,
+      "step": 15170
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.334192757431038,
+      "learning_rate": 1.2009673023933005e-06,
+      "loss": 0.6061,
+      "step": 15171
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.039355686377249,
+      "learning_rate": 1.200872664544196e-06,
+      "loss": 0.6341,
+      "step": 15172
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.4156435642093097,
+      "learning_rate": 1.2007780248203297e-06,
+      "loss": 0.6071,
+      "step": 15173
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.99330179413182,
+      "learning_rate": 1.2006833832225837e-06,
+      "loss": 0.623,
+      "step": 15174
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.878689940139553,
+      "learning_rate": 1.2005887397518424e-06,
+      "loss": 0.6712,
+      "step": 15175
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.754331800687567,
+      "learning_rate": 1.2004940944089885e-06,
+      "loss": 0.6315,
+      "step": 15176
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.1218672576263473,
+      "learning_rate": 1.2003994471949054e-06,
+      "loss": 0.6478,
+      "step": 15177
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.8929386755860835,
+      "learning_rate": 1.2003047981104768e-06,
+      "loss": 0.5983,
+      "step": 15178
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.6739256572096357,
+      "learning_rate": 1.2002101471565855e-06,
+      "loss": 0.6165,
+      "step": 15179
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.8695644981873687,
+      "learning_rate": 1.2001154943341152e-06,
+      "loss": 0.6217,
+      "step": 15180
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.557747435575275,
+      "learning_rate": 1.2000208396439494e-06,
+      "loss": 0.6178,
+      "step": 15181
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.77615279879227,
+      "learning_rate": 1.1999261830869714e-06,
+      "loss": 0.6615,
+      "step": 15182
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.218522623064391,
+      "learning_rate": 1.1998315246640647e-06,
+      "loss": 0.6758,
+      "step": 15183
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.7314414476839524,
+      "learning_rate": 1.1997368643761127e-06,
+      "loss": 0.6576,
+      "step": 15184
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.9492866564321627,
+      "learning_rate": 1.1996422022239986e-06,
+      "loss": 0.5977,
+      "step": 15185
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 3.67490697887356,
+      "learning_rate": 1.1995475382086068e-06,
+      "loss": 0.64,
+      "step": 15186
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.758097519343222,
+      "learning_rate": 1.1994528723308198e-06,
+      "loss": 0.668,
+      "step": 15187
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.849846123074867,
+      "learning_rate": 1.1993582045915213e-06,
+      "loss": 0.6517,
+      "step": 15188
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.747647280513092,
+      "learning_rate": 1.1992635349915956e-06,
+      "loss": 0.5996,
+      "step": 15189
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.347355382547056,
+      "learning_rate": 1.1991688635319252e-06,
+      "loss": 0.651,
+      "step": 15190
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.5564635257832604,
+      "learning_rate": 1.1990741902133943e-06,
+      "loss": 0.6159,
+      "step": 15191
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.5349967398406514,
+      "learning_rate": 1.1989795150368864e-06,
+      "loss": 0.6432,
+      "step": 15192
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.8759686541462277,
+      "learning_rate": 1.1988848380032851e-06,
+      "loss": 0.5853,
+      "step": 15193
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.335255340362606,
+      "learning_rate": 1.198790159113474e-06,
+      "loss": 0.735,
+      "step": 15194
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.7226565320405225,
+      "learning_rate": 1.1986954783683367e-06,
+      "loss": 0.6354,
+      "step": 15195
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.1221246396999387,
+      "learning_rate": 1.1986007957687572e-06,
+      "loss": 0.651,
+      "step": 15196
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.799865752252739,
+      "learning_rate": 1.1985061113156187e-06,
+      "loss": 0.6341,
+      "step": 15197
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.437636435289581,
+      "learning_rate": 1.1984114250098051e-06,
+      "loss": 0.6777,
+      "step": 15198
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.6693213378768874,
+      "learning_rate": 1.1983167368522002e-06,
+      "loss": 0.599,
+      "step": 15199
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.1553038252943795,
+      "learning_rate": 1.1982220468436878e-06,
+      "loss": 0.6341,
+      "step": 15200
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.948260688208051,
+      "learning_rate": 1.1981273549851511e-06,
+      "loss": 0.6361,
+      "step": 15201
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.814395484904959,
+      "learning_rate": 1.1980326612774744e-06,
+      "loss": 0.6361,
+      "step": 15202
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.8769763184434165,
+      "learning_rate": 1.1979379657215418e-06,
+      "loss": 0.5918,
+      "step": 15203
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.308587168365515,
+      "learning_rate": 1.1978432683182362e-06,
+      "loss": 0.6133,
+      "step": 15204
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.813579721128315,
+      "learning_rate": 1.1977485690684418e-06,
+      "loss": 0.6289,
+      "step": 15205
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.1152054670884555,
+      "learning_rate": 1.1976538679730431e-06,
+      "loss": 0.6393,
+      "step": 15206
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.6563358111565294,
+      "learning_rate": 1.1975591650329229e-06,
+      "loss": 0.6224,
+      "step": 15207
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.2346904886142185,
+      "learning_rate": 1.1974644602489657e-06,
+      "loss": 0.6413,
+      "step": 15208
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.858387967029551,
+      "learning_rate": 1.197369753622055e-06,
+      "loss": 0.6328,
+      "step": 15209
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.2263241602528554,
+      "learning_rate": 1.1972750451530753e-06,
+      "loss": 0.6354,
+      "step": 15210
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.3673174274595947,
+      "learning_rate": 1.1971803348429098e-06,
+      "loss": 0.6654,
+      "step": 15211
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.7815535828192104,
+      "learning_rate": 1.1970856226924432e-06,
+      "loss": 0.666,
+      "step": 15212
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.303695407467361,
+      "learning_rate": 1.1969909087025588e-06,
+      "loss": 0.6556,
+      "step": 15213
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.992518105907452,
+      "learning_rate": 1.1968961928741406e-06,
+      "loss": 0.6289,
+      "step": 15214
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.182609953750445,
+      "learning_rate": 1.196801475208073e-06,
+      "loss": 0.6471,
+      "step": 15215
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.7821587663170173,
+      "learning_rate": 1.19670675570524e-06,
+      "loss": 0.6543,
+      "step": 15216
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.8068582347477165,
+      "learning_rate": 1.196612034366525e-06,
+      "loss": 0.5964,
+      "step": 15217
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.073448620640498,
+      "learning_rate": 1.196517311192813e-06,
+      "loss": 0.6341,
+      "step": 15218
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.175520329184716,
+      "learning_rate": 1.1964225861849876e-06,
+      "loss": 0.6751,
+      "step": 15219
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.962204940353816,
+      "learning_rate": 1.1963278593439325e-06,
+      "loss": 0.694,
+      "step": 15220
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.9797780553637847,
+      "learning_rate": 1.1962331306705322e-06,
+      "loss": 0.6426,
+      "step": 15221
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.00014784390881,
+      "learning_rate": 1.196138400165671e-06,
+      "loss": 0.6159,
+      "step": 15222
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.972839734395516,
+      "learning_rate": 1.1960436678302327e-06,
+      "loss": 0.6797,
+      "step": 15223
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.9757010340516565,
+      "learning_rate": 1.1959489336651013e-06,
+      "loss": 0.625,
+      "step": 15224
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.366549185718354,
+      "learning_rate": 1.195854197671161e-06,
+      "loss": 0.6315,
+      "step": 15225
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.255186792161503,
+      "learning_rate": 1.1957594598492967e-06,
+      "loss": 0.6217,
+      "step": 15226
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.7101067043232336,
+      "learning_rate": 1.1956647202003914e-06,
+      "loss": 0.6348,
+      "step": 15227
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 5.4214442282388555,
+      "learning_rate": 1.1955699787253304e-06,
+      "loss": 0.6797,
+      "step": 15228
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.8076369555606413,
+      "learning_rate": 1.1954752354249976e-06,
+      "loss": 0.6335,
+      "step": 15229
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 7.341640349540955,
+      "learning_rate": 1.1953804903002768e-06,
+      "loss": 0.6478,
+      "step": 15230
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.0629830592459926,
+      "learning_rate": 1.1952857433520526e-06,
+      "loss": 0.6536,
+      "step": 15231
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.6037242433332084,
+      "learning_rate": 1.1951909945812096e-06,
+      "loss": 0.6426,
+      "step": 15232
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.766913213247672,
+      "learning_rate": 1.1950962439886315e-06,
+      "loss": 0.6699,
+      "step": 15233
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.399900999984255,
+      "learning_rate": 1.1950014915752031e-06,
+      "loss": 0.6328,
+      "step": 15234
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.086097865543194,
+      "learning_rate": 1.1949067373418083e-06,
+      "loss": 0.6191,
+      "step": 15235
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.3375481277348444,
+      "learning_rate": 1.1948119812893318e-06,
+      "loss": 0.6263,
+      "step": 15236
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.201622806870857,
+      "learning_rate": 1.194717223418658e-06,
+      "loss": 0.6257,
+      "step": 15237
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.3548750538927927,
+      "learning_rate": 1.1946224637306708e-06,
+      "loss": 0.623,
+      "step": 15238
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.2883317647606303,
+      "learning_rate": 1.1945277022262555e-06,
+      "loss": 0.6895,
+      "step": 15239
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.178215119965613,
+      "learning_rate": 1.1944329389062954e-06,
+      "loss": 0.6497,
+      "step": 15240
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.746648146495417,
+      "learning_rate": 1.1943381737716758e-06,
+      "loss": 0.7005,
+      "step": 15241
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.735903972122468,
+      "learning_rate": 1.1942434068232807e-06,
+      "loss": 0.6048,
+      "step": 15242
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.8724466454654713,
+      "learning_rate": 1.1941486380619947e-06,
+      "loss": 0.6204,
+      "step": 15243
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.834583132813355,
+      "learning_rate": 1.1940538674887024e-06,
+      "loss": 0.6198,
+      "step": 15244
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.655772992538555,
+      "learning_rate": 1.1939590951042878e-06,
+      "loss": 0.6484,
+      "step": 15245
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.911540428355448,
+      "learning_rate": 1.1938643209096364e-06,
+      "loss": 0.6484,
+      "step": 15246
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.630733391525835,
+      "learning_rate": 1.193769544905632e-06,
+      "loss": 0.6172,
+      "step": 15247
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 4.799190010819467,
+      "learning_rate": 1.1936747670931587e-06,
+      "loss": 0.6439,
+      "step": 15248
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.8300254831350737,
+      "learning_rate": 1.1935799874731024e-06,
+      "loss": 0.6413,
+      "step": 15249
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.3940247489742164,
+      "learning_rate": 1.1934852060463467e-06,
+      "loss": 0.6341,
+      "step": 15250
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.704972666189697,
+      "learning_rate": 1.1933904228137764e-06,
+      "loss": 0.6029,
+      "step": 15251
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.7197058144880097,
+      "learning_rate": 1.1932956377762767e-06,
+      "loss": 0.707,
+      "step": 15252
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.4373079073071806,
+      "learning_rate": 1.1932008509347311e-06,
+      "loss": 0.6855,
+      "step": 15253
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 7.091177659742945,
+      "learning_rate": 1.193106062290025e-06,
+      "loss": 0.6406,
+      "step": 15254
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.1661378215919775,
+      "learning_rate": 1.193011271843043e-06,
+      "loss": 0.6927,
+      "step": 15255
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.6305578442318422,
+      "learning_rate": 1.1929164795946696e-06,
+      "loss": 0.6146,
+      "step": 15256
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.581230184004895,
+      "learning_rate": 1.1928216855457897e-06,
+      "loss": 0.6055,
+      "step": 15257
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.1441006793821016,
+      "learning_rate": 1.192726889697288e-06,
+      "loss": 0.6296,
+      "step": 15258
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.815135157113227,
+      "learning_rate": 1.1926320920500492e-06,
+      "loss": 0.6432,
+      "step": 15259
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.125397619948988,
+      "learning_rate": 1.1925372926049581e-06,
+      "loss": 0.6738,
+      "step": 15260
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 6.004850708962287,
+      "learning_rate": 1.1924424913628995e-06,
+      "loss": 0.6419,
+      "step": 15261
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.64253137449058,
+      "learning_rate": 1.192347688324758e-06,
+      "loss": 0.6706,
+      "step": 15262
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.97798429032833,
+      "learning_rate": 1.1922528834914185e-06,
+      "loss": 0.6165,
+      "step": 15263
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.872111831993428,
+      "learning_rate": 1.1921580768637656e-06,
+      "loss": 0.6348,
+      "step": 15264
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.126925818790896,
+      "learning_rate": 1.1920632684426848e-06,
+      "loss": 0.6426,
+      "step": 15265
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.909633347444344,
+      "learning_rate": 1.1919684582290603e-06,
+      "loss": 0.6953,
+      "step": 15266
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 8.960953336937344,
+      "learning_rate": 1.191873646223777e-06,
+      "loss": 0.668,
+      "step": 15267
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.0755361168377133,
+      "learning_rate": 1.1917788324277206e-06,
+      "loss": 0.6484,
+      "step": 15268
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.159398379102945,
+      "learning_rate": 1.191684016841775e-06,
+      "loss": 0.6055,
+      "step": 15269
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.9964780369184143,
+      "learning_rate": 1.1915891994668253e-06,
+      "loss": 0.6745,
+      "step": 15270
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.652000565185574,
+      "learning_rate": 1.191494380303757e-06,
+      "loss": 0.6426,
+      "step": 15271
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.5305625747138367,
+      "learning_rate": 1.1913995593534547e-06,
+      "loss": 0.64,
+      "step": 15272
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.695738099020112,
+      "learning_rate": 1.1913047366168034e-06,
+      "loss": 0.6862,
+      "step": 15273
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.5487817117636196,
+      "learning_rate": 1.191209912094688e-06,
+      "loss": 0.6283,
+      "step": 15274
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.9334252410378925,
+      "learning_rate": 1.1911150857879934e-06,
+      "loss": 0.6387,
+      "step": 15275
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.673309010108496,
+      "learning_rate": 1.191020257697605e-06,
+      "loss": 0.6667,
+      "step": 15276
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.7478693429637437,
+      "learning_rate": 1.1909254278244073e-06,
+      "loss": 0.6387,
+      "step": 15277
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.289894156366346,
+      "learning_rate": 1.1908305961692864e-06,
+      "loss": 0.6322,
+      "step": 15278
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.43957211925086,
+      "learning_rate": 1.1907357627331263e-06,
+      "loss": 0.6725,
+      "step": 15279
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.4532447407706854,
+      "learning_rate": 1.1906409275168123e-06,
+      "loss": 0.6198,
+      "step": 15280
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.465377683445044,
+      "learning_rate": 1.19054609052123e-06,
+      "loss": 0.6354,
+      "step": 15281
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.659678742690491,
+      "learning_rate": 1.1904512517472639e-06,
+      "loss": 0.6315,
+      "step": 15282
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.746331127689491,
+      "learning_rate": 1.1903564111957993e-06,
+      "loss": 0.6777,
+      "step": 15283
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.4636574797522166,
+      "learning_rate": 1.1902615688677217e-06,
+      "loss": 0.6445,
+      "step": 15284
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.8521624473800657,
+      "learning_rate": 1.1901667247639162e-06,
+      "loss": 0.6354,
+      "step": 15285
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.5066822034664535,
+      "learning_rate": 1.1900718788852675e-06,
+      "loss": 0.6803,
+      "step": 15286
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.4697157968031074,
+      "learning_rate": 1.1899770312326614e-06,
+      "loss": 0.6471,
+      "step": 15287
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.475270817403986,
+      "learning_rate": 1.1898821818069828e-06,
+      "loss": 0.5898,
+      "step": 15288
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.3241181604763574,
+      "learning_rate": 1.189787330609117e-06,
+      "loss": 0.5768,
+      "step": 15289
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.3684493830404487,
+      "learning_rate": 1.1896924776399492e-06,
+      "loss": 0.6393,
+      "step": 15290
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.8066298649517223,
+      "learning_rate": 1.1895976229003645e-06,
+      "loss": 0.6243,
+      "step": 15291
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.559674144949094,
+      "learning_rate": 1.189502766391249e-06,
+      "loss": 0.6647,
+      "step": 15292
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.365362093310124,
+      "learning_rate": 1.189407908113487e-06,
+      "loss": 0.6686,
+      "step": 15293
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.6982166246342776,
+      "learning_rate": 1.1893130480679645e-06,
+      "loss": 0.6217,
+      "step": 15294
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.9791011633822997,
+      "learning_rate": 1.1892181862555667e-06,
+      "loss": 0.7012,
+      "step": 15295
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.1946986487529907,
+      "learning_rate": 1.1891233226771784e-06,
+      "loss": 0.6426,
+      "step": 15296
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.0115486791851573,
+      "learning_rate": 1.1890284573336854e-06,
+      "loss": 0.6829,
+      "step": 15297
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.084448131754785,
+      "learning_rate": 1.1889335902259736e-06,
+      "loss": 0.6641,
+      "step": 15298
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 5.040639015060798,
+      "learning_rate": 1.1888387213549273e-06,
+      "loss": 0.6484,
+      "step": 15299
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.5224942832341597,
+      "learning_rate": 1.188743850721433e-06,
+      "loss": 0.6094,
+      "step": 15300
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.0174482495537673,
+      "learning_rate": 1.188648978326375e-06,
+      "loss": 0.6107,
+      "step": 15301
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.8006417404912347,
+      "learning_rate": 1.1885541041706404e-06,
+      "loss": 0.6387,
+      "step": 15302
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.7561716066605735,
+      "learning_rate": 1.1884592282551128e-06,
+      "loss": 0.6595,
+      "step": 15303
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 4.018750692387599,
+      "learning_rate": 1.1883643505806788e-06,
+      "loss": 0.6732,
+      "step": 15304
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.1715276920801037,
+      "learning_rate": 1.188269471148224e-06,
+      "loss": 0.6335,
+      "step": 15305
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.8860244284324486,
+      "learning_rate": 1.1881745899586333e-06,
+      "loss": 0.6602,
+      "step": 15306
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.8092488272112908,
+      "learning_rate": 1.1880797070127926e-06,
+      "loss": 0.6673,
+      "step": 15307
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.613616563619354,
+      "learning_rate": 1.187984822311587e-06,
+      "loss": 0.6357,
+      "step": 15308
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.6306369681799926,
+      "learning_rate": 1.1878899358559028e-06,
+      "loss": 0.6569,
+      "step": 15309
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.5924581064073133,
+      "learning_rate": 1.1877950476466252e-06,
+      "loss": 0.6139,
+      "step": 15310
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.1924852864930364,
+      "learning_rate": 1.1877001576846395e-06,
+      "loss": 0.6393,
+      "step": 15311
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.0554842365802544,
+      "learning_rate": 1.187605265970832e-06,
+      "loss": 0.6556,
+      "step": 15312
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.4436107239282117,
+      "learning_rate": 1.1875103725060878e-06,
+      "loss": 0.6217,
+      "step": 15313
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.6811459397760706,
+      "learning_rate": 1.1874154772912927e-06,
+      "loss": 0.6556,
+      "step": 15314
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.8654668499969915,
+      "learning_rate": 1.1873205803273324e-06,
+      "loss": 0.6348,
+      "step": 15315
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.646247282372318,
+      "learning_rate": 1.1872256816150928e-06,
+      "loss": 0.6387,
+      "step": 15316
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.8249902320708435,
+      "learning_rate": 1.1871307811554587e-06,
+      "loss": 0.6465,
+      "step": 15317
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.530946961572825,
+      "learning_rate": 1.1870358789493174e-06,
+      "loss": 0.6426,
+      "step": 15318
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.292022148634445,
+      "learning_rate": 1.186940974997553e-06,
+      "loss": 0.6328,
+      "step": 15319
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 3.3856658504233303,
+      "learning_rate": 1.186846069301052e-06,
+      "loss": 0.6038,
+      "step": 15320
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.114017307336941,
+      "learning_rate": 1.1867511618607002e-06,
+      "loss": 0.6641,
+      "step": 15321
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.6919861347519367,
+      "learning_rate": 1.1866562526773836e-06,
+      "loss": 0.6868,
+      "step": 15322
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.7150638982505697,
+      "learning_rate": 1.1865613417519872e-06,
+      "loss": 0.6471,
+      "step": 15323
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.135790671411815,
+      "learning_rate": 1.1864664290853977e-06,
+      "loss": 0.6198,
+      "step": 15324
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.924749911334911,
+      "learning_rate": 1.1863715146785003e-06,
+      "loss": 0.6471,
+      "step": 15325
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.194978402250479,
+      "learning_rate": 1.1862765985321813e-06,
+      "loss": 0.6517,
+      "step": 15326
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.5468015321582675,
+      "learning_rate": 1.1861816806473259e-06,
+      "loss": 0.6315,
+      "step": 15327
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.333241547070015,
+      "learning_rate": 1.1860867610248207e-06,
+      "loss": 0.6227,
+      "step": 15328
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.223531866688676,
+      "learning_rate": 1.1859918396655514e-06,
+      "loss": 0.612,
+      "step": 15329
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.014945103757636,
+      "learning_rate": 1.1858969165704036e-06,
+      "loss": 0.6328,
+      "step": 15330
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.7858884916516793,
+      "learning_rate": 1.1858019917402635e-06,
+      "loss": 0.6211,
+      "step": 15331
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.6918263171431818,
+      "learning_rate": 1.185707065176017e-06,
+      "loss": 0.6237,
+      "step": 15332
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.1612155414833367,
+      "learning_rate": 1.1856121368785499e-06,
+      "loss": 0.6445,
+      "step": 15333
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.5009692060980484,
+      "learning_rate": 1.1855172068487487e-06,
+      "loss": 0.6465,
+      "step": 15334
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.258195465194532,
+      "learning_rate": 1.1854222750874986e-06,
+      "loss": 0.6777,
+      "step": 15335
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.912982876162719,
+      "learning_rate": 1.1853273415956862e-06,
+      "loss": 0.6595,
+      "step": 15336
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.282978408926545,
+      "learning_rate": 1.1852324063741973e-06,
+      "loss": 0.6413,
+      "step": 15337
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.6002560372724703,
+      "learning_rate": 1.185137469423918e-06,
+      "loss": 0.6074,
+      "step": 15338
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.2943744999203624,
+      "learning_rate": 1.1850425307457344e-06,
+      "loss": 0.6217,
+      "step": 15339
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.146761422414615,
+      "learning_rate": 1.1849475903405322e-06,
+      "loss": 0.5885,
+      "step": 15340
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.246137536774571,
+      "learning_rate": 1.1848526482091981e-06,
+      "loss": 0.5951,
+      "step": 15341
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.68543344424926,
+      "learning_rate": 1.1847577043526179e-06,
+      "loss": 0.666,
+      "step": 15342
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.8776134381632126,
+      "learning_rate": 1.1846627587716774e-06,
+      "loss": 0.7142,
+      "step": 15343
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.6641026895826414,
+      "learning_rate": 1.184567811467263e-06,
+      "loss": 0.6868,
+      "step": 15344
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 8.55902418471175,
+      "learning_rate": 1.1844728624402613e-06,
+      "loss": 0.6543,
+      "step": 15345
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.414726489290211,
+      "learning_rate": 1.1843779116915578e-06,
+      "loss": 0.7018,
+      "step": 15346
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.019146428733983,
+      "learning_rate": 1.1842829592220388e-06,
+      "loss": 0.6992,
+      "step": 15347
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.716288820714218,
+      "learning_rate": 1.184188005032591e-06,
+      "loss": 0.6393,
+      "step": 15348
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.7091253404153557,
+      "learning_rate": 1.1840930491241002e-06,
+      "loss": 0.6029,
+      "step": 15349
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0263634614648782,
+      "learning_rate": 1.1839980914974525e-06,
+      "loss": 0.6615,
+      "step": 15350
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.1177137471636707,
+      "learning_rate": 1.1839031321535344e-06,
+      "loss": 0.6829,
+      "step": 15351
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.710087097613653,
+      "learning_rate": 1.1838081710932322e-06,
+      "loss": 0.6302,
+      "step": 15352
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.781504549307216,
+      "learning_rate": 1.183713208317432e-06,
+      "loss": 0.6439,
+      "step": 15353
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.282174041144426,
+      "learning_rate": 1.18361824382702e-06,
+      "loss": 0.6615,
+      "step": 15354
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0499957529420643,
+      "learning_rate": 1.1835232776228833e-06,
+      "loss": 0.6452,
+      "step": 15355
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.867569584947434,
+      "learning_rate": 1.183428309705907e-06,
+      "loss": 0.6888,
+      "step": 15356
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.015541822897121,
+      "learning_rate": 1.1833333400769781e-06,
+      "loss": 0.5837,
+      "step": 15357
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0214348055265057,
+      "learning_rate": 1.1832383687369833e-06,
+      "loss": 0.6296,
+      "step": 15358
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.661427006823941,
+      "learning_rate": 1.1831433956868085e-06,
+      "loss": 0.6439,
+      "step": 15359
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 4.818658151249231,
+      "learning_rate": 1.1830484209273402e-06,
+      "loss": 0.6549,
+      "step": 15360
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.4477702485054103,
+      "learning_rate": 1.1829534444594646e-06,
+      "loss": 0.6354,
+      "step": 15361
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.309964886226435,
+      "learning_rate": 1.1828584662840684e-06,
+      "loss": 0.6432,
+      "step": 15362
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.570371419861426,
+      "learning_rate": 1.1827634864020379e-06,
+      "loss": 0.5977,
+      "step": 15363
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.997444859568011,
+      "learning_rate": 1.1826685048142598e-06,
+      "loss": 0.6289,
+      "step": 15364
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.29931093401615,
+      "learning_rate": 1.1825735215216201e-06,
+      "loss": 0.6602,
+      "step": 15365
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.836716885895707,
+      "learning_rate": 1.1824785365250058e-06,
+      "loss": 0.6465,
+      "step": 15366
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.1218756070407374,
+      "learning_rate": 1.1823835498253032e-06,
+      "loss": 0.6686,
+      "step": 15367
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.874276464453335,
+      "learning_rate": 1.1822885614233987e-06,
+      "loss": 0.6816,
+      "step": 15368
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.902337748410941,
+      "learning_rate": 1.1821935713201788e-06,
+      "loss": 0.7174,
+      "step": 15369
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0105352399090113,
+      "learning_rate": 1.1820985795165305e-06,
+      "loss": 0.6628,
+      "step": 15370
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 6.799473507269287,
+      "learning_rate": 1.18200358601334e-06,
+      "loss": 0.6387,
+      "step": 15371
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.186913563838606,
+      "learning_rate": 1.1819085908114939e-06,
+      "loss": 0.6986,
+      "step": 15372
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0357671733449334,
+      "learning_rate": 1.181813593911879e-06,
+      "loss": 0.6188,
+      "step": 15373
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.8754977182359007,
+      "learning_rate": 1.1817185953153813e-06,
+      "loss": 0.6302,
+      "step": 15374
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.9055020298622396,
+      "learning_rate": 1.181623595022888e-06,
+      "loss": 0.6686,
+      "step": 15375
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.315294194602545,
+      "learning_rate": 1.181528593035286e-06,
+      "loss": 0.6484,
+      "step": 15376
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.454353245784931,
+      "learning_rate": 1.1814335893534614e-06,
+      "loss": 0.639,
+      "step": 15377
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.9333042172137476,
+      "learning_rate": 1.1813385839783007e-06,
+      "loss": 0.6536,
+      "step": 15378
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.832360709348862,
+      "learning_rate": 1.1812435769106914e-06,
+      "loss": 0.6211,
+      "step": 15379
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.574916911103472,
+      "learning_rate": 1.1811485681515196e-06,
+      "loss": 0.6816,
+      "step": 15380
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.7446872752612337,
+      "learning_rate": 1.1810535577016723e-06,
+      "loss": 0.6777,
+      "step": 15381
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0083065596611562,
+      "learning_rate": 1.180958545562036e-06,
+      "loss": 0.6654,
+      "step": 15382
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.8683058216562713,
+      "learning_rate": 1.1808635317334977e-06,
+      "loss": 0.653,
+      "step": 15383
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.8424009070649974,
+      "learning_rate": 1.1807685162169443e-06,
+      "loss": 0.6738,
+      "step": 15384
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 3.0372284978722615,
+      "learning_rate": 1.1806734990132619e-06,
+      "loss": 0.6855,
+      "step": 15385
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.2786440424281036,
+      "learning_rate": 1.1805784801233376e-06,
+      "loss": 0.6471,
+      "step": 15386
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 5.2932168318700095,
+      "learning_rate": 1.180483459548059e-06,
+      "loss": 0.6191,
+      "step": 15387
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.464488084255283,
+      "learning_rate": 1.1803884372883118e-06,
+      "loss": 0.6523,
+      "step": 15388
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.9137432748834056,
+      "learning_rate": 1.1802934133449837e-06,
+      "loss": 0.627,
+      "step": 15389
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.782125624349598,
+      "learning_rate": 1.180198387718961e-06,
+      "loss": 0.638,
+      "step": 15390
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.8080451065056216,
+      "learning_rate": 1.180103360411131e-06,
+      "loss": 0.6263,
+      "step": 15391
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.437352617599294,
+      "learning_rate": 1.1800083314223804e-06,
+      "loss": 0.6849,
+      "step": 15392
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.205143590614912,
+      "learning_rate": 1.179913300753596e-06,
+      "loss": 0.6387,
+      "step": 15393
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.916474168807981,
+      "learning_rate": 1.1798182684056647e-06,
+      "loss": 0.668,
+      "step": 15394
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.2307830232053036,
+      "learning_rate": 1.179723234379474e-06,
+      "loss": 0.5996,
+      "step": 15395
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.367972674590326,
+      "learning_rate": 1.17962819867591e-06,
+      "loss": 0.6973,
+      "step": 15396
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.5595770284115105,
+      "learning_rate": 1.1795331612958606e-06,
+      "loss": 0.6309,
+      "step": 15397
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.6950754499905885,
+      "learning_rate": 1.179438122240212e-06,
+      "loss": 0.6602,
+      "step": 15398
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.3828844266695404,
+      "learning_rate": 1.1793430815098518e-06,
+      "loss": 0.6602,
+      "step": 15399
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.364764235659605,
+      "learning_rate": 1.1792480391056665e-06,
+      "loss": 0.6439,
+      "step": 15400
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.002878668054556,
+      "learning_rate": 1.1791529950285437e-06,
+      "loss": 0.6471,
+      "step": 15401
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.23352796075945,
+      "learning_rate": 1.1790579492793698e-06,
+      "loss": 0.6243,
+      "step": 15402
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.3025937402971444,
+      "learning_rate": 1.1789629018590323e-06,
+      "loss": 0.6569,
+      "step": 15403
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.84328866884314,
+      "learning_rate": 1.1788678527684187e-06,
+      "loss": 0.6699,
+      "step": 15404
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.687537262945748,
+      "learning_rate": 1.1787728020084151e-06,
+      "loss": 0.6374,
+      "step": 15405
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.1824994912515465,
+      "learning_rate": 1.1786777495799091e-06,
+      "loss": 0.668,
+      "step": 15406
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.513836333850433,
+      "learning_rate": 1.178582695483788e-06,
+      "loss": 0.6465,
+      "step": 15407
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.988531636818401,
+      "learning_rate": 1.178487639720939e-06,
+      "loss": 0.6165,
+      "step": 15408
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.403542112654442,
+      "learning_rate": 1.1783925822922486e-06,
+      "loss": 0.6771,
+      "step": 15409
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.01178822151872,
+      "learning_rate": 1.178297523198605e-06,
+      "loss": 0.6504,
+      "step": 15410
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.1033063046621012,
+      "learning_rate": 1.1782024624408947e-06,
+      "loss": 0.6126,
+      "step": 15411
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.9298033042823177,
+      "learning_rate": 1.1781074000200045e-06,
+      "loss": 0.6263,
+      "step": 15412
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.753801257536937,
+      "learning_rate": 1.1780123359368228e-06,
+      "loss": 0.6589,
+      "step": 15413
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.285497506059499,
+      "learning_rate": 1.1779172701922362e-06,
+      "loss": 0.6419,
+      "step": 15414
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.04635384870484,
+      "learning_rate": 1.1778222027871317e-06,
+      "loss": 0.6641,
+      "step": 15415
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.8725919902389436,
+      "learning_rate": 1.1777271337223972e-06,
+      "loss": 0.6361,
+      "step": 15416
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.7073717401935706,
+      "learning_rate": 1.1776320629989193e-06,
+      "loss": 0.6133,
+      "step": 15417
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.8855323651212457,
+      "learning_rate": 1.1775369906175858e-06,
+      "loss": 0.6849,
+      "step": 15418
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.952697693072487,
+      "learning_rate": 1.1774419165792839e-06,
+      "loss": 0.6634,
+      "step": 15419
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.993016688956522,
+      "learning_rate": 1.1773468408849008e-06,
+      "loss": 0.6562,
+      "step": 15420
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.5774171985089676,
+      "learning_rate": 1.1772517635353242e-06,
+      "loss": 0.6478,
+      "step": 15421
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.294351929014379,
+      "learning_rate": 1.1771566845314409e-06,
+      "loss": 0.6426,
+      "step": 15422
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.5999585534949476,
+      "learning_rate": 1.1770616038741385e-06,
+      "loss": 0.6471,
+      "step": 15423
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.5877800973572,
+      "learning_rate": 1.176966521564305e-06,
+      "loss": 0.5944,
+      "step": 15424
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.0322794995064415,
+      "learning_rate": 1.1768714376028267e-06,
+      "loss": 0.6517,
+      "step": 15425
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.907867429032107,
+      "learning_rate": 1.1767763519905923e-06,
+      "loss": 0.6217,
+      "step": 15426
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.54199149027121,
+      "learning_rate": 1.1766812647284883e-06,
+      "loss": 0.6328,
+      "step": 15427
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.9065285286687748,
+      "learning_rate": 1.176586175817402e-06,
+      "loss": 0.6523,
+      "step": 15428
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.7857983718681005,
+      "learning_rate": 1.1764910852582216e-06,
+      "loss": 0.6725,
+      "step": 15429
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.8814630141314286,
+      "learning_rate": 1.1763959930518344e-06,
+      "loss": 0.6628,
+      "step": 15430
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.8826969138044194,
+      "learning_rate": 1.176300899199128e-06,
+      "loss": 0.6536,
+      "step": 15431
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 6.671325362038667,
+      "learning_rate": 1.1762058037009895e-06,
+      "loss": 0.6393,
+      "step": 15432
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.832396643313422,
+      "learning_rate": 1.1761107065583062e-06,
+      "loss": 0.6393,
+      "step": 15433
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.2347380690248917,
+      "learning_rate": 1.176015607771967e-06,
+      "loss": 0.6165,
+      "step": 15434
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.8746566808696707,
+      "learning_rate": 1.175920507342858e-06,
+      "loss": 0.5885,
+      "step": 15435
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.7039424680148922,
+      "learning_rate": 1.1758254052718673e-06,
+      "loss": 0.597,
+      "step": 15436
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.0287702727245787,
+      "learning_rate": 1.1757303015598828e-06,
+      "loss": 0.5892,
+      "step": 15437
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.1159395395947005,
+      "learning_rate": 1.1756351962077916e-06,
+      "loss": 0.6328,
+      "step": 15438
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.042057127649672,
+      "learning_rate": 1.1755400892164818e-06,
+      "loss": 0.6367,
+      "step": 15439
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.91018962026587,
+      "learning_rate": 1.1754449805868406e-06,
+      "loss": 0.6348,
+      "step": 15440
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.7631203924966545,
+      "learning_rate": 1.175349870319756e-06,
+      "loss": 0.6159,
+      "step": 15441
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.760893129383927,
+      "learning_rate": 1.1752547584161157e-06,
+      "loss": 0.6328,
+      "step": 15442
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.0022891084486565,
+      "learning_rate": 1.175159644876807e-06,
+      "loss": 0.6595,
+      "step": 15443
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.8217458880196804,
+      "learning_rate": 1.175064529702718e-06,
+      "loss": 0.6048,
+      "step": 15444
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.132607355803799,
+      "learning_rate": 1.1749694128947363e-06,
+      "loss": 0.6243,
+      "step": 15445
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.375846434958951,
+      "learning_rate": 1.1748742944537493e-06,
+      "loss": 0.6647,
+      "step": 15446
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.0564460702549834,
+      "learning_rate": 1.1747791743806454e-06,
+      "loss": 0.6367,
+      "step": 15447
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.2609243865723143,
+      "learning_rate": 1.1746840526763117e-06,
+      "loss": 0.6201,
+      "step": 15448
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.258018316293525,
+      "learning_rate": 1.1745889293416365e-06,
+      "loss": 0.6497,
+      "step": 15449
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 7.3384193611253945,
+      "learning_rate": 1.1744938043775077e-06,
+      "loss": 0.6543,
+      "step": 15450
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.321514430176881,
+      "learning_rate": 1.1743986777848123e-06,
+      "loss": 0.6276,
+      "step": 15451
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.4428253017393633,
+      "learning_rate": 1.1743035495644384e-06,
+      "loss": 0.6562,
+      "step": 15452
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.9541424558552647,
+      "learning_rate": 1.1742084197172746e-06,
+      "loss": 0.6647,
+      "step": 15453
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 5.196268565928766,
+      "learning_rate": 1.174113288244208e-06,
+      "loss": 0.6947,
+      "step": 15454
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 3.484118370589236,
+      "learning_rate": 1.1740181551461267e-06,
+      "loss": 0.6328,
+      "step": 15455
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.230788896436972,
+      "learning_rate": 1.1739230204239186e-06,
+      "loss": 0.6432,
+      "step": 15456
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.106460364882983,
+      "learning_rate": 1.1738278840784715e-06,
+      "loss": 0.6973,
+      "step": 15457
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.743346797243414,
+      "learning_rate": 1.1737327461106734e-06,
+      "loss": 0.61,
+      "step": 15458
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.724314403310257,
+      "learning_rate": 1.1736376065214122e-06,
+      "loss": 0.6426,
+      "step": 15459
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.882968850265827,
+      "learning_rate": 1.173542465311576e-06,
+      "loss": 0.6113,
+      "step": 15460
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.927350889840967,
+      "learning_rate": 1.1734473224820524e-06,
+      "loss": 0.6732,
+      "step": 15461
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.176851103848874,
+      "learning_rate": 1.1733521780337297e-06,
+      "loss": 0.6484,
+      "step": 15462
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.693521513836052,
+      "learning_rate": 1.173257031967496e-06,
+      "loss": 0.6335,
+      "step": 15463
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.9590005769383567,
+      "learning_rate": 1.1731618842842387e-06,
+      "loss": 0.6338,
+      "step": 15464
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.8308126867074184,
+      "learning_rate": 1.1730667349848463e-06,
+      "loss": 0.6901,
+      "step": 15465
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.48393011368664,
+      "learning_rate": 1.1729715840702073e-06,
+      "loss": 0.6257,
+      "step": 15466
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.9455679699142525,
+      "learning_rate": 1.1728764315412087e-06,
+      "loss": 0.6406,
+      "step": 15467
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.161026882708681,
+      "learning_rate": 1.172781277398739e-06,
+      "loss": 0.6165,
+      "step": 15468
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.56991419860545,
+      "learning_rate": 1.1726861216436868e-06,
+      "loss": 0.6309,
+      "step": 15469
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.389968274317099,
+      "learning_rate": 1.1725909642769394e-06,
+      "loss": 0.6751,
+      "step": 15470
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.0006586290534876,
+      "learning_rate": 1.1724958052993854e-06,
+      "loss": 0.653,
+      "step": 15471
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.0287740796383815,
+      "learning_rate": 1.1724006447119126e-06,
+      "loss": 0.64,
+      "step": 15472
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.4435234233445846,
+      "learning_rate": 1.1723054825154093e-06,
+      "loss": 0.6523,
+      "step": 15473
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.751887971973845,
+      "learning_rate": 1.1722103187107641e-06,
+      "loss": 0.6296,
+      "step": 15474
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.0090017902107324,
+      "learning_rate": 1.1721151532988645e-06,
+      "loss": 0.6224,
+      "step": 15475
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.8806217625694495,
+      "learning_rate": 1.172019986280599e-06,
+      "loss": 0.6074,
+      "step": 15476
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.414870966663561,
+      "learning_rate": 1.171924817656856e-06,
+      "loss": 0.6081,
+      "step": 15477
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.3629909844766335,
+      "learning_rate": 1.1718296474285229e-06,
+      "loss": 0.6309,
+      "step": 15478
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.8924512222840546,
+      "learning_rate": 1.171734475596489e-06,
+      "loss": 0.6556,
+      "step": 15479
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.6919058259567072,
+      "learning_rate": 1.171639302161642e-06,
+      "loss": 0.6439,
+      "step": 15480
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.7715152631485616,
+      "learning_rate": 1.1715441271248698e-06,
+      "loss": 0.651,
+      "step": 15481
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.049036182715633,
+      "learning_rate": 1.1714489504870615e-06,
+      "loss": 0.6647,
+      "step": 15482
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.472514428800762,
+      "learning_rate": 1.171353772249105e-06,
+      "loss": 0.6712,
+      "step": 15483
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.2189506839286777,
+      "learning_rate": 1.1712585924118882e-06,
+      "loss": 0.6393,
+      "step": 15484
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.9738768047646933,
+      "learning_rate": 1.1711634109763003e-06,
+      "loss": 0.651,
+      "step": 15485
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.573603000966073,
+      "learning_rate": 1.1710682279432287e-06,
+      "loss": 0.6354,
+      "step": 15486
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.8074454794794246,
+      "learning_rate": 1.1709730433135627e-06,
+      "loss": 0.6517,
+      "step": 15487
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.922672808804402,
+      "learning_rate": 1.1708778570881898e-06,
+      "loss": 0.5794,
+      "step": 15488
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 6.3444267193281645,
+      "learning_rate": 1.1707826692679987e-06,
+      "loss": 0.5918,
+      "step": 15489
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.9984370756799565,
+      "learning_rate": 1.1706874798538783e-06,
+      "loss": 0.6152,
+      "step": 15490
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.4442963800713957,
+      "learning_rate": 1.1705922888467162e-06,
+      "loss": 0.5999,
+      "step": 15491
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.2310296751556047,
+      "learning_rate": 1.1704970962474013e-06,
+      "loss": 0.5931,
+      "step": 15492
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.2367750855569475,
+      "learning_rate": 1.1704019020568217e-06,
+      "loss": 0.6693,
+      "step": 15493
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.0775494656014493,
+      "learning_rate": 1.1703067062758662e-06,
+      "loss": 0.6855,
+      "step": 15494
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.031575081423354,
+      "learning_rate": 1.1702115089054232e-06,
+      "loss": 0.6589,
+      "step": 15495
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.045830479421231,
+      "learning_rate": 1.1701163099463808e-06,
+      "loss": 0.6257,
+      "step": 15496
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.726510828556007,
+      "learning_rate": 1.1700211093996285e-06,
+      "loss": 0.6536,
+      "step": 15497
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.140346782384717,
+      "learning_rate": 1.1699259072660538e-06,
+      "loss": 0.6257,
+      "step": 15498
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.337720149252819,
+      "learning_rate": 1.1698307035465454e-06,
+      "loss": 0.6693,
+      "step": 15499
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.246048175439891,
+      "learning_rate": 1.1697354982419925e-06,
+      "loss": 0.6693,
+      "step": 15500
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.6059651655418135,
+      "learning_rate": 1.1696402913532825e-06,
+      "loss": 0.6621,
+      "step": 15501
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.675183040373272,
+      "learning_rate": 1.169545082881305e-06,
+      "loss": 0.5951,
+      "step": 15502
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.6616752976925437,
+      "learning_rate": 1.1694498728269486e-06,
+      "loss": 0.7025,
+      "step": 15503
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.501290624343809,
+      "learning_rate": 1.1693546611911012e-06,
+      "loss": 0.64,
+      "step": 15504
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.7217990003250576,
+      "learning_rate": 1.1692594479746517e-06,
+      "loss": 0.6191,
+      "step": 15505
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.4298494583370593,
+      "learning_rate": 1.1691642331784888e-06,
+      "loss": 0.638,
+      "step": 15506
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 5.822439304104232,
+      "learning_rate": 1.1690690168035014e-06,
+      "loss": 0.6895,
+      "step": 15507
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.9421283983760054,
+      "learning_rate": 1.168973798850578e-06,
+      "loss": 0.6751,
+      "step": 15508
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.665911864137881,
+      "learning_rate": 1.1688785793206067e-06,
+      "loss": 0.6126,
+      "step": 15509
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.1192013257297804,
+      "learning_rate": 1.1687833582144771e-06,
+      "loss": 0.6172,
+      "step": 15510
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.938957873641972,
+      "learning_rate": 1.1686881355330774e-06,
+      "loss": 0.6829,
+      "step": 15511
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.560910993971844,
+      "learning_rate": 1.168592911277296e-06,
+      "loss": 0.627,
+      "step": 15512
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.943589296764702,
+      "learning_rate": 1.1684976854480227e-06,
+      "loss": 0.6204,
+      "step": 15513
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.8073656152004163,
+      "learning_rate": 1.1684024580461454e-06,
+      "loss": 0.638,
+      "step": 15514
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 4.649642800447774,
+      "learning_rate": 1.1683072290725526e-06,
+      "loss": 0.6562,
+      "step": 15515
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.9396998391144638,
+      "learning_rate": 1.1682119985281341e-06,
+      "loss": 0.6432,
+      "step": 15516
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.679680255480015,
+      "learning_rate": 1.168116766413778e-06,
+      "loss": 0.6543,
+      "step": 15517
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.2811097490483783,
+      "learning_rate": 1.168021532730373e-06,
+      "loss": 0.6185,
+      "step": 15518
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.8066066141267507,
+      "learning_rate": 1.1679262974788085e-06,
+      "loss": 0.6426,
+      "step": 15519
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.1350179320591036,
+      "learning_rate": 1.167831060659973e-06,
+      "loss": 0.625,
+      "step": 15520
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 3.247844943453554,
+      "learning_rate": 1.1677358222747552e-06,
+      "loss": 0.6211,
+      "step": 15521
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.5085084540237856,
+      "learning_rate": 1.1676405823240442e-06,
+      "loss": 0.5804,
+      "step": 15522
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.258447201046247,
+      "learning_rate": 1.1675453408087288e-06,
+      "loss": 0.6641,
+      "step": 15523
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1444944790007865,
+      "learning_rate": 1.167450097729698e-06,
+      "loss": 0.6133,
+      "step": 15524
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.088613821124058,
+      "learning_rate": 1.1673548530878406e-06,
+      "loss": 0.6484,
+      "step": 15525
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.8666406753812836,
+      "learning_rate": 1.1672596068840454e-06,
+      "loss": 0.6523,
+      "step": 15526
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.751503879782037,
+      "learning_rate": 1.1671643591192017e-06,
+      "loss": 0.6641,
+      "step": 15527
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.151236289057366,
+      "learning_rate": 1.1670691097941983e-06,
+      "loss": 0.6862,
+      "step": 15528
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1323269451381655,
+      "learning_rate": 1.1669738589099241e-06,
+      "loss": 0.6608,
+      "step": 15529
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.4014426526516273,
+      "learning_rate": 1.166878606467268e-06,
+      "loss": 0.6074,
+      "step": 15530
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.123862170943287,
+      "learning_rate": 1.166783352467119e-06,
+      "loss": 0.6243,
+      "step": 15531
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.517404568168167,
+      "learning_rate": 1.1666880969103662e-06,
+      "loss": 0.6309,
+      "step": 15532
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.675660656259796,
+      "learning_rate": 1.166592839797899e-06,
+      "loss": 0.5788,
+      "step": 15533
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.3638448668149223,
+      "learning_rate": 1.1664975811306057e-06,
+      "loss": 0.6699,
+      "step": 15534
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.2382629343139366,
+      "learning_rate": 1.1664023209093762e-06,
+      "loss": 0.6979,
+      "step": 15535
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.9252928467745845,
+      "learning_rate": 1.1663070591350988e-06,
+      "loss": 0.6328,
+      "step": 15536
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.80395563727962,
+      "learning_rate": 1.166211795808663e-06,
+      "loss": 0.6719,
+      "step": 15537
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.2938944414883435,
+      "learning_rate": 1.1661165309309576e-06,
+      "loss": 0.6745,
+      "step": 15538
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1276305658881394,
+      "learning_rate": 1.166021264502872e-06,
+      "loss": 0.5996,
+      "step": 15539
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.0648572318353615,
+      "learning_rate": 1.1659259965252957e-06,
+      "loss": 0.6113,
+      "step": 15540
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 6.3983097620726745,
+      "learning_rate": 1.165830726999117e-06,
+      "loss": 0.6654,
+      "step": 15541
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.7023817191536565,
+      "learning_rate": 1.1657354559252254e-06,
+      "loss": 0.6094,
+      "step": 15542
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1002199208157166,
+      "learning_rate": 1.1656401833045103e-06,
+      "loss": 0.6146,
+      "step": 15543
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.295983172922668,
+      "learning_rate": 1.1655449091378606e-06,
+      "loss": 0.6536,
+      "step": 15544
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.4651236446571114,
+      "learning_rate": 1.1654496334261658e-06,
+      "loss": 0.6191,
+      "step": 15545
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.2483468223097707,
+      "learning_rate": 1.165354356170315e-06,
+      "loss": 0.6484,
+      "step": 15546
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.6740345142175337,
+      "learning_rate": 1.165259077371197e-06,
+      "loss": 0.6504,
+      "step": 15547
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.295642863772294,
+      "learning_rate": 1.1651637970297018e-06,
+      "loss": 0.6276,
+      "step": 15548
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.1943809181182035,
+      "learning_rate": 1.1650685151467179e-06,
+      "loss": 0.6217,
+      "step": 15549
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.318611490173517,
+      "learning_rate": 1.1649732317231352e-06,
+      "loss": 0.6094,
+      "step": 15550
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.729747485937904,
+      "learning_rate": 1.1648779467598427e-06,
+      "loss": 0.6413,
+      "step": 15551
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.6563331232219802,
+      "learning_rate": 1.1647826602577295e-06,
+      "loss": 0.6283,
+      "step": 15552
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.446404825702435,
+      "learning_rate": 1.1646873722176858e-06,
+      "loss": 0.6133,
+      "step": 15553
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.139237756373081,
+      "learning_rate": 1.1645920826405997e-06,
+      "loss": 0.6283,
+      "step": 15554
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.749708012756754,
+      "learning_rate": 1.1644967915273612e-06,
+      "loss": 0.6497,
+      "step": 15555
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.817155826265496,
+      "learning_rate": 1.16440149887886e-06,
+      "loss": 0.6113,
+      "step": 15556
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.260749637770168,
+      "learning_rate": 1.1643062046959846e-06,
+      "loss": 0.5977,
+      "step": 15557
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.2976235814094954,
+      "learning_rate": 1.1642109089796253e-06,
+      "loss": 0.6536,
+      "step": 15558
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.8065925816809925,
+      "learning_rate": 1.1641156117306707e-06,
+      "loss": 0.6322,
+      "step": 15559
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.71955516292312,
+      "learning_rate": 1.1640203129500105e-06,
+      "loss": 0.6562,
+      "step": 15560
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.983268399630613,
+      "learning_rate": 1.1639250126385344e-06,
+      "loss": 0.6491,
+      "step": 15561
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.285562912202388,
+      "learning_rate": 1.1638297107971319e-06,
+      "loss": 0.6289,
+      "step": 15562
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.0983879041379407,
+      "learning_rate": 1.163734407426692e-06,
+      "loss": 0.6491,
+      "step": 15563
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.505653376066035,
+      "learning_rate": 1.1636391025281042e-06,
+      "loss": 0.6393,
+      "step": 15564
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.846386766454327,
+      "learning_rate": 1.1635437961022585e-06,
+      "loss": 0.625,
+      "step": 15565
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.119367151518711,
+      "learning_rate": 1.1634484881500441e-06,
+      "loss": 0.6419,
+      "step": 15566
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.219266549804639,
+      "learning_rate": 1.1633531786723503e-06,
+      "loss": 0.6299,
+      "step": 15567
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.631615346310551,
+      "learning_rate": 1.1632578676700667e-06,
+      "loss": 0.6315,
+      "step": 15568
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.0080077523796347,
+      "learning_rate": 1.1631625551440836e-06,
+      "loss": 0.6058,
+      "step": 15569
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.4678097398228678,
+      "learning_rate": 1.1630672410952893e-06,
+      "loss": 0.6243,
+      "step": 15570
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.2316637504737775,
+      "learning_rate": 1.1629719255245743e-06,
+      "loss": 0.6562,
+      "step": 15571
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.141131394938036,
+      "learning_rate": 1.1628766084328277e-06,
+      "loss": 0.6966,
+      "step": 15572
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.7661529759220107,
+      "learning_rate": 1.1627812898209396e-06,
+      "loss": 0.6348,
+      "step": 15573
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.3430642367417764,
+      "learning_rate": 1.1626859696897992e-06,
+      "loss": 0.6452,
+      "step": 15574
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.465673458185593,
+      "learning_rate": 1.162590648040296e-06,
+      "loss": 0.6471,
+      "step": 15575
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.871871662272028,
+      "learning_rate": 1.1624953248733203e-06,
+      "loss": 0.6107,
+      "step": 15576
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1392768405231335,
+      "learning_rate": 1.1624000001897613e-06,
+      "loss": 0.6146,
+      "step": 15577
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.2470991214754803,
+      "learning_rate": 1.1623046739905085e-06,
+      "loss": 0.6484,
+      "step": 15578
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.1513065543568706,
+      "learning_rate": 1.1622093462764522e-06,
+      "loss": 0.6315,
+      "step": 15579
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 10.570859081691436,
+      "learning_rate": 1.1621140170484816e-06,
+      "loss": 0.653,
+      "step": 15580
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.66029543468407,
+      "learning_rate": 1.1620186863074862e-06,
+      "loss": 0.6484,
+      "step": 15581
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.9689065347664765,
+      "learning_rate": 1.1619233540543563e-06,
+      "loss": 0.6647,
+      "step": 15582
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.689593929012432,
+      "learning_rate": 1.1618280202899816e-06,
+      "loss": 0.6523,
+      "step": 15583
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.233298754882079,
+      "learning_rate": 1.1617326850152514e-06,
+      "loss": 0.6406,
+      "step": 15584
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 4.349632281746875,
+      "learning_rate": 1.1616373482310557e-06,
+      "loss": 0.6289,
+      "step": 15585
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.3705331817813007,
+      "learning_rate": 1.1615420099382847e-06,
+      "loss": 0.6719,
+      "step": 15586
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.845510109952966,
+      "learning_rate": 1.1614466701378276e-06,
+      "loss": 0.6484,
+      "step": 15587
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 3.107643943422341,
+      "learning_rate": 1.1613513288305745e-06,
+      "loss": 0.5947,
+      "step": 15588
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.600539535306128,
+      "learning_rate": 1.1612559860174152e-06,
+      "loss": 0.6335,
+      "step": 15589
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.371476311086528,
+      "learning_rate": 1.1611606416992395e-06,
+      "loss": 0.6576,
+      "step": 15590
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.769747930814936,
+      "learning_rate": 1.1610652958769372e-06,
+      "loss": 0.6621,
+      "step": 15591
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.6376827644893406,
+      "learning_rate": 1.1609699485513981e-06,
+      "loss": 0.5807,
+      "step": 15592
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.336419585603276,
+      "learning_rate": 1.1608745997235127e-06,
+      "loss": 0.6497,
+      "step": 15593
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.5382980237935566,
+      "learning_rate": 1.16077924939417e-06,
+      "loss": 0.6504,
+      "step": 15594
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.631646277246079,
+      "learning_rate": 1.1606838975642606e-06,
+      "loss": 0.6139,
+      "step": 15595
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.067400721662484,
+      "learning_rate": 1.1605885442346742e-06,
+      "loss": 0.6276,
+      "step": 15596
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.7316451112522113,
+      "learning_rate": 1.1604931894063004e-06,
+      "loss": 0.6699,
+      "step": 15597
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.00377917373943,
+      "learning_rate": 1.1603978330800296e-06,
+      "loss": 0.7057,
+      "step": 15598
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.730310796217601,
+      "learning_rate": 1.160302475256752e-06,
+      "loss": 0.6341,
+      "step": 15599
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.1020737462068233,
+      "learning_rate": 1.1602071159373567e-06,
+      "loss": 0.6185,
+      "step": 15600
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.075294713224141,
+      "learning_rate": 1.1601117551227345e-06,
+      "loss": 0.6484,
+      "step": 15601
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.3934602368782367,
+      "learning_rate": 1.1600163928137751e-06,
+      "loss": 0.6589,
+      "step": 15602
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.203728038372765,
+      "learning_rate": 1.1599210290113687e-06,
+      "loss": 0.6882,
+      "step": 15603
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.7223471556472805,
+      "learning_rate": 1.159825663716405e-06,
+      "loss": 0.6452,
+      "step": 15604
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.7603236683477403,
+      "learning_rate": 1.1597302969297742e-06,
+      "loss": 0.6341,
+      "step": 15605
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.788121241915303,
+      "learning_rate": 1.1596349286523665e-06,
+      "loss": 0.638,
+      "step": 15606
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.245162280728777,
+      "learning_rate": 1.1595395588850717e-06,
+      "loss": 0.6628,
+      "step": 15607
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.3133875075271324,
+      "learning_rate": 1.1594441876287807e-06,
+      "loss": 0.6696,
+      "step": 15608
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.974308015473831,
+      "learning_rate": 1.1593488148843825e-06,
+      "loss": 0.6966,
+      "step": 15609
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 8.584846437903323,
+      "learning_rate": 1.1592534406527676e-06,
+      "loss": 0.6621,
+      "step": 15610
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.5501152317537805,
+      "learning_rate": 1.1591580649348264e-06,
+      "loss": 0.6699,
+      "step": 15611
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.6981541508733464,
+      "learning_rate": 1.159062687731449e-06,
+      "loss": 0.6159,
+      "step": 15612
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.4257401497322677,
+      "learning_rate": 1.1589673090435256e-06,
+      "loss": 0.6361,
+      "step": 15613
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 5.967355384353739,
+      "learning_rate": 1.158871928871946e-06,
+      "loss": 0.6237,
+      "step": 15614
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.4415263001969016,
+      "learning_rate": 1.1587765472176006e-06,
+      "loss": 0.6185,
+      "step": 15615
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.8595726547687574,
+      "learning_rate": 1.1586811640813798e-06,
+      "loss": 0.6693,
+      "step": 15616
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.4679009405122,
+      "learning_rate": 1.1585857794641736e-06,
+      "loss": 0.7253,
+      "step": 15617
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.8811675496205575,
+      "learning_rate": 1.1584903933668722e-06,
+      "loss": 0.6302,
+      "step": 15618
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.7362368927815135,
+      "learning_rate": 1.1583950057903664e-06,
+      "loss": 0.6328,
+      "step": 15619
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.3370949878375744,
+      "learning_rate": 1.1582996167355455e-06,
+      "loss": 0.6217,
+      "step": 15620
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.4475800636037865,
+      "learning_rate": 1.1582042262033003e-06,
+      "loss": 0.6341,
+      "step": 15621
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.6174637070290188,
+      "learning_rate": 1.1581088341945214e-06,
+      "loss": 0.6777,
+      "step": 15622
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.2578410064514003,
+      "learning_rate": 1.1580134407100984e-06,
+      "loss": 0.6595,
+      "step": 15623
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.9716912787785574,
+      "learning_rate": 1.1579180457509223e-06,
+      "loss": 0.6263,
+      "step": 15624
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.653961442170425,
+      "learning_rate": 1.1578226493178828e-06,
+      "loss": 0.6452,
+      "step": 15625
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.759931779833622,
+      "learning_rate": 1.1577272514118708e-06,
+      "loss": 0.6595,
+      "step": 15626
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.451458423763547,
+      "learning_rate": 1.1576318520337764e-06,
+      "loss": 0.6107,
+      "step": 15627
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.206193281676228,
+      "learning_rate": 1.1575364511844898e-06,
+      "loss": 0.6497,
+      "step": 15628
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.0345320646376077,
+      "learning_rate": 1.1574410488649017e-06,
+      "loss": 0.638,
+      "step": 15629
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.497763921651016,
+      "learning_rate": 1.1573456450759024e-06,
+      "loss": 0.6289,
+      "step": 15630
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.7062004271982967,
+      "learning_rate": 1.1572502398183822e-06,
+      "loss": 0.6686,
+      "step": 15631
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.5192676974779418,
+      "learning_rate": 1.157154833093232e-06,
+      "loss": 0.6543,
+      "step": 15632
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.9850753587083965,
+      "learning_rate": 1.1570594249013412e-06,
+      "loss": 0.6882,
+      "step": 15633
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.9983276751449255,
+      "learning_rate": 1.156964015243601e-06,
+      "loss": 0.6354,
+      "step": 15634
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.6448652632290317,
+      "learning_rate": 1.1568686041209024e-06,
+      "loss": 0.625,
+      "step": 15635
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.051479581226954,
+      "learning_rate": 1.1567731915341345e-06,
+      "loss": 0.6243,
+      "step": 15636
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.6207315568241056,
+      "learning_rate": 1.1566777774841888e-06,
+      "loss": 0.6406,
+      "step": 15637
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.248444891576933,
+      "learning_rate": 1.1565823619719554e-06,
+      "loss": 0.6426,
+      "step": 15638
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.705995412343959,
+      "learning_rate": 1.1564869449983255e-06,
+      "loss": 0.6354,
+      "step": 15639
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.708449692597205,
+      "learning_rate": 1.1563915265641886e-06,
+      "loss": 0.6536,
+      "step": 15640
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.642969487717022,
+      "learning_rate": 1.1562961066704358e-06,
+      "loss": 0.6517,
+      "step": 15641
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.582060617689002,
+      "learning_rate": 1.1562006853179578e-06,
+      "loss": 0.6029,
+      "step": 15642
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.5542440686431758,
+      "learning_rate": 1.1561052625076451e-06,
+      "loss": 0.6406,
+      "step": 15643
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.920405351416478,
+      "learning_rate": 1.1560098382403879e-06,
+      "loss": 0.6159,
+      "step": 15644
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.995526949334935,
+      "learning_rate": 1.155914412517077e-06,
+      "loss": 0.668,
+      "step": 15645
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.6541117291788017,
+      "learning_rate": 1.1558189853386033e-06,
+      "loss": 0.6322,
+      "step": 15646
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 7.68065223401273,
+      "learning_rate": 1.1557235567058572e-06,
+      "loss": 0.6354,
+      "step": 15647
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.2987030323354283,
+      "learning_rate": 1.1556281266197295e-06,
+      "loss": 0.6758,
+      "step": 15648
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.810645808808605,
+      "learning_rate": 1.1555326950811107e-06,
+      "loss": 0.5905,
+      "step": 15649
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.506281007939244,
+      "learning_rate": 1.1554372620908911e-06,
+      "loss": 0.6602,
+      "step": 15650
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.337397709285941,
+      "learning_rate": 1.1553418276499622e-06,
+      "loss": 0.6312,
+      "step": 15651
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.1026690917465958,
+      "learning_rate": 1.1552463917592143e-06,
+      "loss": 0.6611,
+      "step": 15652
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.1482640763510736,
+      "learning_rate": 1.1551509544195379e-06,
+      "loss": 0.6361,
+      "step": 15653
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.9648057578641693,
+      "learning_rate": 1.155055515631824e-06,
+      "loss": 0.679,
+      "step": 15654
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 3.125424422118839,
+      "learning_rate": 1.1549600753969633e-06,
+      "loss": 0.597,
+      "step": 15655
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 4.416872396725043,
+      "learning_rate": 1.1548646337158465e-06,
+      "loss": 0.6914,
+      "step": 15656
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.731814983181606,
+      "learning_rate": 1.1547691905893642e-06,
+      "loss": 0.6413,
+      "step": 15657
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.667004047656661,
+      "learning_rate": 1.1546737460184075e-06,
+      "loss": 0.6966,
+      "step": 15658
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.1084580879960244,
+      "learning_rate": 1.1545783000038671e-06,
+      "loss": 0.6068,
+      "step": 15659
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.448179718986582,
+      "learning_rate": 1.1544828525466337e-06,
+      "loss": 0.638,
+      "step": 15660
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.7820270984941575,
+      "learning_rate": 1.1543874036475983e-06,
+      "loss": 0.627,
+      "step": 15661
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 8.515203991664343,
+      "learning_rate": 1.1542919533076513e-06,
+      "loss": 0.6816,
+      "step": 15662
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.338987374519392,
+      "learning_rate": 1.154196501527684e-06,
+      "loss": 0.6855,
+      "step": 15663
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.4597504000458303,
+      "learning_rate": 1.1541010483085873e-06,
+      "loss": 0.6224,
+      "step": 15664
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.8587129575264316,
+      "learning_rate": 1.1540055936512518e-06,
+      "loss": 0.6185,
+      "step": 15665
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.2497843426636828,
+      "learning_rate": 1.1539101375565683e-06,
+      "loss": 0.6595,
+      "step": 15666
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 5.754718187596582,
+      "learning_rate": 1.153814680025428e-06,
+      "loss": 0.6276,
+      "step": 15667
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.447118504370182,
+      "learning_rate": 1.1537192210587218e-06,
+      "loss": 0.6185,
+      "step": 15668
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.328710757410439,
+      "learning_rate": 1.1536237606573404e-06,
+      "loss": 0.6413,
+      "step": 15669
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.6661153897159395,
+      "learning_rate": 1.1535282988221749e-06,
+      "loss": 0.6185,
+      "step": 15670
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.7405932491888594,
+      "learning_rate": 1.1534328355541159e-06,
+      "loss": 0.6439,
+      "step": 15671
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.4621293812382055,
+      "learning_rate": 1.1533373708540553e-06,
+      "loss": 0.5898,
+      "step": 15672
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.022729422097801,
+      "learning_rate": 1.153241904722883e-06,
+      "loss": 0.6374,
+      "step": 15673
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.84084717925837,
+      "learning_rate": 1.1531464371614904e-06,
+      "loss": 0.61,
+      "step": 15674
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.8161507912385177,
+      "learning_rate": 1.1530509681707689e-06,
+      "loss": 0.6549,
+      "step": 15675
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.987189004999574,
+      "learning_rate": 1.152955497751609e-06,
+      "loss": 0.6654,
+      "step": 15676
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.88356479831321,
+      "learning_rate": 1.1528600259049018e-06,
+      "loss": 0.6615,
+      "step": 15677
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.9346950892720485,
+      "learning_rate": 1.1527645526315386e-06,
+      "loss": 0.6732,
+      "step": 15678
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.2141904745029244,
+      "learning_rate": 1.1526690779324102e-06,
+      "loss": 0.694,
+      "step": 15679
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.897935208374883,
+      "learning_rate": 1.152573601808408e-06,
+      "loss": 0.6224,
+      "step": 15680
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.9372173983179355,
+      "learning_rate": 1.1524781242604226e-06,
+      "loss": 0.6432,
+      "step": 15681
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.925257485355486,
+      "learning_rate": 1.1523826452893456e-06,
+      "loss": 0.6582,
+      "step": 15682
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.1952486376687053,
+      "learning_rate": 1.1522871648960679e-06,
+      "loss": 0.6406,
+      "step": 15683
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.886616846939566,
+      "learning_rate": 1.1521916830814805e-06,
+      "loss": 0.6986,
+      "step": 15684
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.4615798638249675,
+      "learning_rate": 1.1520961998464748e-06,
+      "loss": 0.6497,
+      "step": 15685
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.129965558469427,
+      "learning_rate": 1.1520007151919418e-06,
+      "loss": 0.6309,
+      "step": 15686
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.080105933625945,
+      "learning_rate": 1.1519052291187724e-06,
+      "loss": 0.6615,
+      "step": 15687
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.3539778567397156,
+      "learning_rate": 1.1518097416278585e-06,
+      "loss": 0.6185,
+      "step": 15688
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.064220468699017,
+      "learning_rate": 1.1517142527200904e-06,
+      "loss": 0.6296,
+      "step": 15689
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.7321470141482638,
+      "learning_rate": 1.1516187623963602e-06,
+      "loss": 0.6309,
+      "step": 15690
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.578949423135296,
+      "learning_rate": 1.1515232706575585e-06,
+      "loss": 0.6484,
+      "step": 15691
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.633675003007874,
+      "learning_rate": 1.1514277775045766e-06,
+      "loss": 0.6263,
+      "step": 15692
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.885838949841728,
+      "learning_rate": 1.1513322829383061e-06,
+      "loss": 0.6478,
+      "step": 15693
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.343078053993023,
+      "learning_rate": 1.1512367869596378e-06,
+      "loss": 0.6191,
+      "step": 15694
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.9842074149548963,
+      "learning_rate": 1.1511412895694633e-06,
+      "loss": 0.6126,
+      "step": 15695
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.521089335142912,
+      "learning_rate": 1.1510457907686737e-06,
+      "loss": 0.6478,
+      "step": 15696
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.482890300966469,
+      "learning_rate": 1.1509502905581604e-06,
+      "loss": 0.6484,
+      "step": 15697
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.9094262528278803,
+      "learning_rate": 1.1508547889388149e-06,
+      "loss": 0.6367,
+      "step": 15698
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.9982784180826307,
+      "learning_rate": 1.150759285911528e-06,
+      "loss": 0.64,
+      "step": 15699
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.1275255547958087,
+      "learning_rate": 1.1506637814771913e-06,
+      "loss": 0.6693,
+      "step": 15700
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.685926229939487,
+      "learning_rate": 1.1505682756366965e-06,
+      "loss": 0.6458,
+      "step": 15701
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.0941180310864707,
+      "learning_rate": 1.1504727683909349e-06,
+      "loss": 0.627,
+      "step": 15702
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.6306704361551527,
+      "learning_rate": 1.1503772597407972e-06,
+      "loss": 0.6055,
+      "step": 15703
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.52695286094051,
+      "learning_rate": 1.1502817496871752e-06,
+      "loss": 0.6536,
+      "step": 15704
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.825363117542417,
+      "learning_rate": 1.1501862382309603e-06,
+      "loss": 0.6582,
+      "step": 15705
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.949685422712726,
+      "learning_rate": 1.1500907253730442e-06,
+      "loss": 0.6185,
+      "step": 15706
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.382802791397897,
+      "learning_rate": 1.149995211114318e-06,
+      "loss": 0.6797,
+      "step": 15707
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.0945034830664593,
+      "learning_rate": 1.1498996954556735e-06,
+      "loss": 0.6797,
+      "step": 15708
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.6186114990174603,
+      "learning_rate": 1.1498041783980016e-06,
+      "loss": 0.6836,
+      "step": 15709
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.051023623626522,
+      "learning_rate": 1.1497086599421939e-06,
+      "loss": 0.6445,
+      "step": 15710
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.78068803607225,
+      "learning_rate": 1.1496131400891424e-06,
+      "loss": 0.6732,
+      "step": 15711
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.6375284002017274,
+      "learning_rate": 1.149517618839738e-06,
+      "loss": 0.6517,
+      "step": 15712
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.372721010788345,
+      "learning_rate": 1.1494220961948725e-06,
+      "loss": 0.6237,
+      "step": 15713
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.152446427751764,
+      "learning_rate": 1.1493265721554373e-06,
+      "loss": 0.6175,
+      "step": 15714
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.8948802829983626,
+      "learning_rate": 1.149231046722324e-06,
+      "loss": 0.6432,
+      "step": 15715
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.283449830443574,
+      "learning_rate": 1.1491355198964244e-06,
+      "loss": 0.6725,
+      "step": 15716
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.851442016057324,
+      "learning_rate": 1.1490399916786295e-06,
+      "loss": 0.6777,
+      "step": 15717
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.6102574492582638,
+      "learning_rate": 1.1489444620698314e-06,
+      "loss": 0.6647,
+      "step": 15718
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.4700270412713095,
+      "learning_rate": 1.1488489310709213e-06,
+      "loss": 0.6204,
+      "step": 15719
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.062447643924152,
+      "learning_rate": 1.1487533986827908e-06,
+      "loss": 0.6764,
+      "step": 15720
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.720075936422059,
+      "learning_rate": 1.1486578649063322e-06,
+      "loss": 0.6549,
+      "step": 15721
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 4.010379650965508,
+      "learning_rate": 1.1485623297424362e-06,
+      "loss": 0.6413,
+      "step": 15722
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.5289676563978225,
+      "learning_rate": 1.1484667931919949e-06,
+      "loss": 0.6354,
+      "step": 15723
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.052904186351656,
+      "learning_rate": 1.1483712552559e-06,
+      "loss": 0.6237,
+      "step": 15724
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.657520446853786,
+      "learning_rate": 1.1482757159350428e-06,
+      "loss": 0.6146,
+      "step": 15725
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.5690870424979657,
+      "learning_rate": 1.1481801752303155e-06,
+      "loss": 0.6338,
+      "step": 15726
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.0889984342859775,
+      "learning_rate": 1.1480846331426096e-06,
+      "loss": 0.6237,
+      "step": 15727
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.5660000761573403,
+      "learning_rate": 1.1479890896728165e-06,
+      "loss": 0.6582,
+      "step": 15728
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.195935981915697,
+      "learning_rate": 1.147893544821828e-06,
+      "loss": 0.6569,
+      "step": 15729
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.3670132970366704,
+      "learning_rate": 1.1477979985905361e-06,
+      "loss": 0.6497,
+      "step": 15730
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.5789485293404772,
+      "learning_rate": 1.1477024509798325e-06,
+      "loss": 0.625,
+      "step": 15731
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.1871775960534143,
+      "learning_rate": 1.1476069019906088e-06,
+      "loss": 0.6367,
+      "step": 15732
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.504116314003527,
+      "learning_rate": 1.147511351623757e-06,
+      "loss": 0.6452,
+      "step": 15733
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.532051763299711,
+      "learning_rate": 1.1474157998801683e-06,
+      "loss": 0.6152,
+      "step": 15734
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.3703383446860156,
+      "learning_rate": 1.1473202467607355e-06,
+      "loss": 0.6178,
+      "step": 15735
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.0636318313708455,
+      "learning_rate": 1.1472246922663493e-06,
+      "loss": 0.6465,
+      "step": 15736
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.8414192162872847,
+      "learning_rate": 1.147129136397902e-06,
+      "loss": 0.6172,
+      "step": 15737
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.0009366635561077,
+      "learning_rate": 1.147033579156286e-06,
+      "loss": 0.6211,
+      "step": 15738
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.6734974543068795,
+      "learning_rate": 1.1469380205423923e-06,
+      "loss": 0.6341,
+      "step": 15739
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.756253153236887,
+      "learning_rate": 1.1468424605571129e-06,
+      "loss": 0.6237,
+      "step": 15740
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.130925840119248,
+      "learning_rate": 1.14674689920134e-06,
+      "loss": 0.6458,
+      "step": 15741
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.803978799320831,
+      "learning_rate": 1.1466513364759652e-06,
+      "loss": 0.6022,
+      "step": 15742
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.7517971539431136,
+      "learning_rate": 1.1465557723818806e-06,
+      "loss": 0.64,
+      "step": 15743
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.330957923988224,
+      "learning_rate": 1.146460206919978e-06,
+      "loss": 0.653,
+      "step": 15744
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.307130169235897,
+      "learning_rate": 1.146364640091149e-06,
+      "loss": 0.6283,
+      "step": 15745
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.5268654147622023,
+      "learning_rate": 1.1462690718962865e-06,
+      "loss": 0.6445,
+      "step": 15746
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.8505924695752096,
+      "learning_rate": 1.1461735023362813e-06,
+      "loss": 0.6725,
+      "step": 15747
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.945243620127022,
+      "learning_rate": 1.146077931412026e-06,
+      "loss": 0.6582,
+      "step": 15748
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.7938599494167944,
+      "learning_rate": 1.1459823591244128e-06,
+      "loss": 0.6185,
+      "step": 15749
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.872090063427313,
+      "learning_rate": 1.1458867854743329e-06,
+      "loss": 0.6445,
+      "step": 15750
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.405852772573199,
+      "learning_rate": 1.1457912104626792e-06,
+      "loss": 0.6543,
+      "step": 15751
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.635019368354188,
+      "learning_rate": 1.1456956340903428e-06,
+      "loss": 0.6445,
+      "step": 15752
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.457014732519546,
+      "learning_rate": 1.145600056358216e-06,
+      "loss": 0.6673,
+      "step": 15753
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.9777934003752535,
+      "learning_rate": 1.1455044772671916e-06,
+      "loss": 0.6465,
+      "step": 15754
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.856658816385258,
+      "learning_rate": 1.1454088968181607e-06,
+      "loss": 0.6517,
+      "step": 15755
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.13808181392674,
+      "learning_rate": 1.1453133150120158e-06,
+      "loss": 0.6628,
+      "step": 15756
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.7631023556847514,
+      "learning_rate": 1.1452177318496485e-06,
+      "loss": 0.6484,
+      "step": 15757
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.282009130556564,
+      "learning_rate": 1.1451221473319518e-06,
+      "loss": 0.6693,
+      "step": 15758
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.9920338650546516,
+      "learning_rate": 1.145026561459817e-06,
+      "loss": 0.6686,
+      "step": 15759
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.1509874262076183,
+      "learning_rate": 1.1449309742341364e-06,
+      "loss": 0.6413,
+      "step": 15760
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.8018267059457003,
+      "learning_rate": 1.1448353856558023e-06,
+      "loss": 0.6439,
+      "step": 15761
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 8.28803853535034,
+      "learning_rate": 1.144739795725707e-06,
+      "loss": 0.7057,
+      "step": 15762
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.73680842260369,
+      "learning_rate": 1.144644204444742e-06,
+      "loss": 0.6647,
+      "step": 15763
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.166416304849486,
+      "learning_rate": 1.1445486118138003e-06,
+      "loss": 0.6458,
+      "step": 15764
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.703528405233583,
+      "learning_rate": 1.1444530178337732e-06,
+      "loss": 0.6621,
+      "step": 15765
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.640757906154732,
+      "learning_rate": 1.1443574225055534e-06,
+      "loss": 0.7018,
+      "step": 15766
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.9754659224642324,
+      "learning_rate": 1.144261825830033e-06,
+      "loss": 0.6348,
+      "step": 15767
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.804486890726017,
+      "learning_rate": 1.1441662278081044e-06,
+      "loss": 0.6543,
+      "step": 15768
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.942736255632947,
+      "learning_rate": 1.1440706284406595e-06,
+      "loss": 0.6719,
+      "step": 15769
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.2033760045583057,
+      "learning_rate": 1.1439750277285908e-06,
+      "loss": 0.6361,
+      "step": 15770
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.8871671658111158,
+      "learning_rate": 1.1438794256727902e-06,
+      "loss": 0.6751,
+      "step": 15771
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.5580586561029928,
+      "learning_rate": 1.1437838222741504e-06,
+      "loss": 0.6413,
+      "step": 15772
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.3123556844308912,
+      "learning_rate": 1.1436882175335633e-06,
+      "loss": 0.6504,
+      "step": 15773
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.063707503027957,
+      "learning_rate": 1.1435926114519214e-06,
+      "loss": 0.6328,
+      "step": 15774
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.1435534451664644,
+      "learning_rate": 1.1434970040301172e-06,
+      "loss": 0.6589,
+      "step": 15775
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.241252064928334,
+      "learning_rate": 1.1434013952690425e-06,
+      "loss": 0.6257,
+      "step": 15776
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.4057395788172693,
+      "learning_rate": 1.14330578516959e-06,
+      "loss": 0.6615,
+      "step": 15777
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 5.299938838644335,
+      "learning_rate": 1.143210173732652e-06,
+      "loss": 0.6836,
+      "step": 15778
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.787373826183395,
+      "learning_rate": 1.1431145609591206e-06,
+      "loss": 0.6387,
+      "step": 15779
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.8994612934185193,
+      "learning_rate": 1.1430189468498885e-06,
+      "loss": 0.6296,
+      "step": 15780
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.9130000465147714,
+      "learning_rate": 1.142923331405848e-06,
+      "loss": 0.6367,
+      "step": 15781
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.725431092429716,
+      "learning_rate": 1.1428277146278912e-06,
+      "loss": 0.6211,
+      "step": 15782
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.5534234391494612,
+      "learning_rate": 1.1427320965169108e-06,
+      "loss": 0.6725,
+      "step": 15783
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.2480678173886406,
+      "learning_rate": 1.142636477073799e-06,
+      "loss": 0.666,
+      "step": 15784
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.581879546519884,
+      "learning_rate": 1.1425408562994484e-06,
+      "loss": 0.6191,
+      "step": 15785
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.317185136606181,
+      "learning_rate": 1.1424452341947515e-06,
+      "loss": 0.6439,
+      "step": 15786
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.964601335554055,
+      "learning_rate": 1.1423496107606005e-06,
+      "loss": 0.6465,
+      "step": 15787
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.201083412378467,
+      "learning_rate": 1.1422539859978881e-06,
+      "loss": 0.6517,
+      "step": 15788
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.557976696816228,
+      "learning_rate": 1.1421583599075067e-06,
+      "loss": 0.625,
+      "step": 15789
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 3.290349810740807,
+      "learning_rate": 1.1420627324903486e-06,
+      "loss": 0.6367,
+      "step": 15790
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.100428662102095,
+      "learning_rate": 1.1419671037473067e-06,
+      "loss": 0.6569,
+      "step": 15791
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.666687669845471,
+      "learning_rate": 1.141871473679273e-06,
+      "loss": 0.6335,
+      "step": 15792
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.20502657655601,
+      "learning_rate": 1.1417758422871404e-06,
+      "loss": 0.6634,
+      "step": 15793
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.808172981889703,
+      "learning_rate": 1.1416802095718014e-06,
+      "loss": 0.6602,
+      "step": 15794
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.5929471755788915,
+      "learning_rate": 1.1415845755341481e-06,
+      "loss": 0.6068,
+      "step": 15795
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.688515094969571,
+      "learning_rate": 1.141488940175074e-06,
+      "loss": 0.6595,
+      "step": 15796
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.887162031681495,
+      "learning_rate": 1.1413933034954708e-06,
+      "loss": 0.6491,
+      "step": 15797
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.703942140329897,
+      "learning_rate": 1.1412976654962314e-06,
+      "loss": 0.6491,
+      "step": 15798
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.8281227006674934,
+      "learning_rate": 1.1412020261782485e-06,
+      "loss": 0.6322,
+      "step": 15799
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.306939327565783,
+      "learning_rate": 1.1411063855424144e-06,
+      "loss": 0.6029,
+      "step": 15800
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.3705056623371785,
+      "learning_rate": 1.1410107435896222e-06,
+      "loss": 0.6471,
+      "step": 15801
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.6121370281135845,
+      "learning_rate": 1.140915100320764e-06,
+      "loss": 0.6465,
+      "step": 15802
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.276677718013321,
+      "learning_rate": 1.1408194557367328e-06,
+      "loss": 0.6719,
+      "step": 15803
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.4974152409941497,
+      "learning_rate": 1.1407238098384214e-06,
+      "loss": 0.6113,
+      "step": 15804
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.8520717013289967,
+      "learning_rate": 1.1406281626267218e-06,
+      "loss": 0.6296,
+      "step": 15805
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.5512460417856415,
+      "learning_rate": 1.1405325141025275e-06,
+      "loss": 0.6413,
+      "step": 15806
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.8520375539935556,
+      "learning_rate": 1.1404368642667305e-06,
+      "loss": 0.6302,
+      "step": 15807
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.573950188715342,
+      "learning_rate": 1.1403412131202238e-06,
+      "loss": 0.6673,
+      "step": 15808
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.046882024338023,
+      "learning_rate": 1.1402455606639006e-06,
+      "loss": 0.6003,
+      "step": 15809
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.2947965362506473,
+      "learning_rate": 1.140149906898653e-06,
+      "loss": 0.6393,
+      "step": 15810
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.179838462470135,
+      "learning_rate": 1.1400542518253737e-06,
+      "loss": 0.6146,
+      "step": 15811
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.7140892324830417,
+      "learning_rate": 1.139958595444956e-06,
+      "loss": 0.6413,
+      "step": 15812
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.5935490671847536,
+      "learning_rate": 1.1398629377582919e-06,
+      "loss": 0.5908,
+      "step": 15813
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.926009256755029,
+      "learning_rate": 1.1397672787662749e-06,
+      "loss": 0.6803,
+      "step": 15814
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.172859224099184,
+      "learning_rate": 1.1396716184697978e-06,
+      "loss": 0.651,
+      "step": 15815
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.1458438290852886,
+      "learning_rate": 1.1395759568697526e-06,
+      "loss": 0.6432,
+      "step": 15816
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.0628061464334415,
+      "learning_rate": 1.1394802939670332e-06,
+      "loss": 0.6289,
+      "step": 15817
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.8751970948541596,
+      "learning_rate": 1.1393846297625315e-06,
+      "loss": 0.6237,
+      "step": 15818
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.354192310415196,
+      "learning_rate": 1.1392889642571406e-06,
+      "loss": 0.6003,
+      "step": 15819
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.869629816777073,
+      "learning_rate": 1.139193297451754e-06,
+      "loss": 0.6335,
+      "step": 15820
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.4859566088744685,
+      "learning_rate": 1.1390976293472636e-06,
+      "loss": 0.6504,
+      "step": 15821
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.157257447486696,
+      "learning_rate": 1.1390019599445631e-06,
+      "loss": 0.6921,
+      "step": 15822
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.894028944016534,
+      "learning_rate": 1.1389062892445444e-06,
+      "loss": 0.6589,
+      "step": 15823
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.3692942838173545,
+      "learning_rate": 1.1388106172481015e-06,
+      "loss": 0.6576,
+      "step": 15824
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.4315443975371744,
+      "learning_rate": 1.1387149439561269e-06,
+      "loss": 0.6523,
+      "step": 15825
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.135536008329789,
+      "learning_rate": 1.1386192693695132e-06,
+      "loss": 0.6335,
+      "step": 15826
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.8581802280731794,
+      "learning_rate": 1.1385235934891538e-06,
+      "loss": 0.6263,
+      "step": 15827
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.9941127203817457,
+      "learning_rate": 1.1384279163159414e-06,
+      "loss": 0.6178,
+      "step": 15828
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.0551097855204254,
+      "learning_rate": 1.1383322378507687e-06,
+      "loss": 0.6654,
+      "step": 15829
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.481017136001444,
+      "learning_rate": 1.1382365580945297e-06,
+      "loss": 0.6647,
+      "step": 15830
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.163907449752612,
+      "learning_rate": 1.138140877048116e-06,
+      "loss": 0.6901,
+      "step": 15831
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.0783049740645065,
+      "learning_rate": 1.1380451947124212e-06,
+      "loss": 0.6719,
+      "step": 15832
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.7912017597442156,
+      "learning_rate": 1.1379495110883388e-06,
+      "loss": 0.638,
+      "step": 15833
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.305357684386184,
+      "learning_rate": 1.1378538261767613e-06,
+      "loss": 0.6354,
+      "step": 15834
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.8745130786079143,
+      "learning_rate": 1.137758139978582e-06,
+      "loss": 0.6419,
+      "step": 15835
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.212757782947894,
+      "learning_rate": 1.1376624524946936e-06,
+      "loss": 0.6107,
+      "step": 15836
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.7874742298146407,
+      "learning_rate": 1.1375667637259896e-06,
+      "loss": 0.6022,
+      "step": 15837
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.0945619462919427,
+      "learning_rate": 1.1374710736733625e-06,
+      "loss": 0.6647,
+      "step": 15838
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.9595730830672546,
+      "learning_rate": 1.1373753823377059e-06,
+      "loss": 0.6165,
+      "step": 15839
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.146828433752816,
+      "learning_rate": 1.1372796897199126e-06,
+      "loss": 0.6693,
+      "step": 15840
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.5734080731958113,
+      "learning_rate": 1.1371839958208759e-06,
+      "loss": 0.6133,
+      "step": 15841
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.10739823715781,
+      "learning_rate": 1.137088300641489e-06,
+      "loss": 0.6315,
+      "step": 15842
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.7811276497030812,
+      "learning_rate": 1.1369926041826446e-06,
+      "loss": 0.6458,
+      "step": 15843
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.804147747330815,
+      "learning_rate": 1.1368969064452364e-06,
+      "loss": 0.6491,
+      "step": 15844
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.9322221472465175,
+      "learning_rate": 1.1368012074301567e-06,
+      "loss": 0.6439,
+      "step": 15845
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.327389978169689,
+      "learning_rate": 1.1367055071382995e-06,
+      "loss": 0.6517,
+      "step": 15846
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.7897626530579944,
+      "learning_rate": 1.136609805570558e-06,
+      "loss": 0.6335,
+      "step": 15847
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.743807995404757,
+      "learning_rate": 1.1365141027278249e-06,
+      "loss": 0.6322,
+      "step": 15848
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.7473845884742656,
+      "learning_rate": 1.1364183986109937e-06,
+      "loss": 0.6341,
+      "step": 15849
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.523371044455536,
+      "learning_rate": 1.1363226932209573e-06,
+      "loss": 0.668,
+      "step": 15850
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.034440062312992,
+      "learning_rate": 1.1362269865586094e-06,
+      "loss": 0.6536,
+      "step": 15851
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.4252794925782224,
+      "learning_rate": 1.1361312786248426e-06,
+      "loss": 0.6309,
+      "step": 15852
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.188381073125677,
+      "learning_rate": 1.136035569420551e-06,
+      "loss": 0.6615,
+      "step": 15853
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 5.362495473130766,
+      "learning_rate": 1.1359398589466275e-06,
+      "loss": 0.6452,
+      "step": 15854
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.5611368692369165,
+      "learning_rate": 1.1358441472039646e-06,
+      "loss": 0.6335,
+      "step": 15855
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 4.7911299203316124,
+      "learning_rate": 1.1357484341934567e-06,
+      "loss": 0.6393,
+      "step": 15856
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.838695112652654,
+      "learning_rate": 1.1356527199159968e-06,
+      "loss": 0.6348,
+      "step": 15857
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 6.535957255986538,
+      "learning_rate": 1.1355570043724775e-06,
+      "loss": 0.6432,
+      "step": 15858
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.3936142952134225,
+      "learning_rate": 1.1354612875637932e-06,
+      "loss": 0.7207,
+      "step": 15859
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.78743223454934,
+      "learning_rate": 1.1353655694908369e-06,
+      "loss": 0.6471,
+      "step": 15860
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.69397850037158,
+      "learning_rate": 1.135269850154501e-06,
+      "loss": 0.6504,
+      "step": 15861
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.3889493402897677,
+      "learning_rate": 1.1351741295556803e-06,
+      "loss": 0.7096,
+      "step": 15862
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.182448437659767,
+      "learning_rate": 1.1350784076952674e-06,
+      "loss": 0.6139,
+      "step": 15863
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.025287179788519,
+      "learning_rate": 1.1349826845741554e-06,
+      "loss": 0.6471,
+      "step": 15864
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.569760339880629,
+      "learning_rate": 1.1348869601932386e-06,
+      "loss": 0.6413,
+      "step": 15865
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.6045568196374487,
+      "learning_rate": 1.1347912345534093e-06,
+      "loss": 0.64,
+      "step": 15866
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.736131967806333,
+      "learning_rate": 1.134695507655562e-06,
+      "loss": 0.6055,
+      "step": 15867
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.99468582354883,
+      "learning_rate": 1.1345997795005893e-06,
+      "loss": 0.6042,
+      "step": 15868
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.7640156324677063,
+      "learning_rate": 1.134504050089385e-06,
+      "loss": 0.6224,
+      "step": 15869
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.0797146889636147,
+      "learning_rate": 1.1344083194228425e-06,
+      "loss": 0.651,
+      "step": 15870
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.777392407644592,
+      "learning_rate": 1.1343125875018553e-06,
+      "loss": 0.6302,
+      "step": 15871
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.239444688259647,
+      "learning_rate": 1.1342168543273168e-06,
+      "loss": 0.6361,
+      "step": 15872
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.784180383397691,
+      "learning_rate": 1.1341211199001209e-06,
+      "loss": 0.6341,
+      "step": 15873
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.310249779437611,
+      "learning_rate": 1.1340253842211601e-06,
+      "loss": 0.6458,
+      "step": 15874
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.293694521614699,
+      "learning_rate": 1.133929647291329e-06,
+      "loss": 0.6074,
+      "step": 15875
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.9632932281772253,
+      "learning_rate": 1.1338339091115205e-06,
+      "loss": 0.6165,
+      "step": 15876
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.911498963109746,
+      "learning_rate": 1.1337381696826282e-06,
+      "loss": 0.5967,
+      "step": 15877
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.6289929855047682,
+      "learning_rate": 1.133642429005546e-06,
+      "loss": 0.6875,
+      "step": 15878
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.1489766152677765,
+      "learning_rate": 1.133546687081167e-06,
+      "loss": 0.6367,
+      "step": 15879
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.8468239618161966,
+      "learning_rate": 1.133450943910385e-06,
+      "loss": 0.6849,
+      "step": 15880
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.534982243184555,
+      "learning_rate": 1.1333551994940936e-06,
+      "loss": 0.6302,
+      "step": 15881
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.078178378655002,
+      "learning_rate": 1.133259453833186e-06,
+      "loss": 0.6419,
+      "step": 15882
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.8773638768870575,
+      "learning_rate": 1.1331637069285568e-06,
+      "loss": 0.6419,
+      "step": 15883
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.7620293993691156,
+      "learning_rate": 1.1330679587810982e-06,
+      "loss": 0.6784,
+      "step": 15884
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.555117029698259,
+      "learning_rate": 1.132972209391705e-06,
+      "loss": 0.6055,
+      "step": 15885
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.033466737531785,
+      "learning_rate": 1.1328764587612702e-06,
+      "loss": 0.6097,
+      "step": 15886
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.042219328337598,
+      "learning_rate": 1.132780706890688e-06,
+      "loss": 0.6608,
+      "step": 15887
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.392516632518395,
+      "learning_rate": 1.1326849537808516e-06,
+      "loss": 0.6706,
+      "step": 15888
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.961284250711164,
+      "learning_rate": 1.1325891994326543e-06,
+      "loss": 0.6543,
+      "step": 15889
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.8681460287775384,
+      "learning_rate": 1.1324934438469908e-06,
+      "loss": 0.6074,
+      "step": 15890
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.0494756028003267,
+      "learning_rate": 1.1323976870247542e-06,
+      "loss": 0.6654,
+      "step": 15891
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.529816430758364,
+      "learning_rate": 1.132301928966838e-06,
+      "loss": 0.6953,
+      "step": 15892
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.8862844058575616,
+      "learning_rate": 1.1322061696741368e-06,
+      "loss": 0.6237,
+      "step": 15893
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.2209552667348422,
+      "learning_rate": 1.1321104091475433e-06,
+      "loss": 0.6862,
+      "step": 15894
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.72895880041363,
+      "learning_rate": 1.1320146473879515e-06,
+      "loss": 0.6608,
+      "step": 15895
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.9396478320682125,
+      "learning_rate": 1.131918884396256e-06,
+      "loss": 0.6732,
+      "step": 15896
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.148769384043705,
+      "learning_rate": 1.1318231201733492e-06,
+      "loss": 0.653,
+      "step": 15897
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.2962009771422327,
+      "learning_rate": 1.1317273547201256e-06,
+      "loss": 0.6471,
+      "step": 15898
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.520487743568668,
+      "learning_rate": 1.1316315880374792e-06,
+      "loss": 0.6393,
+      "step": 15899
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.278244560681304,
+      "learning_rate": 1.1315358201263034e-06,
+      "loss": 0.6315,
+      "step": 15900
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.269449686826547,
+      "learning_rate": 1.1314400509874925e-06,
+      "loss": 0.6074,
+      "step": 15901
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.6007146760572435,
+      "learning_rate": 1.1313442806219396e-06,
+      "loss": 0.627,
+      "step": 15902
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.6962125797987073,
+      "learning_rate": 1.131248509030539e-06,
+      "loss": 0.6243,
+      "step": 15903
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.2157323529812376,
+      "learning_rate": 1.1311527362141847e-06,
+      "loss": 0.6243,
+      "step": 15904
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.9193321465564037,
+      "learning_rate": 1.1310569621737699e-06,
+      "loss": 0.6966,
+      "step": 15905
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.304618995516055,
+      "learning_rate": 1.130961186910189e-06,
+      "loss": 0.6322,
+      "step": 15906
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.7437406784059326,
+      "learning_rate": 1.1308654104243361e-06,
+      "loss": 0.696,
+      "step": 15907
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.221913687051314,
+      "learning_rate": 1.1307696327171042e-06,
+      "loss": 0.6413,
+      "step": 15908
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.8362109930634443,
+      "learning_rate": 1.130673853789388e-06,
+      "loss": 0.653,
+      "step": 15909
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 4.4415557697326244,
+      "learning_rate": 1.1305780736420813e-06,
+      "loss": 0.651,
+      "step": 15910
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.0959303538892415,
+      "learning_rate": 1.1304822922760776e-06,
+      "loss": 0.6999,
+      "step": 15911
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.579657961169723,
+      "learning_rate": 1.1303865096922717e-06,
+      "loss": 0.6139,
+      "step": 15912
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.6999772839685936,
+      "learning_rate": 1.1302907258915565e-06,
+      "loss": 0.6445,
+      "step": 15913
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.984905840534368,
+      "learning_rate": 1.1301949408748263e-06,
+      "loss": 0.6576,
+      "step": 15914
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.68412456516553,
+      "learning_rate": 1.1300991546429756e-06,
+      "loss": 0.6882,
+      "step": 15915
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.2862465981914606,
+      "learning_rate": 1.1300033671968981e-06,
+      "loss": 0.5827,
+      "step": 15916
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.8980314052684344,
+      "learning_rate": 1.1299075785374874e-06,
+      "loss": 0.5983,
+      "step": 15917
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.420323316350745,
+      "learning_rate": 1.1298117886656376e-06,
+      "loss": 0.6263,
+      "step": 15918
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.3264476586519303,
+      "learning_rate": 1.1297159975822427e-06,
+      "loss": 0.6146,
+      "step": 15919
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.586910110136496,
+      "learning_rate": 1.1296202052881977e-06,
+      "loss": 0.6296,
+      "step": 15920
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.1691436034217277,
+      "learning_rate": 1.1295244117843954e-06,
+      "loss": 0.6686,
+      "step": 15921
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.1756549485068892,
+      "learning_rate": 1.1294286170717304e-06,
+      "loss": 0.6449,
+      "step": 15922
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.7703726995010727,
+      "learning_rate": 1.1293328211510966e-06,
+      "loss": 0.6322,
+      "step": 15923
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.610535435982195,
+      "learning_rate": 1.1292370240233882e-06,
+      "loss": 0.6471,
+      "step": 15924
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.7102247833893967,
+      "learning_rate": 1.1291412256894993e-06,
+      "loss": 0.6302,
+      "step": 15925
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.683518268704734,
+      "learning_rate": 1.129045426150324e-06,
+      "loss": 0.6484,
+      "step": 15926
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.85041928147152,
+      "learning_rate": 1.1289496254067558e-06,
+      "loss": 0.6589,
+      "step": 15927
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.060261737069827,
+      "learning_rate": 1.1288538234596899e-06,
+      "loss": 0.7057,
+      "step": 15928
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.151229271133701,
+      "learning_rate": 1.1287580203100199e-06,
+      "loss": 0.5846,
+      "step": 15929
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.347921214316046,
+      "learning_rate": 1.1286622159586392e-06,
+      "loss": 0.6549,
+      "step": 15930
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.818109952572884,
+      "learning_rate": 1.1285664104064435e-06,
+      "loss": 0.6354,
+      "step": 15931
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.516593196405455,
+      "learning_rate": 1.1284706036543254e-06,
+      "loss": 0.6497,
+      "step": 15932
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.648431812257964,
+      "learning_rate": 1.1283747957031805e-06,
+      "loss": 0.6536,
+      "step": 15933
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.7030268458017934,
+      "learning_rate": 1.1282789865539018e-06,
+      "loss": 0.5983,
+      "step": 15934
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.1146956595623947,
+      "learning_rate": 1.128183176207384e-06,
+      "loss": 0.6152,
+      "step": 15935
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.3537400021975876,
+      "learning_rate": 1.1280873646645217e-06,
+      "loss": 0.6497,
+      "step": 15936
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.199663106743153,
+      "learning_rate": 1.1279915519262083e-06,
+      "loss": 0.6862,
+      "step": 15937
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.72473371856909,
+      "learning_rate": 1.1278957379933385e-06,
+      "loss": 0.6367,
+      "step": 15938
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.0411363010506207,
+      "learning_rate": 1.1277999228668065e-06,
+      "loss": 0.6087,
+      "step": 15939
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.17421930294263,
+      "learning_rate": 1.1277041065475063e-06,
+      "loss": 0.6927,
+      "step": 15940
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.390496714832426,
+      "learning_rate": 1.1276082890363327e-06,
+      "loss": 0.5957,
+      "step": 15941
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.295812400248977,
+      "learning_rate": 1.1275124703341794e-06,
+      "loss": 0.694,
+      "step": 15942
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.0083738480547475,
+      "learning_rate": 1.1274166504419412e-06,
+      "loss": 0.6608,
+      "step": 15943
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.0318904459347955,
+      "learning_rate": 1.127320829360512e-06,
+      "loss": 0.6465,
+      "step": 15944
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.9782931183573833,
+      "learning_rate": 1.127225007090786e-06,
+      "loss": 0.6628,
+      "step": 15945
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.0025654423306944,
+      "learning_rate": 1.1271291836336582e-06,
+      "loss": 0.6211,
+      "step": 15946
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.267129140141695,
+      "learning_rate": 1.1270333589900223e-06,
+      "loss": 0.7005,
+      "step": 15947
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 9.473915089093387,
+      "learning_rate": 1.1269375331607726e-06,
+      "loss": 0.7376,
+      "step": 15948
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.0353836659594706,
+      "learning_rate": 1.1268417061468042e-06,
+      "loss": 0.5801,
+      "step": 15949
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.814409653992196,
+      "learning_rate": 1.1267458779490107e-06,
+      "loss": 0.6374,
+      "step": 15950
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.73925987622777,
+      "learning_rate": 1.1266500485682865e-06,
+      "loss": 0.6445,
+      "step": 15951
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.1118327128741576,
+      "learning_rate": 1.1265542180055263e-06,
+      "loss": 0.6693,
+      "step": 15952
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.8620688949316597,
+      "learning_rate": 1.1264583862616243e-06,
+      "loss": 0.6374,
+      "step": 15953
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.714676099440628,
+      "learning_rate": 1.1263625533374754e-06,
+      "loss": 0.6634,
+      "step": 15954
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.491328275974797,
+      "learning_rate": 1.1262667192339732e-06,
+      "loss": 0.6152,
+      "step": 15955
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.557864691331143,
+      "learning_rate": 1.1261708839520128e-06,
+      "loss": 0.5977,
+      "step": 15956
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.984145693881013,
+      "learning_rate": 1.1260750474924884e-06,
+      "loss": 0.6302,
+      "step": 15957
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.97179597802095,
+      "learning_rate": 1.1259792098562942e-06,
+      "loss": 0.6745,
+      "step": 15958
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.9074186933901665,
+      "learning_rate": 1.125883371044325e-06,
+      "loss": 0.6322,
+      "step": 15959
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 6.771173283201224,
+      "learning_rate": 1.125787531057475e-06,
+      "loss": 0.6536,
+      "step": 15960
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.913632850965144,
+      "learning_rate": 1.1256916898966391e-06,
+      "loss": 0.5872,
+      "step": 15961
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.5265707299778337,
+      "learning_rate": 1.1255958475627115e-06,
+      "loss": 0.6068,
+      "step": 15962
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.668743129347392,
+      "learning_rate": 1.1255000040565868e-06,
+      "loss": 0.6185,
+      "step": 15963
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.6847375622058953,
+      "learning_rate": 1.1254041593791593e-06,
+      "loss": 0.6693,
+      "step": 15964
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.852563075570596,
+      "learning_rate": 1.1253083135313237e-06,
+      "loss": 0.6257,
+      "step": 15965
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.144809142907136,
+      "learning_rate": 1.1252124665139747e-06,
+      "loss": 0.6738,
+      "step": 15966
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.0574801540987244,
+      "learning_rate": 1.1251166183280065e-06,
+      "loss": 0.6497,
+      "step": 15967
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.3330607053067824,
+      "learning_rate": 1.1250207689743136e-06,
+      "loss": 0.6523,
+      "step": 15968
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.096843486567493,
+      "learning_rate": 1.1249249184537913e-06,
+      "loss": 0.5879,
+      "step": 15969
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.222385414843731,
+      "learning_rate": 1.1248290667673335e-06,
+      "loss": 0.6953,
+      "step": 15970
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.9042043471230543,
+      "learning_rate": 1.124733213915835e-06,
+      "loss": 0.64,
+      "step": 15971
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.361953733917618,
+      "learning_rate": 1.1246373599001903e-06,
+      "loss": 0.6549,
+      "step": 15972
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.308312301632412,
+      "learning_rate": 1.1245415047212943e-06,
+      "loss": 0.6374,
+      "step": 15973
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.4630636186468324,
+      "learning_rate": 1.124445648380041e-06,
+      "loss": 0.6302,
+      "step": 15974
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.7827649962197976,
+      "learning_rate": 1.1243497908773258e-06,
+      "loss": 0.6003,
+      "step": 15975
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.0841089733464364,
+      "learning_rate": 1.1242539322140433e-06,
+      "loss": 0.5924,
+      "step": 15976
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.832639113902774,
+      "learning_rate": 1.124158072391087e-06,
+      "loss": 0.6562,
+      "step": 15977
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.116939860749533,
+      "learning_rate": 1.1240622114093532e-06,
+      "loss": 0.6273,
+      "step": 15978
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.229429171491874,
+      "learning_rate": 1.1239663492697355e-06,
+      "loss": 0.6393,
+      "step": 15979
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.4140929168467435,
+      "learning_rate": 1.123870485973129e-06,
+      "loss": 0.6035,
+      "step": 15980
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.0849633651683477,
+      "learning_rate": 1.1237746215204282e-06,
+      "loss": 0.638,
+      "step": 15981
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.949226397470719,
+      "learning_rate": 1.1236787559125281e-06,
+      "loss": 0.5892,
+      "step": 15982
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.493257444750188,
+      "learning_rate": 1.123582889150323e-06,
+      "loss": 0.5846,
+      "step": 15983
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.346193044721229,
+      "learning_rate": 1.123487021234708e-06,
+      "loss": 0.64,
+      "step": 15984
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.959407806404358,
+      "learning_rate": 1.1233911521665773e-06,
+      "loss": 0.6914,
+      "step": 15985
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.2826494646437285,
+      "learning_rate": 1.1232952819468266e-06,
+      "loss": 0.6803,
+      "step": 15986
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.547748330237594,
+      "learning_rate": 1.12319941057635e-06,
+      "loss": 0.6289,
+      "step": 15987
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 4.751843283729108,
+      "learning_rate": 1.1231035380560423e-06,
+      "loss": 0.6647,
+      "step": 15988
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.5135935520778547,
+      "learning_rate": 1.1230076643867988e-06,
+      "loss": 0.625,
+      "step": 15989
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 5.19786064295236,
+      "learning_rate": 1.1229117895695134e-06,
+      "loss": 0.6257,
+      "step": 15990
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 3.163834061291581,
+      "learning_rate": 1.1228159136050815e-06,
+      "loss": 0.6484,
+      "step": 15991
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.7905035612720064,
+      "learning_rate": 1.1227200364943978e-06,
+      "loss": 0.6283,
+      "step": 15992
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.2534703283875044,
+      "learning_rate": 1.122624158238357e-06,
+      "loss": 0.638,
+      "step": 15993
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.7566997319918616,
+      "learning_rate": 1.1225282788378544e-06,
+      "loss": 0.679,
+      "step": 15994
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.5881351070549714,
+      "learning_rate": 1.1224323982937845e-06,
+      "loss": 0.6263,
+      "step": 15995
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 7.046149292564972,
+      "learning_rate": 1.122336516607042e-06,
+      "loss": 0.6777,
+      "step": 15996
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.191495725158638,
+      "learning_rate": 1.122240633778522e-06,
+      "loss": 0.6543,
+      "step": 15997
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.02553166109281,
+      "learning_rate": 1.1221447498091194e-06,
+      "loss": 0.6543,
+      "step": 15998
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.477925237376793,
+      "learning_rate": 1.1220488646997295e-06,
+      "loss": 0.625,
+      "step": 15999
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.884157513761074,
+      "learning_rate": 1.1219529784512463e-06,
+      "loss": 0.6562,
+      "step": 16000
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.0065588730448707,
+      "learning_rate": 1.121857091064565e-06,
+      "loss": 0.6276,
+      "step": 16001
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.23295493786255,
+      "learning_rate": 1.1217612025405812e-06,
+      "loss": 0.6771,
+      "step": 16002
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.086968375910149,
+      "learning_rate": 1.121665312880189e-06,
+      "loss": 0.5827,
+      "step": 16003
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.043155370643646,
+      "learning_rate": 1.1215694220842838e-06,
+      "loss": 0.6523,
+      "step": 16004
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.6448588514813585,
+      "learning_rate": 1.1214735301537603e-06,
+      "loss": 0.6348,
+      "step": 16005
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.213952559827501,
+      "learning_rate": 1.1213776370895137e-06,
+      "loss": 0.6419,
+      "step": 16006
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.650437505480731,
+      "learning_rate": 1.1212817428924388e-06,
+      "loss": 0.6237,
+      "step": 16007
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.0581947966255765,
+      "learning_rate": 1.1211858475634304e-06,
+      "loss": 0.6823,
+      "step": 16008
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.5155968354425697,
+      "learning_rate": 1.1210899511033842e-06,
+      "loss": 0.6875,
+      "step": 16009
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.386549504661448,
+      "learning_rate": 1.1209940535131947e-06,
+      "loss": 0.6315,
+      "step": 16010
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.915655688357336,
+      "learning_rate": 1.120898154793757e-06,
+      "loss": 0.6576,
+      "step": 16011
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.220314836028879,
+      "learning_rate": 1.1208022549459658e-06,
+      "loss": 0.6159,
+      "step": 16012
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.099946020111482,
+      "learning_rate": 1.120706353970717e-06,
+      "loss": 0.6458,
+      "step": 16013
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.734100900502355,
+      "learning_rate": 1.1206104518689048e-06,
+      "loss": 0.6432,
+      "step": 16014
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.3274394004280685,
+      "learning_rate": 1.120514548641425e-06,
+      "loss": 0.6523,
+      "step": 16015
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.370699317034133,
+      "learning_rate": 1.1204186442891717e-06,
+      "loss": 0.6641,
+      "step": 16016
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.206739447317396,
+      "learning_rate": 1.1203227388130405e-06,
+      "loss": 0.6178,
+      "step": 16017
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.0331001285566295,
+      "learning_rate": 1.1202268322139269e-06,
+      "loss": 0.6348,
+      "step": 16018
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.9724805114777535,
+      "learning_rate": 1.1201309244927254e-06,
+      "loss": 0.6126,
+      "step": 16019
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.27581916328456,
+      "learning_rate": 1.1200350156503317e-06,
+      "loss": 0.6803,
+      "step": 16020
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.88957525263457,
+      "learning_rate": 1.1199391056876403e-06,
+      "loss": 0.6387,
+      "step": 16021
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.869931868642469,
+      "learning_rate": 1.1198431946055467e-06,
+      "loss": 0.6074,
+      "step": 16022
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 6.045725759492785,
+      "learning_rate": 1.1197472824049462e-06,
+      "loss": 0.6458,
+      "step": 16023
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.88935164115915,
+      "learning_rate": 1.1196513690867334e-06,
+      "loss": 0.6009,
+      "step": 16024
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.7606487753342623,
+      "learning_rate": 1.1195554546518037e-06,
+      "loss": 0.6543,
+      "step": 16025
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.560158258541504,
+      "learning_rate": 1.1194595391010526e-06,
+      "loss": 0.6764,
+      "step": 16026
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.088141965698427,
+      "learning_rate": 1.119363622435375e-06,
+      "loss": 0.6361,
+      "step": 16027
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.7752546848311375,
+      "learning_rate": 1.1192677046556662e-06,
+      "loss": 0.6595,
+      "step": 16028
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.41125274391606,
+      "learning_rate": 1.1191717857628214e-06,
+      "loss": 0.61,
+      "step": 16029
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.1631689351239682,
+      "learning_rate": 1.1190758657577356e-06,
+      "loss": 0.6204,
+      "step": 16030
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.443853966430382,
+      "learning_rate": 1.1189799446413044e-06,
+      "loss": 0.6615,
+      "step": 16031
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.608314763563522,
+      "learning_rate": 1.1188840224144228e-06,
+      "loss": 0.6178,
+      "step": 16032
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.5646241895787294,
+      "learning_rate": 1.118788099077986e-06,
+      "loss": 0.6771,
+      "step": 16033
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.9799535688595595,
+      "learning_rate": 1.1186921746328894e-06,
+      "loss": 0.6497,
+      "step": 16034
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.000872368681424,
+      "learning_rate": 1.1185962490800285e-06,
+      "loss": 0.64,
+      "step": 16035
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.4286238794489177,
+      "learning_rate": 1.118500322420298e-06,
+      "loss": 0.6641,
+      "step": 16036
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.4438528339846624,
+      "learning_rate": 1.1184043946545936e-06,
+      "loss": 0.6296,
+      "step": 16037
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.4083722784136445,
+      "learning_rate": 1.1183084657838107e-06,
+      "loss": 0.696,
+      "step": 16038
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.4318393104438485,
+      "learning_rate": 1.118212535808844e-06,
+      "loss": 0.6602,
+      "step": 16039
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.322258070430384,
+      "learning_rate": 1.1181166047305897e-06,
+      "loss": 0.6191,
+      "step": 16040
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.309777537975936,
+      "learning_rate": 1.1180206725499424e-06,
+      "loss": 0.6611,
+      "step": 16041
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.3367862938638617,
+      "learning_rate": 1.117924739267798e-06,
+      "loss": 0.6322,
+      "step": 16042
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.8888467385844456,
+      "learning_rate": 1.117828804885051e-06,
+      "loss": 0.6445,
+      "step": 16043
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.9183995338337607,
+      "learning_rate": 1.117732869402598e-06,
+      "loss": 0.6283,
+      "step": 16044
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.5974500131339155,
+      "learning_rate": 1.1176369328213336e-06,
+      "loss": 0.6426,
+      "step": 16045
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.130826455416171,
+      "learning_rate": 1.117540995142153e-06,
+      "loss": 0.6615,
+      "step": 16046
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.332951879514919,
+      "learning_rate": 1.1174450563659522e-06,
+      "loss": 0.6257,
+      "step": 16047
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.582548716346851,
+      "learning_rate": 1.1173491164936262e-06,
+      "loss": 0.6172,
+      "step": 16048
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 5.491006833165741,
+      "learning_rate": 1.1172531755260703e-06,
+      "loss": 0.696,
+      "step": 16049
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.9653946871430894,
+      "learning_rate": 1.1171572334641802e-06,
+      "loss": 0.6367,
+      "step": 16050
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.573852627950711,
+      "learning_rate": 1.1170612903088513e-06,
+      "loss": 0.6829,
+      "step": 16051
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.6273692195634855,
+      "learning_rate": 1.1169653460609793e-06,
+      "loss": 0.6569,
+      "step": 16052
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.302592504127437,
+      "learning_rate": 1.1168694007214593e-06,
+      "loss": 0.623,
+      "step": 16053
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 4.544258025597939,
+      "learning_rate": 1.1167734542911867e-06,
+      "loss": 0.6289,
+      "step": 16054
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.8682192691422275,
+      "learning_rate": 1.1166775067710572e-06,
+      "loss": 0.6556,
+      "step": 16055
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.620803959607594,
+      "learning_rate": 1.1165815581619663e-06,
+      "loss": 0.6556,
+      "step": 16056
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 3.6310827364992715,
+      "learning_rate": 1.116485608464809e-06,
+      "loss": 0.653,
+      "step": 16057
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.6722177643468394,
+      "learning_rate": 1.1163896576804818e-06,
+      "loss": 0.666,
+      "step": 16058
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.628701003502692,
+      "learning_rate": 1.116293705809879e-06,
+      "loss": 0.6608,
+      "step": 16059
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.4792867342717377,
+      "learning_rate": 1.1161977528538972e-06,
+      "loss": 0.6302,
+      "step": 16060
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.457442902616609,
+      "learning_rate": 1.1161017988134313e-06,
+      "loss": 0.584,
+      "step": 16061
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.319661045176083,
+      "learning_rate": 1.1160058436893774e-06,
+      "loss": 0.651,
+      "step": 16062
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.7762898642342115,
+      "learning_rate": 1.1159098874826304e-06,
+      "loss": 0.6719,
+      "step": 16063
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.925049043107803,
+      "learning_rate": 1.115813930194086e-06,
+      "loss": 0.6842,
+      "step": 16064
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.9731440515808654,
+      "learning_rate": 1.1157179718246406e-06,
+      "loss": 0.6191,
+      "step": 16065
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.489855041459923,
+      "learning_rate": 1.1156220123751886e-06,
+      "loss": 0.623,
+      "step": 16066
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.4978096043321,
+      "learning_rate": 1.115526051846626e-06,
+      "loss": 0.6204,
+      "step": 16067
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.7531943832651327,
+      "learning_rate": 1.115430090239849e-06,
+      "loss": 0.6458,
+      "step": 16068
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.7074860618754304,
+      "learning_rate": 1.1153341275557525e-06,
+      "loss": 0.5788,
+      "step": 16069
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.109046785666444,
+      "learning_rate": 1.1152381637952324e-06,
+      "loss": 0.6289,
+      "step": 16070
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.892918346553909,
+      "learning_rate": 1.1151421989591842e-06,
+      "loss": 0.6751,
+      "step": 16071
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.0026069920595737,
+      "learning_rate": 1.115046233048504e-06,
+      "loss": 0.6686,
+      "step": 16072
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.8462673658940525,
+      "learning_rate": 1.114950266064087e-06,
+      "loss": 0.6074,
+      "step": 16073
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.6181908238906124,
+      "learning_rate": 1.1148542980068287e-06,
+      "loss": 0.6188,
+      "step": 16074
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.570680211376234,
+      "learning_rate": 1.1147583288776254e-06,
+      "loss": 0.623,
+      "step": 16075
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.109927371591423,
+      "learning_rate": 1.1146623586773724e-06,
+      "loss": 0.6341,
+      "step": 16076
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.809847481773316,
+      "learning_rate": 1.1145663874069655e-06,
+      "loss": 0.6764,
+      "step": 16077
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.25937442640351,
+      "learning_rate": 1.1144704150673004e-06,
+      "loss": 0.6016,
+      "step": 16078
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.177219241621853,
+      "learning_rate": 1.1143744416592729e-06,
+      "loss": 0.653,
+      "step": 16079
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.8442636019896956,
+      "learning_rate": 1.1142784671837782e-06,
+      "loss": 0.6335,
+      "step": 16080
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.7151273901560367,
+      "learning_rate": 1.114182491641713e-06,
+      "loss": 0.6523,
+      "step": 16081
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.0221392589290828,
+      "learning_rate": 1.114086515033972e-06,
+      "loss": 0.6712,
+      "step": 16082
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.0492546095904043,
+      "learning_rate": 1.1139905373614515e-06,
+      "loss": 0.6283,
+      "step": 16083
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.608583685624029,
+      "learning_rate": 1.1138945586250476e-06,
+      "loss": 0.6569,
+      "step": 16084
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.696945914182056,
+      "learning_rate": 1.1137985788256557e-06,
+      "loss": 0.6758,
+      "step": 16085
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.9129281218208187,
+      "learning_rate": 1.1137025979641715e-06,
+      "loss": 0.6576,
+      "step": 16086
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.772586364488629,
+      "learning_rate": 1.1136066160414905e-06,
+      "loss": 0.6165,
+      "step": 16087
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.970791043798506,
+      "learning_rate": 1.1135106330585094e-06,
+      "loss": 0.6536,
+      "step": 16088
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.8888371102544594,
+      "learning_rate": 1.1134146490161236e-06,
+      "loss": 0.6146,
+      "step": 16089
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.1587352835903366,
+      "learning_rate": 1.1133186639152282e-06,
+      "loss": 0.6328,
+      "step": 16090
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.1056294860043985,
+      "learning_rate": 1.1132226777567204e-06,
+      "loss": 0.6771,
+      "step": 16091
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.037284597149811,
+      "learning_rate": 1.113126690541495e-06,
+      "loss": 0.6348,
+      "step": 16092
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.792537725471638,
+      "learning_rate": 1.113030702270448e-06,
+      "loss": 0.6325,
+      "step": 16093
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.131600525668189,
+      "learning_rate": 1.112934712944476e-06,
+      "loss": 0.6159,
+      "step": 16094
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.1567915833800346,
+      "learning_rate": 1.112838722564474e-06,
+      "loss": 0.6107,
+      "step": 16095
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.457384362619689,
+      "learning_rate": 1.112742731131338e-06,
+      "loss": 0.6263,
+      "step": 16096
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.024693205730018,
+      "learning_rate": 1.1126467386459646e-06,
+      "loss": 0.6465,
+      "step": 16097
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.533287469315511,
+      "learning_rate": 1.1125507451092491e-06,
+      "loss": 0.6243,
+      "step": 16098
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.2348750096225474,
+      "learning_rate": 1.1124547505220873e-06,
+      "loss": 0.6419,
+      "step": 16099
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.9108850745921164,
+      "learning_rate": 1.1123587548853754e-06,
+      "loss": 0.6751,
+      "step": 16100
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.346490961588511,
+      "learning_rate": 1.1122627582000094e-06,
+      "loss": 0.5853,
+      "step": 16101
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.021360763139583,
+      "learning_rate": 1.1121667604668854e-06,
+      "loss": 0.6432,
+      "step": 16102
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.4183334169343387,
+      "learning_rate": 1.1120707616868987e-06,
+      "loss": 0.6523,
+      "step": 16103
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.2289521170082773,
+      "learning_rate": 1.1119747618609458e-06,
+      "loss": 0.6693,
+      "step": 16104
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.111777379747224,
+      "learning_rate": 1.1118787609899227e-06,
+      "loss": 0.651,
+      "step": 16105
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.268333703581339,
+      "learning_rate": 1.1117827590747252e-06,
+      "loss": 0.6217,
+      "step": 16106
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.478071334037668,
+      "learning_rate": 1.111686756116249e-06,
+      "loss": 0.5863,
+      "step": 16107
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.445358166183908,
+      "learning_rate": 1.1115907521153909e-06,
+      "loss": 0.64,
+      "step": 16108
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.1160696466125186,
+      "learning_rate": 1.111494747073046e-06,
+      "loss": 0.6035,
+      "step": 16109
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.098171483568357,
+      "learning_rate": 1.1113987409901112e-06,
+      "loss": 0.651,
+      "step": 16110
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.2549449723257218,
+      "learning_rate": 1.111302733867482e-06,
+      "loss": 0.6576,
+      "step": 16111
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.1866334996270003,
+      "learning_rate": 1.111206725706054e-06,
+      "loss": 0.6875,
+      "step": 16112
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.2963936269469873,
+      "learning_rate": 1.1111107165067243e-06,
+      "loss": 0.6406,
+      "step": 16113
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 4.156253028028251,
+      "learning_rate": 1.1110147062703884e-06,
+      "loss": 0.6055,
+      "step": 16114
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.0983786821156243,
+      "learning_rate": 1.1109186949979425e-06,
+      "loss": 0.6615,
+      "step": 16115
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.109273505085939,
+      "learning_rate": 1.1108226826902824e-06,
+      "loss": 0.7044,
+      "step": 16116
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.4205672944452794,
+      "learning_rate": 1.1107266693483045e-06,
+      "loss": 0.6621,
+      "step": 16117
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.93497811568179,
+      "learning_rate": 1.110630654972905e-06,
+      "loss": 0.599,
+      "step": 16118
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.721290824161096,
+      "learning_rate": 1.1105346395649794e-06,
+      "loss": 0.653,
+      "step": 16119
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.6098812727512843,
+      "learning_rate": 1.1104386231254246e-06,
+      "loss": 0.6126,
+      "step": 16120
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 6.470414672864971,
+      "learning_rate": 1.1103426056551362e-06,
+      "loss": 0.6289,
+      "step": 16121
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 7.76254598681364,
+      "learning_rate": 1.1102465871550105e-06,
+      "loss": 0.6491,
+      "step": 16122
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.2776089126451877,
+      "learning_rate": 1.1101505676259437e-06,
+      "loss": 0.6237,
+      "step": 16123
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.145882994185805,
+      "learning_rate": 1.1100545470688317e-06,
+      "loss": 0.6484,
+      "step": 16124
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 5.104109444245109,
+      "learning_rate": 1.1099585254845712e-06,
+      "loss": 0.6367,
+      "step": 16125
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.784042391310777,
+      "learning_rate": 1.109862502874058e-06,
+      "loss": 0.6387,
+      "step": 16126
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.1805952272187836,
+      "learning_rate": 1.109766479238188e-06,
+      "loss": 0.5833,
+      "step": 16127
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.419156216826375,
+      "learning_rate": 1.109670454577858e-06,
+      "loss": 0.597,
+      "step": 16128
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.694927049689944,
+      "learning_rate": 1.1095744288939642e-06,
+      "loss": 0.6322,
+      "step": 16129
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.8149108670565597,
+      "learning_rate": 1.1094784021874022e-06,
+      "loss": 0.623,
+      "step": 16130
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.5465843319570896,
+      "learning_rate": 1.109382374459069e-06,
+      "loss": 0.6413,
+      "step": 16131
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.206310401785116,
+      "learning_rate": 1.10928634570986e-06,
+      "loss": 0.6243,
+      "step": 16132
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.523610159236279,
+      "learning_rate": 1.1091903159406716e-06,
+      "loss": 0.5931,
+      "step": 16133
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.426689057464205,
+      "learning_rate": 1.1090942851524012e-06,
+      "loss": 0.6719,
+      "step": 16134
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.550485808956085,
+      "learning_rate": 1.1089982533459437e-06,
+      "loss": 0.6341,
+      "step": 16135
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.22077015244825,
+      "learning_rate": 1.108902220522196e-06,
+      "loss": 0.611,
+      "step": 16136
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.1212141190392693,
+      "learning_rate": 1.1088061866820537e-06,
+      "loss": 0.6517,
+      "step": 16137
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.6932200526493277,
+      "learning_rate": 1.1087101518264144e-06,
+      "loss": 0.6738,
+      "step": 16138
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.3678866623471526,
+      "learning_rate": 1.1086141159561733e-06,
+      "loss": 0.6406,
+      "step": 16139
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.471805719338849,
+      "learning_rate": 1.1085180790722268e-06,
+      "loss": 0.6699,
+      "step": 16140
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.878813440108217,
+      "learning_rate": 1.1084220411754718e-06,
+      "loss": 0.707,
+      "step": 16141
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.8771446254875785,
+      "learning_rate": 1.1083260022668043e-06,
+      "loss": 0.6478,
+      "step": 16142
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.1423557783566345,
+      "learning_rate": 1.1082299623471205e-06,
+      "loss": 0.6257,
+      "step": 16143
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.6192449345676176,
+      "learning_rate": 1.1081339214173172e-06,
+      "loss": 0.6634,
+      "step": 16144
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.056715115322761,
+      "learning_rate": 1.1080378794782902e-06,
+      "loss": 0.6458,
+      "step": 16145
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.7501126463989127,
+      "learning_rate": 1.1079418365309359e-06,
+      "loss": 0.6335,
+      "step": 16146
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.675256133109593,
+      "learning_rate": 1.107845792576151e-06,
+      "loss": 0.5794,
+      "step": 16147
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.1738217754833395,
+      "learning_rate": 1.1077497476148321e-06,
+      "loss": 0.6569,
+      "step": 16148
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.8909742493649278,
+      "learning_rate": 1.1076537016478748e-06,
+      "loss": 0.6074,
+      "step": 16149
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.952110884279853,
+      "learning_rate": 1.1075576546761764e-06,
+      "loss": 0.6374,
+      "step": 16150
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.0462029502589965,
+      "learning_rate": 1.1074616067006329e-06,
+      "loss": 0.6048,
+      "step": 16151
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.146374743840393,
+      "learning_rate": 1.1073655577221406e-06,
+      "loss": 0.6094,
+      "step": 16152
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.6056987756239787,
+      "learning_rate": 1.107269507741596e-06,
+      "loss": 0.6673,
+      "step": 16153
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.586788531421628,
+      "learning_rate": 1.1071734567598955e-06,
+      "loss": 0.6953,
+      "step": 16154
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.4878436269578197,
+      "learning_rate": 1.1070774047779362e-06,
+      "loss": 0.6523,
+      "step": 16155
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.443551079600942,
+      "learning_rate": 1.1069813517966133e-06,
+      "loss": 0.612,
+      "step": 16156
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.839815272856742,
+      "learning_rate": 1.1068852978168244e-06,
+      "loss": 0.5729,
+      "step": 16157
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.214225851403352,
+      "learning_rate": 1.1067892428394654e-06,
+      "loss": 0.6165,
+      "step": 16158
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.2265590549103798,
+      "learning_rate": 1.1066931868654328e-06,
+      "loss": 0.6165,
+      "step": 16159
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.5306549531350315,
+      "learning_rate": 1.1065971298956236e-06,
+      "loss": 0.6465,
+      "step": 16160
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.7115687824432744,
+      "learning_rate": 1.1065010719309338e-06,
+      "loss": 0.6823,
+      "step": 16161
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.285568869994458,
+      "learning_rate": 1.1064050129722599e-06,
+      "loss": 0.6725,
+      "step": 16162
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.721479183307361,
+      "learning_rate": 1.106308953020499e-06,
+      "loss": 0.653,
+      "step": 16163
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.122746997596649,
+      "learning_rate": 1.106212892076547e-06,
+      "loss": 0.6549,
+      "step": 16164
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.2867862077414447,
+      "learning_rate": 1.106116830141301e-06,
+      "loss": 0.6478,
+      "step": 16165
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.02516919687932,
+      "learning_rate": 1.1060207672156564e-06,
+      "loss": 0.6081,
+      "step": 16166
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.065917041858896,
+      "learning_rate": 1.1059247033005114e-06,
+      "loss": 0.6302,
+      "step": 16167
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.769705302767921,
+      "learning_rate": 1.1058286383967616e-06,
+      "loss": 0.6667,
+      "step": 16168
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.3652364991495336,
+      "learning_rate": 1.1057325725053035e-06,
+      "loss": 0.6901,
+      "step": 16169
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.322921039444759,
+      "learning_rate": 1.1056365056270342e-06,
+      "loss": 0.6185,
+      "step": 16170
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.9182572649189167,
+      "learning_rate": 1.10554043776285e-06,
+      "loss": 0.707,
+      "step": 16171
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.5405346450966335,
+      "learning_rate": 1.1054443689136475e-06,
+      "loss": 0.6745,
+      "step": 16172
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 5.2621690774010625,
+      "learning_rate": 1.1053482990803235e-06,
+      "loss": 0.6973,
+      "step": 16173
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.771418625633773,
+      "learning_rate": 1.1052522282637747e-06,
+      "loss": 0.6471,
+      "step": 16174
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.271025790962827,
+      "learning_rate": 1.105156156464897e-06,
+      "loss": 0.6842,
+      "step": 16175
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.518470712156387,
+      "learning_rate": 1.1050600836845879e-06,
+      "loss": 0.6335,
+      "step": 16176
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.4817669410357404,
+      "learning_rate": 1.1049640099237439e-06,
+      "loss": 0.5768,
+      "step": 16177
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.6494382550073055,
+      "learning_rate": 1.1048679351832611e-06,
+      "loss": 0.666,
+      "step": 16178
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.5002146257825992,
+      "learning_rate": 1.104771859464037e-06,
+      "loss": 0.6608,
+      "step": 16179
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.119049057913742,
+      "learning_rate": 1.1046757827669674e-06,
+      "loss": 0.6849,
+      "step": 16180
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.7584972848147795,
+      "learning_rate": 1.10457970509295e-06,
+      "loss": 0.6257,
+      "step": 16181
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.7850878289544325,
+      "learning_rate": 1.1044836264428808e-06,
+      "loss": 0.666,
+      "step": 16182
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.3843156169647566,
+      "learning_rate": 1.1043875468176563e-06,
+      "loss": 0.6165,
+      "step": 16183
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.094403122957257,
+      "learning_rate": 1.1042914662181742e-06,
+      "loss": 0.6048,
+      "step": 16184
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.013973769845632,
+      "learning_rate": 1.10419538464533e-06,
+      "loss": 0.666,
+      "step": 16185
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.3796108324547363,
+      "learning_rate": 1.1040993021000216e-06,
+      "loss": 0.6178,
+      "step": 16186
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.540014950275787,
+      "learning_rate": 1.1040032185831449e-06,
+      "loss": 0.6374,
+      "step": 16187
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 3.7771911670116323,
+      "learning_rate": 1.103907134095597e-06,
+      "loss": 0.6855,
+      "step": 16188
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.0460338720699065,
+      "learning_rate": 1.1038110486382749e-06,
+      "loss": 0.6947,
+      "step": 16189
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.677394153356149,
+      "learning_rate": 1.1037149622120746e-06,
+      "loss": 0.6022,
+      "step": 16190
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 4.031349011947018,
+      "learning_rate": 1.1036188748178937e-06,
+      "loss": 0.6224,
+      "step": 16191
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.513396735277156,
+      "learning_rate": 1.1035227864566289e-06,
+      "loss": 0.6029,
+      "step": 16192
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.648816118159312,
+      "learning_rate": 1.1034266971291762e-06,
+      "loss": 0.6289,
+      "step": 16193
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.005271997663111,
+      "learning_rate": 1.1033306068364333e-06,
+      "loss": 0.6361,
+      "step": 16194
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.755911477149863,
+      "learning_rate": 1.1032345155792969e-06,
+      "loss": 0.6341,
+      "step": 16195
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.8557220187699777,
+      "learning_rate": 1.1031384233586632e-06,
+      "loss": 0.6699,
+      "step": 16196
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.116158875397881,
+      "learning_rate": 1.1030423301754301e-06,
+      "loss": 0.6549,
+      "step": 16197
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.384526077141439,
+      "learning_rate": 1.1029462360304933e-06,
+      "loss": 0.6152,
+      "step": 16198
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.846512272221521,
+      "learning_rate": 1.1028501409247502e-06,
+      "loss": 0.6686,
+      "step": 16199
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.541131141217938,
+      "learning_rate": 1.1027540448590979e-06,
+      "loss": 0.6556,
+      "step": 16200
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.633037803364924,
+      "learning_rate": 1.1026579478344328e-06,
+      "loss": 0.6784,
+      "step": 16201
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.720346961225457,
+      "learning_rate": 1.102561849851652e-06,
+      "loss": 0.6595,
+      "step": 16202
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.534629859162402,
+      "learning_rate": 1.1024657509116523e-06,
+      "loss": 0.6048,
+      "step": 16203
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.959878757358566,
+      "learning_rate": 1.102369651015331e-06,
+      "loss": 0.6123,
+      "step": 16204
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.953136894234399,
+      "learning_rate": 1.1022735501635847e-06,
+      "loss": 0.6699,
+      "step": 16205
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.360232134382735,
+      "learning_rate": 1.1021774483573098e-06,
+      "loss": 0.6523,
+      "step": 16206
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.870207450309303,
+      "learning_rate": 1.1020813455974043e-06,
+      "loss": 0.6172,
+      "step": 16207
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.136728706131551,
+      "learning_rate": 1.1019852418847644e-06,
+      "loss": 0.681,
+      "step": 16208
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.431088926542497,
+      "learning_rate": 1.1018891372202869e-06,
+      "loss": 0.6517,
+      "step": 16209
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.4640254963097767,
+      "learning_rate": 1.1017930316048696e-06,
+      "loss": 0.6621,
+      "step": 16210
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.8826023370729996,
+      "learning_rate": 1.1016969250394086e-06,
+      "loss": 0.6289,
+      "step": 16211
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.9251996836481444,
+      "learning_rate": 1.1016008175248011e-06,
+      "loss": 0.6302,
+      "step": 16212
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.249268248990979,
+      "learning_rate": 1.1015047090619446e-06,
+      "loss": 0.5931,
+      "step": 16213
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.083647012558321,
+      "learning_rate": 1.1014085996517354e-06,
+      "loss": 0.6198,
+      "step": 16214
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.9029387173816765,
+      "learning_rate": 1.1013124892950707e-06,
+      "loss": 0.6536,
+      "step": 16215
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.0703260603389686,
+      "learning_rate": 1.101216377992848e-06,
+      "loss": 0.6211,
+      "step": 16216
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.619636424248097,
+      "learning_rate": 1.1011202657459633e-06,
+      "loss": 0.6497,
+      "step": 16217
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.127611482869359,
+      "learning_rate": 1.1010241525553145e-06,
+      "loss": 0.5957,
+      "step": 16218
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.3068200118206774,
+      "learning_rate": 1.1009280384217982e-06,
+      "loss": 0.6908,
+      "step": 16219
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.557049396578043,
+      "learning_rate": 1.100831923346312e-06,
+      "loss": 0.6921,
+      "step": 16220
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.680039186971334,
+      "learning_rate": 1.1007358073297523e-06,
+      "loss": 0.6387,
+      "step": 16221
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.900498216604118,
+      "learning_rate": 1.1006396903730162e-06,
+      "loss": 0.6016,
+      "step": 16222
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.232324685298537,
+      "learning_rate": 1.1005435724770013e-06,
+      "loss": 0.6198,
+      "step": 16223
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.9400945855978566,
+      "learning_rate": 1.1004474536426043e-06,
+      "loss": 0.6198,
+      "step": 16224
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.619092659014568,
+      "learning_rate": 1.100351333870722e-06,
+      "loss": 0.6621,
+      "step": 16225
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.7310670760841207,
+      "learning_rate": 1.1002552131622523e-06,
+      "loss": 0.6152,
+      "step": 16226
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.710773906966693,
+      "learning_rate": 1.1001590915180915e-06,
+      "loss": 0.6986,
+      "step": 16227
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.820845447331387,
+      "learning_rate": 1.100062968939137e-06,
+      "loss": 0.6289,
+      "step": 16228
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.698851721903911,
+      "learning_rate": 1.0999668454262862e-06,
+      "loss": 0.6406,
+      "step": 16229
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.003222252348611,
+      "learning_rate": 1.099870720980436e-06,
+      "loss": 0.6123,
+      "step": 16230
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.218076808759493,
+      "learning_rate": 1.0997745956024834e-06,
+      "loss": 0.6152,
+      "step": 16231
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.2063671029342853,
+      "learning_rate": 1.0996784692933254e-06,
+      "loss": 0.5944,
+      "step": 16232
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.863605541177366,
+      "learning_rate": 1.09958234205386e-06,
+      "loss": 0.6458,
+      "step": 16233
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.181892653573195,
+      "learning_rate": 1.0994862138849835e-06,
+      "loss": 0.623,
+      "step": 16234
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.279731258041385,
+      "learning_rate": 1.0993900847875933e-06,
+      "loss": 0.6029,
+      "step": 16235
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.9841900299123894,
+      "learning_rate": 1.0992939547625868e-06,
+      "loss": 0.6556,
+      "step": 16236
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.537177919730715,
+      "learning_rate": 1.099197823810861e-06,
+      "loss": 0.6582,
+      "step": 16237
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.3181160922505333,
+      "learning_rate": 1.0991016919333129e-06,
+      "loss": 0.6289,
+      "step": 16238
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.7979001713987475,
+      "learning_rate": 1.0990055591308403e-06,
+      "loss": 0.5771,
+      "step": 16239
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.3973225495514705,
+      "learning_rate": 1.0989094254043398e-06,
+      "loss": 0.6452,
+      "step": 16240
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.516983623006026,
+      "learning_rate": 1.098813290754709e-06,
+      "loss": 0.6113,
+      "step": 16241
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.171010165900602,
+      "learning_rate": 1.0987171551828452e-06,
+      "loss": 0.6022,
+      "step": 16242
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.418820654271641,
+      "learning_rate": 1.098621018689645e-06,
+      "loss": 0.6875,
+      "step": 16243
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.723969779387697,
+      "learning_rate": 1.0985248812760064e-06,
+      "loss": 0.6621,
+      "step": 16244
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.134400999533632,
+      "learning_rate": 1.0984287429428265e-06,
+      "loss": 0.6582,
+      "step": 16245
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.6276698639047447,
+      "learning_rate": 1.0983326036910022e-06,
+      "loss": 0.6576,
+      "step": 16246
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.4965730024125925,
+      "learning_rate": 1.0982364635214313e-06,
+      "loss": 0.679,
+      "step": 16247
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.5778063544134273,
+      "learning_rate": 1.0981403224350106e-06,
+      "loss": 0.666,
+      "step": 16248
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.520560714689388,
+      "learning_rate": 1.0980441804326372e-06,
+      "loss": 0.7507,
+      "step": 16249
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.2343033005989503,
+      "learning_rate": 1.0979480375152097e-06,
+      "loss": 0.6458,
+      "step": 16250
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.2960960117162195,
+      "learning_rate": 1.0978518936836238e-06,
+      "loss": 0.6237,
+      "step": 16251
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.8330284184704366,
+      "learning_rate": 1.097755748938778e-06,
+      "loss": 0.6445,
+      "step": 16252
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.737859188421782,
+      "learning_rate": 1.0976596032815688e-06,
+      "loss": 0.6022,
+      "step": 16253
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 5.166403881364413,
+      "learning_rate": 1.0975634567128937e-06,
+      "loss": 0.6276,
+      "step": 16254
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 6.1172586798019015,
+      "learning_rate": 1.0974673092336506e-06,
+      "loss": 0.7253,
+      "step": 16255
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.9622524891070445,
+      "learning_rate": 1.0973711608447363e-06,
+      "loss": 0.6302,
+      "step": 16256
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.99547947162132,
+      "learning_rate": 1.0972750115470483e-06,
+      "loss": 0.6908,
+      "step": 16257
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 4.310000470408356,
+      "learning_rate": 1.0971788613414842e-06,
+      "loss": 0.6484,
+      "step": 16258
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 3.048441843511526,
+      "learning_rate": 1.0970827102289412e-06,
+      "loss": 0.6556,
+      "step": 16259
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.7160967308869717,
+      "learning_rate": 1.0969865582103166e-06,
+      "loss": 0.6426,
+      "step": 16260
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.4742486554210217,
+      "learning_rate": 1.0968904052865077e-06,
+      "loss": 0.6185,
+      "step": 16261
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.590785839126898,
+      "learning_rate": 1.0967942514584124e-06,
+      "loss": 0.6458,
+      "step": 16262
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.637189318598999,
+      "learning_rate": 1.0966980967269278e-06,
+      "loss": 0.5814,
+      "step": 16263
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.684753932982316,
+      "learning_rate": 1.096601941092951e-06,
+      "loss": 0.651,
+      "step": 16264
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.5645232523278185,
+      "learning_rate": 1.0965057845573797e-06,
+      "loss": 0.668,
+      "step": 16265
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.146638056992002,
+      "learning_rate": 1.0964096271211116e-06,
+      "loss": 0.6309,
+      "step": 16266
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.054995515632718,
+      "learning_rate": 1.0963134687850443e-06,
+      "loss": 0.6328,
+      "step": 16267
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.0758390342298405,
+      "learning_rate": 1.0962173095500744e-06,
+      "loss": 0.6621,
+      "step": 16268
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 6.015995893488659,
+      "learning_rate": 1.0961211494170997e-06,
+      "loss": 0.6706,
+      "step": 16269
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.8492153178923134,
+      "learning_rate": 1.096024988387018e-06,
+      "loss": 0.6335,
+      "step": 16270
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.224736512370168,
+      "learning_rate": 1.0959288264607268e-06,
+      "loss": 0.6074,
+      "step": 16271
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.149821148493789,
+      "learning_rate": 1.0958326636391232e-06,
+      "loss": 0.6406,
+      "step": 16272
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.9663861415839796,
+      "learning_rate": 1.0957364999231049e-06,
+      "loss": 0.6367,
+      "step": 16273
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.779477897805329,
+      "learning_rate": 1.0956403353135693e-06,
+      "loss": 0.6107,
+      "step": 16274
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.2337194390336306,
+      "learning_rate": 1.095544169811414e-06,
+      "loss": 0.6914,
+      "step": 16275
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.6755465926264943,
+      "learning_rate": 1.0954480034175367e-06,
+      "loss": 0.6654,
+      "step": 16276
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.430746067746198,
+      "learning_rate": 1.0953518361328344e-06,
+      "loss": 0.681,
+      "step": 16277
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.839158563562915,
+      "learning_rate": 1.0952556679582053e-06,
+      "loss": 0.6354,
+      "step": 16278
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.445400867019921,
+      "learning_rate": 1.0951594988945466e-06,
+      "loss": 0.5742,
+      "step": 16279
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.5925337807861126,
+      "learning_rate": 1.0950633289427558e-06,
+      "loss": 0.6263,
+      "step": 16280
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.10190441777177,
+      "learning_rate": 1.0949671581037304e-06,
+      "loss": 0.6693,
+      "step": 16281
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.9018236894506018,
+      "learning_rate": 1.0948709863783682e-06,
+      "loss": 0.6706,
+      "step": 16282
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.108010632442052,
+      "learning_rate": 1.094774813767567e-06,
+      "loss": 0.6055,
+      "step": 16283
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.874359275801509,
+      "learning_rate": 1.0946786402722236e-06,
+      "loss": 0.6257,
+      "step": 16284
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.749806580189776,
+      "learning_rate": 1.0945824658932364e-06,
+      "loss": 0.6615,
+      "step": 16285
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.3172097505160716,
+      "learning_rate": 1.0944862906315028e-06,
+      "loss": 0.6458,
+      "step": 16286
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.1069815100283065,
+      "learning_rate": 1.09439011448792e-06,
+      "loss": 0.666,
+      "step": 16287
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.117939136843455,
+      "learning_rate": 1.0942939374633857e-06,
+      "loss": 0.6478,
+      "step": 16288
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.773664191045305,
+      "learning_rate": 1.0941977595587983e-06,
+      "loss": 0.651,
+      "step": 16289
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.4909785588902986,
+      "learning_rate": 1.0941015807750547e-06,
+      "loss": 0.6712,
+      "step": 16290
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.270424012178166,
+      "learning_rate": 1.0940054011130525e-06,
+      "loss": 0.6602,
+      "step": 16291
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.020142900387604,
+      "learning_rate": 1.0939092205736897e-06,
+      "loss": 0.6582,
+      "step": 16292
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.7958589722438787,
+      "learning_rate": 1.0938130391578638e-06,
+      "loss": 0.6491,
+      "step": 16293
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.7427214741959434,
+      "learning_rate": 1.0937168568664725e-06,
+      "loss": 0.6693,
+      "step": 16294
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.9709493071947772,
+      "learning_rate": 1.0936206737004136e-06,
+      "loss": 0.5944,
+      "step": 16295
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.685297698589531,
+      "learning_rate": 1.093524489660585e-06,
+      "loss": 0.6504,
+      "step": 16296
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.32228871693566,
+      "learning_rate": 1.0934283047478837e-06,
+      "loss": 0.6478,
+      "step": 16297
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.992250198129525,
+      "learning_rate": 1.0933321189632073e-06,
+      "loss": 0.6497,
+      "step": 16298
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.7333144274467576,
+      "learning_rate": 1.0932359323074544e-06,
+      "loss": 0.6921,
+      "step": 16299
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.8383191135999626,
+      "learning_rate": 1.0931397447815227e-06,
+      "loss": 0.6081,
+      "step": 16300
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.6598641959003406,
+      "learning_rate": 1.093043556386309e-06,
+      "loss": 0.6087,
+      "step": 16301
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.000603901105713,
+      "learning_rate": 1.092947367122712e-06,
+      "loss": 0.6243,
+      "step": 16302
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.075104597478175,
+      "learning_rate": 1.0928511769916288e-06,
+      "loss": 0.5996,
+      "step": 16303
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.051662844363592,
+      "learning_rate": 1.092754985993957e-06,
+      "loss": 0.6426,
+      "step": 16304
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.110144956732667,
+      "learning_rate": 1.092658794130595e-06,
+      "loss": 0.6198,
+      "step": 16305
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.251930090849647,
+      "learning_rate": 1.0925626014024405e-06,
+      "loss": 0.668,
+      "step": 16306
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.325094902419621,
+      "learning_rate": 1.0924664078103907e-06,
+      "loss": 0.6328,
+      "step": 16307
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.8812876184946674,
+      "learning_rate": 1.0923702133553438e-06,
+      "loss": 0.6016,
+      "step": 16308
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.740947567560114,
+      "learning_rate": 1.0922740180381977e-06,
+      "loss": 0.6113,
+      "step": 16309
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.603343751898623,
+      "learning_rate": 1.0921778218598498e-06,
+      "loss": 0.6732,
+      "step": 16310
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 5.157405236508481,
+      "learning_rate": 1.0920816248211984e-06,
+      "loss": 0.6126,
+      "step": 16311
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.7526992484302557,
+      "learning_rate": 1.0919854269231407e-06,
+      "loss": 0.6634,
+      "step": 16312
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.336500327621149,
+      "learning_rate": 1.0918892281665756e-06,
+      "loss": 0.6107,
+      "step": 16313
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.750431572880712,
+      "learning_rate": 1.0917930285523993e-06,
+      "loss": 0.5977,
+      "step": 16314
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.1967485667564994,
+      "learning_rate": 1.0916968280815108e-06,
+      "loss": 0.6309,
+      "step": 16315
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.340962817892974,
+      "learning_rate": 1.0916006267548082e-06,
+      "loss": 0.6263,
+      "step": 16316
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.6315349286721648,
+      "learning_rate": 1.0915044245731883e-06,
+      "loss": 0.6445,
+      "step": 16317
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.529691816420575,
+      "learning_rate": 1.0914082215375497e-06,
+      "loss": 0.668,
+      "step": 16318
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.1039263296531603,
+      "learning_rate": 1.0913120176487903e-06,
+      "loss": 0.6348,
+      "step": 16319
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.109587591698981,
+      "learning_rate": 1.0912158129078074e-06,
+      "loss": 0.6191,
+      "step": 16320
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.5489291834942165,
+      "learning_rate": 1.0911196073154995e-06,
+      "loss": 0.6973,
+      "step": 16321
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.0507792494955948,
+      "learning_rate": 1.0910234008727639e-06,
+      "loss": 0.61,
+      "step": 16322
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 7.227747831310424,
+      "learning_rate": 1.0909271935804993e-06,
+      "loss": 0.6348,
+      "step": 16323
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.504351599181026,
+      "learning_rate": 1.090830985439603e-06,
+      "loss": 0.6758,
+      "step": 16324
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.580627731678024,
+      "learning_rate": 1.090734776450973e-06,
+      "loss": 0.7005,
+      "step": 16325
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 4.120880154798373,
+      "learning_rate": 1.0906385666155074e-06,
+      "loss": 0.6582,
+      "step": 16326
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.620476988232404,
+      "learning_rate": 1.0905423559341041e-06,
+      "loss": 0.6784,
+      "step": 16327
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 6.848483543778698,
+      "learning_rate": 1.0904461444076607e-06,
+      "loss": 0.6419,
+      "step": 16328
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.088885249639158,
+      "learning_rate": 1.090349932037076e-06,
+      "loss": 0.6784,
+      "step": 16329
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.402394284707644,
+      "learning_rate": 1.0902537188232468e-06,
+      "loss": 0.625,
+      "step": 16330
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.881104112893777,
+      "learning_rate": 1.0901575047670717e-06,
+      "loss": 0.6562,
+      "step": 16331
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.443533750411454,
+      "learning_rate": 1.0900612898694489e-06,
+      "loss": 0.6348,
+      "step": 16332
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.8089016533068816,
+      "learning_rate": 1.089965074131276e-06,
+      "loss": 0.6022,
+      "step": 16333
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1785308219446646,
+      "learning_rate": 1.089868857553451e-06,
+      "loss": 0.6107,
+      "step": 16334
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.9650354772437635,
+      "learning_rate": 1.089772640136872e-06,
+      "loss": 0.6126,
+      "step": 16335
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.5930329838409616,
+      "learning_rate": 1.0896764218824375e-06,
+      "loss": 0.6198,
+      "step": 16336
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.670980069448908,
+      "learning_rate": 1.0895802027910446e-06,
+      "loss": 0.6387,
+      "step": 16337
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.4896656528191055,
+      "learning_rate": 1.0894839828635918e-06,
+      "loss": 0.6276,
+      "step": 16338
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.749159603587681,
+      "learning_rate": 1.089387762100977e-06,
+      "loss": 0.7031,
+      "step": 16339
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.663866642277847,
+      "learning_rate": 1.0892915405040986e-06,
+      "loss": 0.6426,
+      "step": 16340
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.2338177832931985,
+      "learning_rate": 1.089195318073854e-06,
+      "loss": 0.638,
+      "step": 16341
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.903731577258681,
+      "learning_rate": 1.0890990948111418e-06,
+      "loss": 0.6367,
+      "step": 16342
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.9466740576310397,
+      "learning_rate": 1.08900287071686e-06,
+      "loss": 0.6367,
+      "step": 16343
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.544644613320119,
+      "learning_rate": 1.088906645791906e-06,
+      "loss": 0.5977,
+      "step": 16344
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.395557033231281,
+      "learning_rate": 1.088810420037179e-06,
+      "loss": 0.6413,
+      "step": 16345
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.119441593292491,
+      "learning_rate": 1.0887141934535764e-06,
+      "loss": 0.6465,
+      "step": 16346
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.9204486402562164,
+      "learning_rate": 1.088617966041996e-06,
+      "loss": 0.6367,
+      "step": 16347
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.774764103208645,
+      "learning_rate": 1.0885217378033366e-06,
+      "loss": 0.6667,
+      "step": 16348
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.980087151098614,
+      "learning_rate": 1.088425508738496e-06,
+      "loss": 0.6621,
+      "step": 16349
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.333088633727813,
+      "learning_rate": 1.0883292788483725e-06,
+      "loss": 0.7122,
+      "step": 16350
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.518022949333238,
+      "learning_rate": 1.0882330481338634e-06,
+      "loss": 0.6426,
+      "step": 16351
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.770125769673651,
+      "learning_rate": 1.088136816595868e-06,
+      "loss": 0.6237,
+      "step": 16352
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.480101021152408,
+      "learning_rate": 1.0880405842352838e-06,
+      "loss": 0.6465,
+      "step": 16353
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1065589487172733,
+      "learning_rate": 1.0879443510530089e-06,
+      "loss": 0.6816,
+      "step": 16354
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.602969362668453,
+      "learning_rate": 1.0878481170499418e-06,
+      "loss": 0.6361,
+      "step": 16355
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.2533111870136073,
+      "learning_rate": 1.0877518822269802e-06,
+      "loss": 0.6654,
+      "step": 16356
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.6405603413819683,
+      "learning_rate": 1.0876556465850226e-06,
+      "loss": 0.6647,
+      "step": 16357
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.3103881059801115,
+      "learning_rate": 1.0875594101249671e-06,
+      "loss": 0.6211,
+      "step": 16358
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.561621843559702,
+      "learning_rate": 1.0874631728477123e-06,
+      "loss": 0.6921,
+      "step": 16359
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.685721928018881,
+      "learning_rate": 1.0873669347541552e-06,
+      "loss": 0.6289,
+      "step": 16360
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.842082089491518,
+      "learning_rate": 1.0872706958451954e-06,
+      "loss": 0.627,
+      "step": 16361
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.5872330911140997,
+      "learning_rate": 1.08717445612173e-06,
+      "loss": 0.6107,
+      "step": 16362
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.539938025946263,
+      "learning_rate": 1.0870782155846584e-06,
+      "loss": 0.6296,
+      "step": 16363
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.7771436643238694,
+      "learning_rate": 1.0869819742348773e-06,
+      "loss": 0.612,
+      "step": 16364
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.7787578161354567,
+      "learning_rate": 1.086885732073286e-06,
+      "loss": 0.6224,
+      "step": 16365
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1256561136378918,
+      "learning_rate": 1.086789489100783e-06,
+      "loss": 0.6387,
+      "step": 16366
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.115352555066573,
+      "learning_rate": 1.0866932453182653e-06,
+      "loss": 0.6484,
+      "step": 16367
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.7198784835814043,
+      "learning_rate": 1.0865970007266323e-06,
+      "loss": 0.6237,
+      "step": 16368
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.7743976183620154,
+      "learning_rate": 1.0865007553267817e-06,
+      "loss": 0.6829,
+      "step": 16369
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.0404260040569606,
+      "learning_rate": 1.0864045091196119e-06,
+      "loss": 0.6914,
+      "step": 16370
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1525230328599654,
+      "learning_rate": 1.0863082621060212e-06,
+      "loss": 0.6934,
+      "step": 16371
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.8817242559792424,
+      "learning_rate": 1.086212014286908e-06,
+      "loss": 0.6328,
+      "step": 16372
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.130242512744457,
+      "learning_rate": 1.08611576566317e-06,
+      "loss": 0.638,
+      "step": 16373
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.559364765328557,
+      "learning_rate": 1.0860195162357063e-06,
+      "loss": 0.7057,
+      "step": 16374
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.5722261881593487,
+      "learning_rate": 1.0859232660054148e-06,
+      "loss": 0.6439,
+      "step": 16375
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1040283480151207,
+      "learning_rate": 1.0858270149731938e-06,
+      "loss": 0.6139,
+      "step": 16376
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.0572213490063507,
+      "learning_rate": 1.0857307631399416e-06,
+      "loss": 0.6165,
+      "step": 16377
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.4178591429256073,
+      "learning_rate": 1.0856345105065567e-06,
+      "loss": 0.6361,
+      "step": 16378
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.8834220980260943,
+      "learning_rate": 1.0855382570739377e-06,
+      "loss": 0.5918,
+      "step": 16379
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.046350713774598,
+      "learning_rate": 1.0854420028429821e-06,
+      "loss": 0.6224,
+      "step": 16380
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.9149568792774114,
+      "learning_rate": 1.0853457478145887e-06,
+      "loss": 0.6419,
+      "step": 16381
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.752907277681982,
+      "learning_rate": 1.0852494919896564e-06,
+      "loss": 0.6374,
+      "step": 16382
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.8479612804665533,
+      "learning_rate": 1.0851532353690826e-06,
+      "loss": 0.6641,
+      "step": 16383
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.4928952179764083,
+      "learning_rate": 1.0850569779537664e-06,
+      "loss": 0.6257,
+      "step": 16384
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 4.338394248006104,
+      "learning_rate": 1.0849607197446058e-06,
+      "loss": 0.6589,
+      "step": 16385
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.621596925142939,
+      "learning_rate": 1.0848644607424991e-06,
+      "loss": 0.6445,
+      "step": 16386
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.03017457969983,
+      "learning_rate": 1.0847682009483449e-06,
+      "loss": 0.6406,
+      "step": 16387
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.3029231164004322,
+      "learning_rate": 1.0846719403630415e-06,
+      "loss": 0.6576,
+      "step": 16388
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.365634682493997,
+      "learning_rate": 1.0845756789874878e-06,
+      "loss": 0.6172,
+      "step": 16389
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.814645647238855,
+      "learning_rate": 1.0844794168225817e-06,
+      "loss": 0.6328,
+      "step": 16390
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.9653284781155587,
+      "learning_rate": 1.0843831538692213e-06,
+      "loss": 0.6172,
+      "step": 16391
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.5907357695683593,
+      "learning_rate": 1.084286890128306e-06,
+      "loss": 0.6725,
+      "step": 16392
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 5.427323467763186,
+      "learning_rate": 1.0841906256007337e-06,
+      "loss": 0.6771,
+      "step": 16393
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.9346180670110806,
+      "learning_rate": 1.0840943602874025e-06,
+      "loss": 0.6077,
+      "step": 16394
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.319481627151488,
+      "learning_rate": 1.0839980941892115e-06,
+      "loss": 0.6549,
+      "step": 16395
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.959320507756627,
+      "learning_rate": 1.0839018273070586e-06,
+      "loss": 0.6497,
+      "step": 16396
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.0571140443531744,
+      "learning_rate": 1.0838055596418424e-06,
+      "loss": 0.6224,
+      "step": 16397
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.960258836414737,
+      "learning_rate": 1.0837092911944619e-06,
+      "loss": 0.627,
+      "step": 16398
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.315268005106289,
+      "learning_rate": 1.0836130219658151e-06,
+      "loss": 0.666,
+      "step": 16399
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.134249698197382,
+      "learning_rate": 1.0835167519568006e-06,
+      "loss": 0.6396,
+      "step": 16400
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.4595581906547377,
+      "learning_rate": 1.0834204811683166e-06,
+      "loss": 0.623,
+      "step": 16401
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.0105214162570433,
+      "learning_rate": 1.083324209601262e-06,
+      "loss": 0.6458,
+      "step": 16402
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.20222350878902,
+      "learning_rate": 1.0832279372565352e-06,
+      "loss": 0.6504,
+      "step": 16403
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.229916748267959,
+      "learning_rate": 1.0831316641350346e-06,
+      "loss": 0.6074,
+      "step": 16404
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.873645898034827,
+      "learning_rate": 1.083035390237659e-06,
+      "loss": 0.6224,
+      "step": 16405
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.6511204151152277,
+      "learning_rate": 1.0829391155653068e-06,
+      "loss": 0.6621,
+      "step": 16406
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.937619763451913,
+      "learning_rate": 1.0828428401188762e-06,
+      "loss": 0.5866,
+      "step": 16407
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.4610041066455555,
+      "learning_rate": 1.0827465638992663e-06,
+      "loss": 0.6048,
+      "step": 16408
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.0208663418518,
+      "learning_rate": 1.0826502869073755e-06,
+      "loss": 0.668,
+      "step": 16409
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.907365221811988,
+      "learning_rate": 1.0825540091441019e-06,
+      "loss": 0.6803,
+      "step": 16410
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.06953764770174,
+      "learning_rate": 1.0824577306103446e-06,
+      "loss": 0.6426,
+      "step": 16411
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.155152987671275,
+      "learning_rate": 1.0823614513070023e-06,
+      "loss": 0.6758,
+      "step": 16412
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.159946956765542,
+      "learning_rate": 1.0822651712349728e-06,
+      "loss": 0.7012,
+      "step": 16413
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.251484762984483,
+      "learning_rate": 1.0821688903951555e-06,
+      "loss": 0.6126,
+      "step": 16414
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.5660586316634597,
+      "learning_rate": 1.0820726087884487e-06,
+      "loss": 0.6478,
+      "step": 16415
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.222180305337158,
+      "learning_rate": 1.081976326415751e-06,
+      "loss": 0.6615,
+      "step": 16416
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.9814590301329242,
+      "learning_rate": 1.0818800432779607e-06,
+      "loss": 0.6341,
+      "step": 16417
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.9178391876100966,
+      "learning_rate": 1.081783759375977e-06,
+      "loss": 0.6374,
+      "step": 16418
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.720578262246573,
+      "learning_rate": 1.0816874747106982e-06,
+      "loss": 0.6315,
+      "step": 16419
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.641967355337304,
+      "learning_rate": 1.0815911892830228e-06,
+      "loss": 0.6081,
+      "step": 16420
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.2349580387338936,
+      "learning_rate": 1.0814949030938495e-06,
+      "loss": 0.6146,
+      "step": 16421
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.0443284044246135,
+      "learning_rate": 1.0813986161440775e-06,
+      "loss": 0.6452,
+      "step": 16422
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.172267113954929,
+      "learning_rate": 1.0813023284346045e-06,
+      "loss": 0.6387,
+      "step": 16423
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.686493161938374,
+      "learning_rate": 1.08120603996633e-06,
+      "loss": 0.651,
+      "step": 16424
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.977900098418853,
+      "learning_rate": 1.0811097507401525e-06,
+      "loss": 0.6283,
+      "step": 16425
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.95673548993502,
+      "learning_rate": 1.0810134607569703e-06,
+      "loss": 0.6654,
+      "step": 16426
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 6.828190608050031,
+      "learning_rate": 1.0809171700176824e-06,
+      "loss": 0.6387,
+      "step": 16427
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.966618295499133,
+      "learning_rate": 1.0808208785231873e-06,
+      "loss": 0.6784,
+      "step": 16428
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.031642841957264,
+      "learning_rate": 1.0807245862743842e-06,
+      "loss": 0.6784,
+      "step": 16429
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.749515839631927,
+      "learning_rate": 1.0806282932721709e-06,
+      "loss": 0.6328,
+      "step": 16430
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.619559911776305,
+      "learning_rate": 1.0805319995174465e-06,
+      "loss": 0.6263,
+      "step": 16431
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.162798651697016,
+      "learning_rate": 1.0804357050111104e-06,
+      "loss": 0.6732,
+      "step": 16432
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.097444800608062,
+      "learning_rate": 1.08033940975406e-06,
+      "loss": 0.6315,
+      "step": 16433
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 7.1103972298137945,
+      "learning_rate": 1.0802431137471954e-06,
+      "loss": 0.6641,
+      "step": 16434
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.780228438666357,
+      "learning_rate": 1.0801468169914148e-06,
+      "loss": 0.6374,
+      "step": 16435
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.894009255158155,
+      "learning_rate": 1.0800505194876163e-06,
+      "loss": 0.6836,
+      "step": 16436
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.9905009700183083,
+      "learning_rate": 1.0799542212366996e-06,
+      "loss": 0.6517,
+      "step": 16437
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.7161365942989573,
+      "learning_rate": 1.0798579222395632e-06,
+      "loss": 0.6035,
+      "step": 16438
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.1105367501975523,
+      "learning_rate": 1.0797616224971054e-06,
+      "loss": 0.6328,
+      "step": 16439
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.592379437716966,
+      "learning_rate": 1.0796653220102254e-06,
+      "loss": 0.5938,
+      "step": 16440
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.890423236315873,
+      "learning_rate": 1.079569020779822e-06,
+      "loss": 0.6133,
+      "step": 16441
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.194180650445173,
+      "learning_rate": 1.0794727188067937e-06,
+      "loss": 0.5872,
+      "step": 16442
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.2634018722350433,
+      "learning_rate": 1.07937641609204e-06,
+      "loss": 0.6836,
+      "step": 16443
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.014230312463819,
+      "learning_rate": 1.0792801126364585e-06,
+      "loss": 0.6328,
+      "step": 16444
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 5.043233499379622,
+      "learning_rate": 1.0791838084409494e-06,
+      "loss": 0.6562,
+      "step": 16445
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.128353066226807,
+      "learning_rate": 1.0790875035064104e-06,
+      "loss": 0.6419,
+      "step": 16446
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.2563412672890535,
+      "learning_rate": 1.0789911978337408e-06,
+      "loss": 0.6237,
+      "step": 16447
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.9403530949836574,
+      "learning_rate": 1.0788948914238396e-06,
+      "loss": 0.653,
+      "step": 16448
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.06980782609871,
+      "learning_rate": 1.0787985842776048e-06,
+      "loss": 0.6432,
+      "step": 16449
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.2999082256659023,
+      "learning_rate": 1.0787022763959364e-06,
+      "loss": 0.6406,
+      "step": 16450
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.1468289714846116,
+      "learning_rate": 1.0786059677797322e-06,
+      "loss": 0.6198,
+      "step": 16451
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.187458230573429,
+      "learning_rate": 1.0785096584298921e-06,
+      "loss": 0.6992,
+      "step": 16452
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.200318923469718,
+      "learning_rate": 1.0784133483473143e-06,
+      "loss": 0.6198,
+      "step": 16453
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.0701673905769606,
+      "learning_rate": 1.0783170375328974e-06,
+      "loss": 0.6224,
+      "step": 16454
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.499155470040558,
+      "learning_rate": 1.0782207259875411e-06,
+      "loss": 0.6523,
+      "step": 16455
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 4.053831887423509,
+      "learning_rate": 1.0781244137121436e-06,
+      "loss": 0.6328,
+      "step": 16456
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.4786737875276526,
+      "learning_rate": 1.0780281007076039e-06,
+      "loss": 0.6523,
+      "step": 16457
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.1963522721825792,
+      "learning_rate": 1.0779317869748214e-06,
+      "loss": 0.6211,
+      "step": 16458
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 3.1086793504698855,
+      "learning_rate": 1.0778354725146945e-06,
+      "loss": 0.6341,
+      "step": 16459
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.9097850725510788,
+      "learning_rate": 1.0777391573281223e-06,
+      "loss": 0.623,
+      "step": 16460
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.9478056339022887,
+      "learning_rate": 1.0776428414160037e-06,
+      "loss": 0.6452,
+      "step": 16461
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.6616999414418974,
+      "learning_rate": 1.0775465247792375e-06,
+      "loss": 0.6354,
+      "step": 16462
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 8.993978098101477,
+      "learning_rate": 1.0774502074187227e-06,
+      "loss": 0.6361,
+      "step": 16463
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.621021062809438,
+      "learning_rate": 1.077353889335358e-06,
+      "loss": 0.6081,
+      "step": 16464
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.0768704349795377,
+      "learning_rate": 1.077257570530043e-06,
+      "loss": 0.6426,
+      "step": 16465
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.0359069686867146,
+      "learning_rate": 1.0771612510036763e-06,
+      "loss": 0.6621,
+      "step": 16466
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.875821343914683,
+      "learning_rate": 1.0770649307571567e-06,
+      "loss": 0.638,
+      "step": 16467
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.1748041138358785,
+      "learning_rate": 1.0769686097913833e-06,
+      "loss": 0.6432,
+      "step": 16468
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.552246076707506,
+      "learning_rate": 1.0768722881072552e-06,
+      "loss": 0.5928,
+      "step": 16469
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.492653999362387,
+      "learning_rate": 1.0767759657056707e-06,
+      "loss": 0.5902,
+      "step": 16470
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.483038799461084,
+      "learning_rate": 1.0766796425875298e-06,
+      "loss": 0.6283,
+      "step": 16471
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.108142497093002,
+      "learning_rate": 1.076583318753731e-06,
+      "loss": 0.6595,
+      "step": 16472
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.9782110149244976,
+      "learning_rate": 1.076486994205173e-06,
+      "loss": 0.5739,
+      "step": 16473
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.659827633945598,
+      "learning_rate": 1.0763906689427555e-06,
+      "loss": 0.6302,
+      "step": 16474
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.046791778665622,
+      "learning_rate": 1.076294342967377e-06,
+      "loss": 0.6341,
+      "step": 16475
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.4488702333030696,
+      "learning_rate": 1.0761980162799365e-06,
+      "loss": 0.6178,
+      "step": 16476
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.2664096222732213,
+      "learning_rate": 1.0761016888813333e-06,
+      "loss": 0.6217,
+      "step": 16477
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.107725060422061,
+      "learning_rate": 1.0760053607724664e-06,
+      "loss": 0.6921,
+      "step": 16478
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.110303727287481,
+      "learning_rate": 1.0759090319542345e-06,
+      "loss": 0.6087,
+      "step": 16479
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.987254609017298,
+      "learning_rate": 1.075812702427537e-06,
+      "loss": 0.6113,
+      "step": 16480
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.507102296229874,
+      "learning_rate": 1.075716372193273e-06,
+      "loss": 0.6628,
+      "step": 16481
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 7.817969017732919,
+      "learning_rate": 1.0756200412523414e-06,
+      "loss": 0.6673,
+      "step": 16482
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.179717510551749,
+      "learning_rate": 1.0755237096056409e-06,
+      "loss": 0.6758,
+      "step": 16483
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.449247266036635,
+      "learning_rate": 1.0754273772540713e-06,
+      "loss": 0.6439,
+      "step": 16484
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.8525375003601683,
+      "learning_rate": 1.0753310441985312e-06,
+      "loss": 0.6536,
+      "step": 16485
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.153711485154768,
+      "learning_rate": 1.0752347104399197e-06,
+      "loss": 0.6621,
+      "step": 16486
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.41112701467824,
+      "learning_rate": 1.075138375979136e-06,
+      "loss": 0.6536,
+      "step": 16487
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.8229161040466164,
+      "learning_rate": 1.0750420408170797e-06,
+      "loss": 0.6458,
+      "step": 16488
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.318089385283903,
+      "learning_rate": 1.0749457049546485e-06,
+      "loss": 0.6569,
+      "step": 16489
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.442860511023687,
+      "learning_rate": 1.0748493683927433e-06,
+      "loss": 0.6322,
+      "step": 16490
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.579696115661697,
+      "learning_rate": 1.074753031132262e-06,
+      "loss": 0.6732,
+      "step": 16491
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.263490055218794,
+      "learning_rate": 1.0746566931741036e-06,
+      "loss": 0.6107,
+      "step": 16492
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.85540961358081,
+      "learning_rate": 1.0745603545191681e-06,
+      "loss": 0.5918,
+      "step": 16493
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.3006835291878867,
+      "learning_rate": 1.0744640151683544e-06,
+      "loss": 0.6227,
+      "step": 16494
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.120207122886225,
+      "learning_rate": 1.0743676751225612e-06,
+      "loss": 0.6634,
+      "step": 16495
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.878513883053347,
+      "learning_rate": 1.0742713343826878e-06,
+      "loss": 0.6283,
+      "step": 16496
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.963285078443818,
+      "learning_rate": 1.0741749929496336e-06,
+      "loss": 0.5983,
+      "step": 16497
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.129372827025567,
+      "learning_rate": 1.0740786508242978e-06,
+      "loss": 0.653,
+      "step": 16498
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.7027195629801626,
+      "learning_rate": 1.0739823080075792e-06,
+      "loss": 0.6055,
+      "step": 16499
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.46439412690948,
+      "learning_rate": 1.0738859645003772e-06,
+      "loss": 0.6139,
+      "step": 16500
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.974873617522218,
+      "learning_rate": 1.0737896203035913e-06,
+      "loss": 0.6191,
+      "step": 16501
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.575636459493583,
+      "learning_rate": 1.0736932754181199e-06,
+      "loss": 0.6315,
+      "step": 16502
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.2399878836413625,
+      "learning_rate": 1.0735969298448626e-06,
+      "loss": 0.6862,
+      "step": 16503
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.8557962569757085,
+      "learning_rate": 1.0735005835847188e-06,
+      "loss": 0.6406,
+      "step": 16504
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.195782026835101,
+      "learning_rate": 1.0734042366385875e-06,
+      "loss": 0.6673,
+      "step": 16505
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.025545544407441,
+      "learning_rate": 1.0733078890073682e-06,
+      "loss": 0.6719,
+      "step": 16506
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.399279240275406,
+      "learning_rate": 1.0732115406919595e-06,
+      "loss": 0.6693,
+      "step": 16507
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.2327925098209094,
+      "learning_rate": 1.0731151916932612e-06,
+      "loss": 0.6243,
+      "step": 16508
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.663094803660011,
+      "learning_rate": 1.0730188420121726e-06,
+      "loss": 0.6315,
+      "step": 16509
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.009745464213995,
+      "learning_rate": 1.0729224916495922e-06,
+      "loss": 0.6484,
+      "step": 16510
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.084937654862896,
+      "learning_rate": 1.0728261406064206e-06,
+      "loss": 0.6172,
+      "step": 16511
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.878843542158494,
+      "learning_rate": 1.0727297888835553e-06,
+      "loss": 0.5866,
+      "step": 16512
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.788208054841044,
+      "learning_rate": 1.0726334364818966e-06,
+      "loss": 0.6484,
+      "step": 16513
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.3024067667441583,
+      "learning_rate": 1.072537083402344e-06,
+      "loss": 0.6335,
+      "step": 16514
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.788377111497587,
+      "learning_rate": 1.072440729645796e-06,
+      "loss": 0.6165,
+      "step": 16515
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.671500036378551,
+      "learning_rate": 1.0723443752131527e-06,
+      "loss": 0.6224,
+      "step": 16516
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.994610962257273,
+      "learning_rate": 1.0722480201053122e-06,
+      "loss": 0.6888,
+      "step": 16517
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 5.5796532328566,
+      "learning_rate": 1.0721516643231753e-06,
+      "loss": 0.625,
+      "step": 16518
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.2067887398195105,
+      "learning_rate": 1.0720553078676403e-06,
+      "loss": 0.7324,
+      "step": 16519
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.4964624094303445,
+      "learning_rate": 1.0719589507396065e-06,
+      "loss": 0.6491,
+      "step": 16520
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.7090735253221343,
+      "learning_rate": 1.0718625929399737e-06,
+      "loss": 0.6068,
+      "step": 16521
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.071473784270163,
+      "learning_rate": 1.071766234469641e-06,
+      "loss": 0.6263,
+      "step": 16522
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.3427718314103365,
+      "learning_rate": 1.0716698753295074e-06,
+      "loss": 0.6517,
+      "step": 16523
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 4.362252454323943,
+      "learning_rate": 1.0715735155204727e-06,
+      "loss": 0.6387,
+      "step": 16524
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.137991938956795,
+      "learning_rate": 1.0714771550434363e-06,
+      "loss": 0.6393,
+      "step": 16525
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 3.6757367837969017,
+      "learning_rate": 1.071380793899297e-06,
+      "loss": 0.6296,
+      "step": 16526
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.78072307830015,
+      "learning_rate": 1.0712844320889542e-06,
+      "loss": 0.6165,
+      "step": 16527
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.7729738191111415,
+      "learning_rate": 1.071188069613308e-06,
+      "loss": 0.6432,
+      "step": 16528
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.2935546550910786,
+      "learning_rate": 1.071091706473257e-06,
+      "loss": 0.6986,
+      "step": 16529
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.8475311136648576,
+      "learning_rate": 1.0709953426697007e-06,
+      "loss": 0.6569,
+      "step": 16530
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.7114351949218305,
+      "learning_rate": 1.070898978203539e-06,
+      "loss": 0.6328,
+      "step": 16531
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.052379883945496,
+      "learning_rate": 1.0708026130756706e-06,
+      "loss": 0.6595,
+      "step": 16532
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.9601434658404626,
+      "learning_rate": 1.070706247286995e-06,
+      "loss": 0.6406,
+      "step": 16533
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.7482988143919522,
+      "learning_rate": 1.0706098808384119e-06,
+      "loss": 0.6406,
+      "step": 16534
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.8694616502788968,
+      "learning_rate": 1.0705135137308206e-06,
+      "loss": 0.651,
+      "step": 16535
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.995931187779108,
+      "learning_rate": 1.0704171459651201e-06,
+      "loss": 0.6335,
+      "step": 16536
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.1159305290989208,
+      "learning_rate": 1.0703207775422106e-06,
+      "loss": 0.6589,
+      "step": 16537
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.887441773441922,
+      "learning_rate": 1.070224408462991e-06,
+      "loss": 0.6361,
+      "step": 16538
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.1463784636278205,
+      "learning_rate": 1.0701280387283604e-06,
+      "loss": 0.6536,
+      "step": 16539
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.2206934691324633,
+      "learning_rate": 1.070031668339219e-06,
+      "loss": 0.651,
+      "step": 16540
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.0075542449671806,
+      "learning_rate": 1.0699352972964657e-06,
+      "loss": 0.6263,
+      "step": 16541
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.770257001413496,
+      "learning_rate": 1.0698389256010001e-06,
+      "loss": 0.6335,
+      "step": 16542
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.8315517737791036,
+      "learning_rate": 1.0697425532537217e-06,
+      "loss": 0.6354,
+      "step": 16543
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.6238783145796982,
+      "learning_rate": 1.0696461802555298e-06,
+      "loss": 0.6647,
+      "step": 16544
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.48225285088224,
+      "learning_rate": 1.0695498066073238e-06,
+      "loss": 0.597,
+      "step": 16545
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.892866371631566,
+      "learning_rate": 1.0694534323100034e-06,
+      "loss": 0.6374,
+      "step": 16546
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.6505883381803237,
+      "learning_rate": 1.0693570573644684e-06,
+      "loss": 0.6198,
+      "step": 16547
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.1819377895701146,
+      "learning_rate": 1.0692606817716174e-06,
+      "loss": 0.6628,
+      "step": 16548
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.997316659430023,
+      "learning_rate": 1.0691643055323503e-06,
+      "loss": 0.696,
+      "step": 16549
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.772234399619201,
+      "learning_rate": 1.0690679286475668e-06,
+      "loss": 0.6374,
+      "step": 16550
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.859487424470108,
+      "learning_rate": 1.0689715511181663e-06,
+      "loss": 0.6387,
+      "step": 16551
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 7.416197683490612,
+      "learning_rate": 1.0688751729450478e-06,
+      "loss": 0.6569,
+      "step": 16552
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.832276799877045,
+      "learning_rate": 1.0687787941291115e-06,
+      "loss": 0.6628,
+      "step": 16553
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.468993312765913,
+      "learning_rate": 1.0686824146712567e-06,
+      "loss": 0.61,
+      "step": 16554
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.852264244765483,
+      "learning_rate": 1.0685860345723827e-06,
+      "loss": 0.7038,
+      "step": 16555
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.399958776631386,
+      "learning_rate": 1.0684896538333894e-06,
+      "loss": 0.6159,
+      "step": 16556
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.4985123134512603,
+      "learning_rate": 1.0683932724551758e-06,
+      "loss": 0.653,
+      "step": 16557
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.9537280156187347,
+      "learning_rate": 1.0682968904386419e-06,
+      "loss": 0.6914,
+      "step": 16558
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.2784181018342657,
+      "learning_rate": 1.0682005077846871e-06,
+      "loss": 0.6914,
+      "step": 16559
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.5346543929172975,
+      "learning_rate": 1.0681041244942108e-06,
+      "loss": 0.6152,
+      "step": 16560
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.7404717708602533,
+      "learning_rate": 1.0680077405681132e-06,
+      "loss": 0.6608,
+      "step": 16561
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.040054061597467,
+      "learning_rate": 1.0679113560072926e-06,
+      "loss": 0.6257,
+      "step": 16562
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.0678031587986805,
+      "learning_rate": 1.0678149708126496e-06,
+      "loss": 0.6439,
+      "step": 16563
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.732235023239454,
+      "learning_rate": 1.0677185849850837e-06,
+      "loss": 0.6243,
+      "step": 16564
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.267265201973663,
+      "learning_rate": 1.0676221985254937e-06,
+      "loss": 0.6576,
+      "step": 16565
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.807034434828246,
+      "learning_rate": 1.0675258114347804e-06,
+      "loss": 0.6348,
+      "step": 16566
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.096826316420475,
+      "learning_rate": 1.0674294237138426e-06,
+      "loss": 0.6198,
+      "step": 16567
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.883176360802321,
+      "learning_rate": 1.0673330353635796e-06,
+      "loss": 0.6348,
+      "step": 16568
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.7468394506646105,
+      "learning_rate": 1.0672366463848919e-06,
+      "loss": 0.6003,
+      "step": 16569
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.2645727429381712,
+      "learning_rate": 1.0671402567786785e-06,
+      "loss": 0.627,
+      "step": 16570
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.90130216538511,
+      "learning_rate": 1.0670438665458388e-06,
+      "loss": 0.6283,
+      "step": 16571
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.902373511354042,
+      "learning_rate": 1.066947475687273e-06,
+      "loss": 0.6315,
+      "step": 16572
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.558463785014613,
+      "learning_rate": 1.0668510842038805e-06,
+      "loss": 0.681,
+      "step": 16573
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.9749889576090616,
+      "learning_rate": 1.0667546920965612e-06,
+      "loss": 0.6354,
+      "step": 16574
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.1129595629367985,
+      "learning_rate": 1.0666582993662142e-06,
+      "loss": 0.6497,
+      "step": 16575
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.687233799991034,
+      "learning_rate": 1.0665619060137393e-06,
+      "loss": 0.6243,
+      "step": 16576
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.846989246163363,
+      "learning_rate": 1.0664655120400366e-06,
+      "loss": 0.6445,
+      "step": 16577
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 5.256628166582968,
+      "learning_rate": 1.0663691174460047e-06,
+      "loss": 0.6185,
+      "step": 16578
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.9633544575835153,
+      "learning_rate": 1.0662727222325444e-06,
+      "loss": 0.6257,
+      "step": 16579
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.968845100247705,
+      "learning_rate": 1.0661763264005553e-06,
+      "loss": 0.6263,
+      "step": 16580
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.4270197132540914,
+      "learning_rate": 1.066079929950936e-06,
+      "loss": 0.6328,
+      "step": 16581
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.4709335076433003,
+      "learning_rate": 1.0659835328845874e-06,
+      "loss": 0.6543,
+      "step": 16582
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.8992954683603243,
+      "learning_rate": 1.0658871352024081e-06,
+      "loss": 0.6263,
+      "step": 16583
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.522212942693985,
+      "learning_rate": 1.065790736905299e-06,
+      "loss": 0.7135,
+      "step": 16584
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.650033454074483,
+      "learning_rate": 1.0656943379941587e-06,
+      "loss": 0.6335,
+      "step": 16585
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.880607070577137,
+      "learning_rate": 1.0655979384698874e-06,
+      "loss": 0.6686,
+      "step": 16586
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.786851010560459,
+      "learning_rate": 1.0655015383333848e-06,
+      "loss": 0.6136,
+      "step": 16587
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.992745691388848,
+      "learning_rate": 1.0654051375855506e-06,
+      "loss": 0.6595,
+      "step": 16588
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.8751694170547575,
+      "learning_rate": 1.0653087362272843e-06,
+      "loss": 0.6029,
+      "step": 16589
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.237057252824357,
+      "learning_rate": 1.0652123342594858e-06,
+      "loss": 0.6602,
+      "step": 16590
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 4.1457898226713015,
+      "learning_rate": 1.065115931683055e-06,
+      "loss": 0.6217,
+      "step": 16591
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.00220083090861,
+      "learning_rate": 1.0650195284988913e-06,
+      "loss": 0.6576,
+      "step": 16592
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.9667644512757634,
+      "learning_rate": 1.0649231247078946e-06,
+      "loss": 0.681,
+      "step": 16593
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 6.513711158711923,
+      "learning_rate": 1.064826720310965e-06,
+      "loss": 0.6504,
+      "step": 16594
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.826766095319304,
+      "learning_rate": 1.0647303153090013e-06,
+      "loss": 0.6061,
+      "step": 16595
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.267594991795164,
+      "learning_rate": 1.0646339097029042e-06,
+      "loss": 0.653,
+      "step": 16596
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.596460502900564,
+      "learning_rate": 1.064537503493573e-06,
+      "loss": 0.627,
+      "step": 16597
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.6735288070569174,
+      "learning_rate": 1.0644410966819078e-06,
+      "loss": 0.6458,
+      "step": 16598
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.906309131968693,
+      "learning_rate": 1.0643446892688077e-06,
+      "loss": 0.6621,
+      "step": 16599
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.191984711962047,
+      "learning_rate": 1.0642482812551732e-06,
+      "loss": 0.6315,
+      "step": 16600
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.713102040467932,
+      "learning_rate": 1.0641518726419038e-06,
+      "loss": 0.6732,
+      "step": 16601
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.275603015691537,
+      "learning_rate": 1.0640554634298993e-06,
+      "loss": 0.6439,
+      "step": 16602
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.880085287487562,
+      "learning_rate": 1.0639590536200596e-06,
+      "loss": 0.6432,
+      "step": 16603
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.7612486557584357,
+      "learning_rate": 1.0638626432132842e-06,
+      "loss": 0.6335,
+      "step": 16604
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.53325963251375,
+      "learning_rate": 1.063766232210473e-06,
+      "loss": 0.6126,
+      "step": 16605
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.49036368003377,
+      "learning_rate": 1.063669820612526e-06,
+      "loss": 0.6393,
+      "step": 16606
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.75863607803986,
+      "learning_rate": 1.0635734084203432e-06,
+      "loss": 0.61,
+      "step": 16607
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.8511102008861546,
+      "learning_rate": 1.0634769956348238e-06,
+      "loss": 0.6156,
+      "step": 16608
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.131394513886257,
+      "learning_rate": 1.0633805822568682e-06,
+      "loss": 0.6035,
+      "step": 16609
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.928504129976515,
+      "learning_rate": 1.063284168287376e-06,
+      "loss": 0.64,
+      "step": 16610
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.2348091076638354,
+      "learning_rate": 1.0631877537272474e-06,
+      "loss": 0.6341,
+      "step": 16611
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.8835181492537374,
+      "learning_rate": 1.0630913385773812e-06,
+      "loss": 0.6152,
+      "step": 16612
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.25550688093306,
+      "learning_rate": 1.0629949228386783e-06,
+      "loss": 0.6406,
+      "step": 16613
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.1847354817506597,
+      "learning_rate": 1.0628985065120383e-06,
+      "loss": 0.64,
+      "step": 16614
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.634429406268665,
+      "learning_rate": 1.0628020895983607e-06,
+      "loss": 0.6257,
+      "step": 16615
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.582719700770154,
+      "learning_rate": 1.0627056720985462e-06,
+      "loss": 0.6693,
+      "step": 16616
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.1925782586386706,
+      "learning_rate": 1.0626092540134936e-06,
+      "loss": 0.6191,
+      "step": 16617
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.782339485103639,
+      "learning_rate": 1.0625128353441034e-06,
+      "loss": 0.6745,
+      "step": 16618
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.448361470873428,
+      "learning_rate": 1.0624164160912755e-06,
+      "loss": 0.6686,
+      "step": 16619
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.4962292525528964,
+      "learning_rate": 1.0623199962559095e-06,
+      "loss": 0.6198,
+      "step": 16620
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.163719522122896,
+      "learning_rate": 1.0622235758389055e-06,
+      "loss": 0.6328,
+      "step": 16621
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.3788994374167447,
+      "learning_rate": 1.0621271548411634e-06,
+      "loss": 0.6257,
+      "step": 16622
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.7187350554082523,
+      "learning_rate": 1.0620307332635833e-06,
+      "loss": 0.625,
+      "step": 16623
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.6322324229673524,
+      "learning_rate": 1.0619343111070645e-06,
+      "loss": 0.6149,
+      "step": 16624
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.967511963074988,
+      "learning_rate": 1.0618378883725077e-06,
+      "loss": 0.6803,
+      "step": 16625
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.9576871597061416,
+      "learning_rate": 1.0617414650608118e-06,
+      "loss": 0.6406,
+      "step": 16626
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.9719710495035265,
+      "learning_rate": 1.061645041172878e-06,
+      "loss": 0.6029,
+      "step": 16627
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.101069648799259,
+      "learning_rate": 1.0615486167096053e-06,
+      "loss": 0.6387,
+      "step": 16628
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.3765219487202445,
+      "learning_rate": 1.0614521916718937e-06,
+      "loss": 0.6569,
+      "step": 16629
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.502456121300937,
+      "learning_rate": 1.0613557660606441e-06,
+      "loss": 0.709,
+      "step": 16630
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.8884898828503958,
+      "learning_rate": 1.0612593398767548e-06,
+      "loss": 0.6777,
+      "step": 16631
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.9732434346235674,
+      "learning_rate": 1.0611629131211273e-06,
+      "loss": 0.6094,
+      "step": 16632
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.877016570371078,
+      "learning_rate": 1.0610664857946607e-06,
+      "loss": 0.6426,
+      "step": 16633
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 6.764970424275453,
+      "learning_rate": 1.060970057898255e-06,
+      "loss": 0.666,
+      "step": 16634
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.830504343679963,
+      "learning_rate": 1.0608736294328106e-06,
+      "loss": 0.6217,
+      "step": 16635
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.4014547161091624,
+      "learning_rate": 1.0607772003992271e-06,
+      "loss": 0.6816,
+      "step": 16636
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.4024078809750793,
+      "learning_rate": 1.0606807707984048e-06,
+      "loss": 0.6289,
+      "step": 16637
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.511539968344674,
+      "learning_rate": 1.0605843406312433e-06,
+      "loss": 0.6673,
+      "step": 16638
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.938726235943024,
+      "learning_rate": 1.0604879098986428e-06,
+      "loss": 0.6374,
+      "step": 16639
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.1863374529974138,
+      "learning_rate": 1.0603914786015033e-06,
+      "loss": 0.5905,
+      "step": 16640
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.3884804385245433,
+      "learning_rate": 1.060295046740725e-06,
+      "loss": 0.6673,
+      "step": 16641
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.352522654112734,
+      "learning_rate": 1.0601986143172076e-06,
+      "loss": 0.6146,
+      "step": 16642
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.582584574422062,
+      "learning_rate": 1.0601021813318514e-06,
+      "loss": 0.6152,
+      "step": 16643
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.03727312199283,
+      "learning_rate": 1.0600057477855557e-06,
+      "loss": 0.6094,
+      "step": 16644
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.7698265030398677,
+      "learning_rate": 1.0599093136792211e-06,
+      "loss": 0.6426,
+      "step": 16645
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.815466822053905,
+      "learning_rate": 1.0598128790137482e-06,
+      "loss": 0.6693,
+      "step": 16646
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.074200274951365,
+      "learning_rate": 1.059716443790036e-06,
+      "loss": 0.6042,
+      "step": 16647
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.200147246107283,
+      "learning_rate": 1.059620008008985e-06,
+      "loss": 0.6341,
+      "step": 16648
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.2370592803999894,
+      "learning_rate": 1.059523571671495e-06,
+      "loss": 0.6777,
+      "step": 16649
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.9271323811889163,
+      "learning_rate": 1.0594271347784664e-06,
+      "loss": 0.6117,
+      "step": 16650
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.5581785425169463,
+      "learning_rate": 1.0593306973307992e-06,
+      "loss": 0.6615,
+      "step": 16651
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.281993730173031,
+      "learning_rate": 1.059234259329393e-06,
+      "loss": 0.6758,
+      "step": 16652
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.560553884726162,
+      "learning_rate": 1.0591378207751488e-06,
+      "loss": 0.6465,
+      "step": 16653
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.036719903257724,
+      "learning_rate": 1.0590413816689657e-06,
+      "loss": 0.6634,
+      "step": 16654
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.2330018953085102,
+      "learning_rate": 1.058944942011744e-06,
+      "loss": 0.627,
+      "step": 16655
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.1181705346646535,
+      "learning_rate": 1.0588485018043842e-06,
+      "loss": 0.6595,
+      "step": 16656
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 5.970278844457859,
+      "learning_rate": 1.0587520610477862e-06,
+      "loss": 0.6517,
+      "step": 16657
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.207339918799984,
+      "learning_rate": 1.0586556197428498e-06,
+      "loss": 0.6686,
+      "step": 16658
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.599215376080897,
+      "learning_rate": 1.0585591778904755e-06,
+      "loss": 0.6139,
+      "step": 16659
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 4.873633616316202,
+      "learning_rate": 1.0584627354915632e-06,
+      "loss": 0.6471,
+      "step": 16660
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.847387235039533,
+      "learning_rate": 1.0583662925470126e-06,
+      "loss": 0.6504,
+      "step": 16661
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.489494735697636,
+      "learning_rate": 1.058269849057725e-06,
+      "loss": 0.6341,
+      "step": 16662
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.7207134074131964,
+      "learning_rate": 1.0581734050245992e-06,
+      "loss": 0.679,
+      "step": 16663
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.044647723352439,
+      "learning_rate": 1.0580769604485363e-06,
+      "loss": 0.5892,
+      "step": 16664
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.5388649268154255,
+      "learning_rate": 1.0579805153304354e-06,
+      "loss": 0.6758,
+      "step": 16665
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.313212804699694,
+      "learning_rate": 1.0578840696711976e-06,
+      "loss": 0.6328,
+      "step": 16666
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.101853052246686,
+      "learning_rate": 1.0577876234717228e-06,
+      "loss": 0.6374,
+      "step": 16667
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.7580819807143233,
+      "learning_rate": 1.0576911767329103e-06,
+      "loss": 0.6829,
+      "step": 16668
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.926695154088618,
+      "learning_rate": 1.0575947294556617e-06,
+      "loss": 0.6549,
+      "step": 16669
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.7749673586828645,
+      "learning_rate": 1.0574982816408762e-06,
+      "loss": 0.6159,
+      "step": 16670
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.124525747595442,
+      "learning_rate": 1.0574018332894538e-06,
+      "loss": 0.6562,
+      "step": 16671
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.2244324434274163,
+      "learning_rate": 1.0573053844022954e-06,
+      "loss": 0.6257,
+      "step": 16672
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.336998121832415,
+      "learning_rate": 1.057208934980301e-06,
+      "loss": 0.6217,
+      "step": 16673
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.019082338294373,
+      "learning_rate": 1.0571124850243699e-06,
+      "loss": 0.6634,
+      "step": 16674
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.4491959148669453,
+      "learning_rate": 1.0570160345354033e-06,
+      "loss": 0.6693,
+      "step": 16675
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.6043250709275116,
+      "learning_rate": 1.056919583514301e-06,
+      "loss": 0.6549,
+      "step": 16676
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.868362418224682,
+      "learning_rate": 1.0568231319619632e-06,
+      "loss": 0.6426,
+      "step": 16677
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.310663782437872,
+      "learning_rate": 1.0567266798792897e-06,
+      "loss": 0.6328,
+      "step": 16678
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.8598871230571565,
+      "learning_rate": 1.0566302272671813e-06,
+      "loss": 0.6393,
+      "step": 16679
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.5012911232871575,
+      "learning_rate": 1.0565337741265385e-06,
+      "loss": 0.64,
+      "step": 16680
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.9277283398713165,
+      "learning_rate": 1.0564373204582603e-06,
+      "loss": 0.653,
+      "step": 16681
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.039165271995656,
+      "learning_rate": 1.056340866263248e-06,
+      "loss": 0.6224,
+      "step": 16682
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.244326660419785,
+      "learning_rate": 1.0562444115424014e-06,
+      "loss": 0.627,
+      "step": 16683
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.4415148151340906,
+      "learning_rate": 1.0561479562966204e-06,
+      "loss": 0.6094,
+      "step": 16684
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.568072653297996,
+      "learning_rate": 1.0560515005268058e-06,
+      "loss": 0.6771,
+      "step": 16685
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.9141406179170173,
+      "learning_rate": 1.0559550442338579e-06,
+      "loss": 0.6315,
+      "step": 16686
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.6346712013069107,
+      "learning_rate": 1.055858587418676e-06,
+      "loss": 0.6458,
+      "step": 16687
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.965162464929123,
+      "learning_rate": 1.0557621300821613e-06,
+      "loss": 0.6178,
+      "step": 16688
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.2566510461460276,
+      "learning_rate": 1.0556656722252137e-06,
+      "loss": 0.623,
+      "step": 16689
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.794409157675358,
+      "learning_rate": 1.0555692138487333e-06,
+      "loss": 0.666,
+      "step": 16690
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.265021112143821,
+      "learning_rate": 1.0554727549536206e-06,
+      "loss": 0.6543,
+      "step": 16691
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.032405073184279,
+      "learning_rate": 1.0553762955407757e-06,
+      "loss": 0.638,
+      "step": 16692
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.337590499103263,
+      "learning_rate": 1.0552798356110994e-06,
+      "loss": 0.6465,
+      "step": 16693
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.158241458150883,
+      "learning_rate": 1.055183375165491e-06,
+      "loss": 0.7031,
+      "step": 16694
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.802378820596823,
+      "learning_rate": 1.0550869142048514e-06,
+      "loss": 0.6699,
+      "step": 16695
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.3010590245009515,
+      "learning_rate": 1.054990452730081e-06,
+      "loss": 0.6237,
+      "step": 16696
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.312385183750721,
+      "learning_rate": 1.0548939907420795e-06,
+      "loss": 0.6569,
+      "step": 16697
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.024915219980326,
+      "learning_rate": 1.0547975282417476e-06,
+      "loss": 0.6237,
+      "step": 16698
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.3885570776936893,
+      "learning_rate": 1.0547010652299857e-06,
+      "loss": 0.6068,
+      "step": 16699
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.769171024613509,
+      "learning_rate": 1.0546046017076938e-06,
+      "loss": 0.6504,
+      "step": 16700
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.0133488664557677,
+      "learning_rate": 1.0545081376757722e-06,
+      "loss": 0.6061,
+      "step": 16701
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.243736316639891,
+      "learning_rate": 1.0544116731351214e-06,
+      "loss": 0.6439,
+      "step": 16702
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.570173735816213,
+      "learning_rate": 1.0543152080866419e-06,
+      "loss": 0.6751,
+      "step": 16703
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.279643109820581,
+      "learning_rate": 1.0542187425312335e-06,
+      "loss": 0.6211,
+      "step": 16704
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.717728366042756,
+      "learning_rate": 1.0541222764697966e-06,
+      "loss": 0.6374,
+      "step": 16705
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.380698415364068,
+      "learning_rate": 1.054025809903232e-06,
+      "loss": 0.6328,
+      "step": 16706
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.231476291625808,
+      "learning_rate": 1.0539293428324397e-06,
+      "loss": 0.6133,
+      "step": 16707
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.616093548162246,
+      "learning_rate": 1.05383287525832e-06,
+      "loss": 0.7266,
+      "step": 16708
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.005052292412537,
+      "learning_rate": 1.0537364071817737e-06,
+      "loss": 0.6497,
+      "step": 16709
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.6767961023494746,
+      "learning_rate": 1.0536399386037003e-06,
+      "loss": 0.6758,
+      "step": 16710
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.5507605047762816,
+      "learning_rate": 1.0535434695250005e-06,
+      "loss": 0.6341,
+      "step": 16711
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.688375936009811,
+      "learning_rate": 1.053446999946575e-06,
+      "loss": 0.6198,
+      "step": 16712
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 6.3035526342155945,
+      "learning_rate": 1.053350529869324e-06,
+      "loss": 0.6478,
+      "step": 16713
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.376976823930491,
+      "learning_rate": 1.053254059294148e-06,
+      "loss": 0.6478,
+      "step": 16714
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.914395199084868,
+      "learning_rate": 1.0531575882219466e-06,
+      "loss": 0.6634,
+      "step": 16715
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.756798056678842,
+      "learning_rate": 1.053061116653621e-06,
+      "loss": 0.6471,
+      "step": 16716
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.944839166750393,
+      "learning_rate": 1.0529646445900715e-06,
+      "loss": 0.64,
+      "step": 16717
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.2655853285664604,
+      "learning_rate": 1.052868172032198e-06,
+      "loss": 0.6243,
+      "step": 16718
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.927548861317072,
+      "learning_rate": 1.0527716989809011e-06,
+      "loss": 0.5977,
+      "step": 16719
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.3443598274521165,
+      "learning_rate": 1.0526752254370816e-06,
+      "loss": 0.6224,
+      "step": 16720
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.063780809139767,
+      "learning_rate": 1.0525787514016393e-06,
+      "loss": 0.6862,
+      "step": 16721
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.7977920798587825,
+      "learning_rate": 1.0524822768754748e-06,
+      "loss": 0.582,
+      "step": 16722
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.182097409983831,
+      "learning_rate": 1.052385801859489e-06,
+      "loss": 0.6673,
+      "step": 16723
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.56261501170903,
+      "learning_rate": 1.0522893263545814e-06,
+      "loss": 0.6445,
+      "step": 16724
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.061784718040665,
+      "learning_rate": 1.0521928503616534e-06,
+      "loss": 0.6706,
+      "step": 16725
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 4.016458520348658,
+      "learning_rate": 1.0520963738816045e-06,
+      "loss": 0.6803,
+      "step": 16726
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 3.0978559758084847,
+      "learning_rate": 1.0519998969153355e-06,
+      "loss": 0.6595,
+      "step": 16727
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 5.608128065479814,
+      "learning_rate": 1.051903419463747e-06,
+      "loss": 0.6452,
+      "step": 16728
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.769415135650263,
+      "learning_rate": 1.0518069415277393e-06,
+      "loss": 0.6387,
+      "step": 16729
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.441555535563476,
+      "learning_rate": 1.051710463108213e-06,
+      "loss": 0.6595,
+      "step": 16730
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.6266426018794027,
+      "learning_rate": 1.051613984206068e-06,
+      "loss": 0.6094,
+      "step": 16731
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.4514590323750554,
+      "learning_rate": 1.0515175048222055e-06,
+      "loss": 0.6517,
+      "step": 16732
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.011522106382112,
+      "learning_rate": 1.0514210249575255e-06,
+      "loss": 0.6289,
+      "step": 16733
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.9578704899716106,
+      "learning_rate": 1.0513245446129282e-06,
+      "loss": 0.6361,
+      "step": 16734
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.299102980851394,
+      "learning_rate": 1.0512280637893147e-06,
+      "loss": 0.6315,
+      "step": 16735
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.8931812331011773,
+      "learning_rate": 1.0511315824875852e-06,
+      "loss": 0.6556,
+      "step": 16736
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.2895052345367954,
+      "learning_rate": 1.0510351007086399e-06,
+      "loss": 0.6315,
+      "step": 16737
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.033732148528424,
+      "learning_rate": 1.0509386184533795e-06,
+      "loss": 0.6745,
+      "step": 16738
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.896568761994149,
+      "learning_rate": 1.0508421357227046e-06,
+      "loss": 0.6413,
+      "step": 16739
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.101849273148128,
+      "learning_rate": 1.0507456525175153e-06,
+      "loss": 0.6445,
+      "step": 16740
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.418778132126597,
+      "learning_rate": 1.0506491688387128e-06,
+      "loss": 0.6764,
+      "step": 16741
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.9682339880868005,
+      "learning_rate": 1.0505526846871968e-06,
+      "loss": 0.6035,
+      "step": 16742
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.895897741437786,
+      "learning_rate": 1.0504562000638683e-06,
+      "loss": 0.6842,
+      "step": 16743
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.7251127461996227,
+      "learning_rate": 1.0503597149696273e-06,
+      "loss": 0.6523,
+      "step": 16744
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.4211880032357396,
+      "learning_rate": 1.0502632294053748e-06,
+      "loss": 0.6621,
+      "step": 16745
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.576379416943239,
+      "learning_rate": 1.0501667433720114e-06,
+      "loss": 0.6302,
+      "step": 16746
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.452159982104053,
+      "learning_rate": 1.050070256870437e-06,
+      "loss": 0.6243,
+      "step": 16747
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.268584769125812,
+      "learning_rate": 1.0499737699015528e-06,
+      "loss": 0.5866,
+      "step": 16748
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.360959484686502,
+      "learning_rate": 1.0498772824662587e-06,
+      "loss": 0.6901,
+      "step": 16749
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.295397658194169,
+      "learning_rate": 1.0497807945654554e-06,
+      "loss": 0.6354,
+      "step": 16750
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.029717481595826,
+      "learning_rate": 1.0496843062000437e-06,
+      "loss": 0.6602,
+      "step": 16751
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.544096062369667,
+      "learning_rate": 1.049587817370924e-06,
+      "loss": 0.6816,
+      "step": 16752
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.941999247644441,
+      "learning_rate": 1.0494913280789968e-06,
+      "loss": 0.6829,
+      "step": 16753
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.7908619958665755,
+      "learning_rate": 1.0493948383251628e-06,
+      "loss": 0.6328,
+      "step": 16754
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.6352911035313995,
+      "learning_rate": 1.0492983481103222e-06,
+      "loss": 0.6432,
+      "step": 16755
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 5.567766226807062,
+      "learning_rate": 1.049201857435376e-06,
+      "loss": 0.6419,
+      "step": 16756
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.022678771846916,
+      "learning_rate": 1.0491053663012242e-06,
+      "loss": 0.623,
+      "step": 16757
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.111370358056637,
+      "learning_rate": 1.0490088747087678e-06,
+      "loss": 0.6523,
+      "step": 16758
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.3622661802496845,
+      "learning_rate": 1.0489123826589074e-06,
+      "loss": 0.6367,
+      "step": 16759
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.9601910504267104,
+      "learning_rate": 1.0488158901525433e-06,
+      "loss": 0.6322,
+      "step": 16760
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.6160107550964744,
+      "learning_rate": 1.0487193971905758e-06,
+      "loss": 0.6152,
+      "step": 16761
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.8576447882195306,
+      "learning_rate": 1.0486229037739064e-06,
+      "loss": 0.6152,
+      "step": 16762
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.7153006546520815,
+      "learning_rate": 1.048526409903435e-06,
+      "loss": 0.6367,
+      "step": 16763
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.0495556127129495,
+      "learning_rate": 1.0484299155800621e-06,
+      "loss": 0.6816,
+      "step": 16764
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.718525053602093,
+      "learning_rate": 1.048333420804689e-06,
+      "loss": 0.6172,
+      "step": 16765
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.3418754184824455,
+      "learning_rate": 1.0482369255782154e-06,
+      "loss": 0.6706,
+      "step": 16766
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.256156679092653,
+      "learning_rate": 1.0481404299015425e-06,
+      "loss": 0.6562,
+      "step": 16767
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.334017989046861,
+      "learning_rate": 1.0480439337755705e-06,
+      "loss": 0.6367,
+      "step": 16768
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.974865920204815,
+      "learning_rate": 1.0479474372012004e-06,
+      "loss": 0.6302,
+      "step": 16769
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.054422475176884,
+      "learning_rate": 1.0478509401793324e-06,
+      "loss": 0.5739,
+      "step": 16770
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.476320939669179,
+      "learning_rate": 1.0477544427108674e-06,
+      "loss": 0.679,
+      "step": 16771
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.496081263166006,
+      "learning_rate": 1.0476579447967061e-06,
+      "loss": 0.6009,
+      "step": 16772
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.9906720425542486,
+      "learning_rate": 1.0475614464377492e-06,
+      "loss": 0.6569,
+      "step": 16773
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.3642529787155286,
+      "learning_rate": 1.0474649476348967e-06,
+      "loss": 0.6322,
+      "step": 16774
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.0537632169519098,
+      "learning_rate": 1.0473684483890499e-06,
+      "loss": 0.6543,
+      "step": 16775
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.792993748673278,
+      "learning_rate": 1.047271948701109e-06,
+      "loss": 0.638,
+      "step": 16776
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.294510271892601,
+      "learning_rate": 1.0471754485719749e-06,
+      "loss": 0.6048,
+      "step": 16777
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.371707552064574,
+      "learning_rate": 1.047078948002548e-06,
+      "loss": 0.6536,
+      "step": 16778
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.384140837912468,
+      "learning_rate": 1.0469824469937297e-06,
+      "loss": 0.6667,
+      "step": 16779
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.149059903133516,
+      "learning_rate": 1.0468859455464197e-06,
+      "loss": 0.6543,
+      "step": 16780
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.295928642888411,
+      "learning_rate": 1.0467894436615188e-06,
+      "loss": 0.6478,
+      "step": 16781
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.8025843787768348,
+      "learning_rate": 1.0466929413399279e-06,
+      "loss": 0.6471,
+      "step": 16782
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.366872975144451,
+      "learning_rate": 1.046596438582548e-06,
+      "loss": 0.6888,
+      "step": 16783
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.389247387421261,
+      "learning_rate": 1.046499935390279e-06,
+      "loss": 0.6315,
+      "step": 16784
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.1100411519302362,
+      "learning_rate": 1.0464034317640226e-06,
+      "loss": 0.6784,
+      "step": 16785
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.4815814861314633,
+      "learning_rate": 1.0463069277046783e-06,
+      "loss": 0.7305,
+      "step": 16786
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.985796946673073,
+      "learning_rate": 1.0462104232131474e-06,
+      "loss": 0.6589,
+      "step": 16787
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.8183534558678915,
+      "learning_rate": 1.0461139182903308e-06,
+      "loss": 0.6615,
+      "step": 16788
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.2151204026361437,
+      "learning_rate": 1.046017412937129e-06,
+      "loss": 0.6452,
+      "step": 16789
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.485129235655617,
+      "learning_rate": 1.0459209071544422e-06,
+      "loss": 0.6582,
+      "step": 16790
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.003686178808569,
+      "learning_rate": 1.0458244009431718e-06,
+      "loss": 0.6595,
+      "step": 16791
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.050748923792225,
+      "learning_rate": 1.0457278943042182e-06,
+      "loss": 0.6178,
+      "step": 16792
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 4.398489534484192,
+      "learning_rate": 1.045631387238482e-06,
+      "loss": 0.6257,
+      "step": 16793
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.813938637735066,
+      "learning_rate": 1.045534879746864e-06,
+      "loss": 0.6758,
+      "step": 16794
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.897223892881052,
+      "learning_rate": 1.045438371830265e-06,
+      "loss": 0.6536,
+      "step": 16795
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.0085196097650293,
+      "learning_rate": 1.0453418634895858e-06,
+      "loss": 0.6777,
+      "step": 16796
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.0780471664656255,
+      "learning_rate": 1.0452453547257268e-06,
+      "loss": 0.6393,
+      "step": 16797
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.517126195748907,
+      "learning_rate": 1.0451488455395892e-06,
+      "loss": 0.6536,
+      "step": 16798
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.5176745214858913,
+      "learning_rate": 1.045052335932073e-06,
+      "loss": 0.6426,
+      "step": 16799
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.4632282609999216,
+      "learning_rate": 1.0449558259040797e-06,
+      "loss": 0.6868,
+      "step": 16800
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.6508244282627706,
+      "learning_rate": 1.0448593154565095e-06,
+      "loss": 0.6478,
+      "step": 16801
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.769249897437627,
+      "learning_rate": 1.0447628045902635e-06,
+      "loss": 0.6309,
+      "step": 16802
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.579583371805833,
+      "learning_rate": 1.044666293306242e-06,
+      "loss": 0.6309,
+      "step": 16803
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.466951277208296,
+      "learning_rate": 1.0445697816053462e-06,
+      "loss": 0.6367,
+      "step": 16804
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.534608687480673,
+      "learning_rate": 1.0444732694884767e-06,
+      "loss": 0.6432,
+      "step": 16805
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.7093044289001287,
+      "learning_rate": 1.0443767569565343e-06,
+      "loss": 0.6458,
+      "step": 16806
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.189676580540413,
+      "learning_rate": 1.0442802440104198e-06,
+      "loss": 0.6133,
+      "step": 16807
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.1595802581092363,
+      "learning_rate": 1.0441837306510337e-06,
+      "loss": 0.6419,
+      "step": 16808
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.377334378186979,
+      "learning_rate": 1.044087216879277e-06,
+      "loss": 0.6094,
+      "step": 16809
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.8307888316350693,
+      "learning_rate": 1.0439907026960504e-06,
+      "loss": 0.6276,
+      "step": 16810
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.7086016944132236,
+      "learning_rate": 1.0438941881022544e-06,
+      "loss": 0.6309,
+      "step": 16811
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.496447590868196,
+      "learning_rate": 1.0437976730987905e-06,
+      "loss": 0.6497,
+      "step": 16812
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.5476043240475006,
+      "learning_rate": 1.0437011576865589e-06,
+      "loss": 0.6237,
+      "step": 16813
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.8274388750962447,
+      "learning_rate": 1.0436046418664603e-06,
+      "loss": 0.6543,
+      "step": 16814
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.4005856977557203,
+      "learning_rate": 1.0435081256393961e-06,
+      "loss": 0.653,
+      "step": 16815
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.85829946354656,
+      "learning_rate": 1.0434116090062663e-06,
+      "loss": 0.6491,
+      "step": 16816
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.0806542652550637,
+      "learning_rate": 1.0433150919679726e-06,
+      "loss": 0.6523,
+      "step": 16817
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.8663948378952506,
+      "learning_rate": 1.043218574525415e-06,
+      "loss": 0.6087,
+      "step": 16818
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.05654041414428,
+      "learning_rate": 1.0431220566794945e-06,
+      "loss": 0.6452,
+      "step": 16819
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.124031016798399,
+      "learning_rate": 1.0430255384311124e-06,
+      "loss": 0.6146,
+      "step": 16820
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.6244175208647893,
+      "learning_rate": 1.0429290197811686e-06,
+      "loss": 0.5983,
+      "step": 16821
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.94898598530027,
+      "learning_rate": 1.042832500730565e-06,
+      "loss": 0.6432,
+      "step": 16822
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.130353366002865,
+      "learning_rate": 1.0427359812802016e-06,
+      "loss": 0.6484,
+      "step": 16823
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.266708555630383,
+      "learning_rate": 1.0426394614309798e-06,
+      "loss": 0.6458,
+      "step": 16824
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.2322756531026196,
+      "learning_rate": 1.0425429411838e-06,
+      "loss": 0.6315,
+      "step": 16825
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.3430377909551647,
+      "learning_rate": 1.0424464205395632e-06,
+      "loss": 0.6289,
+      "step": 16826
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.6507643804923835,
+      "learning_rate": 1.0423498994991697e-06,
+      "loss": 0.61,
+      "step": 16827
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.5917225126610304,
+      "learning_rate": 1.0422533780635216e-06,
+      "loss": 0.6367,
+      "step": 16828
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.6280323187597783,
+      "learning_rate": 1.0421568562335187e-06,
+      "loss": 0.6549,
+      "step": 16829
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.0619797283483043,
+      "learning_rate": 1.042060334010062e-06,
+      "loss": 0.6387,
+      "step": 16830
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.2535396578254043,
+      "learning_rate": 1.0419638113940523e-06,
+      "loss": 0.6296,
+      "step": 16831
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.375827777913524,
+      "learning_rate": 1.041867288386391e-06,
+      "loss": 0.6328,
+      "step": 16832
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.652719385605637,
+      "learning_rate": 1.0417707649879786e-06,
+      "loss": 0.6283,
+      "step": 16833
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.289764564048605,
+      "learning_rate": 1.0416742411997157e-06,
+      "loss": 0.6064,
+      "step": 16834
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.6811402740105716,
+      "learning_rate": 1.0415777170225036e-06,
+      "loss": 0.6732,
+      "step": 16835
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.04704352376943,
+      "learning_rate": 1.041481192457243e-06,
+      "loss": 0.5729,
+      "step": 16836
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.172247387647493,
+      "learning_rate": 1.0413846675048345e-06,
+      "loss": 0.6465,
+      "step": 16837
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.169652268384149,
+      "learning_rate": 1.0412881421661797e-06,
+      "loss": 0.6393,
+      "step": 16838
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.2611872674461173,
+      "learning_rate": 1.0411916164421789e-06,
+      "loss": 0.6406,
+      "step": 16839
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.912198703220977,
+      "learning_rate": 1.0410950903337325e-06,
+      "loss": 0.6745,
+      "step": 16840
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 4.90658028254536,
+      "learning_rate": 1.0409985638417428e-06,
+      "loss": 0.6107,
+      "step": 16841
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.168900241529142,
+      "learning_rate": 1.0409020369671096e-06,
+      "loss": 0.6836,
+      "step": 16842
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.905280853854452,
+      "learning_rate": 1.0408055097107337e-06,
+      "loss": 0.6113,
+      "step": 16843
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.480622800575258,
+      "learning_rate": 1.040708982073517e-06,
+      "loss": 0.5729,
+      "step": 16844
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.715339276559126,
+      "learning_rate": 1.0406124540563591e-06,
+      "loss": 0.6576,
+      "step": 16845
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.262382940583709,
+      "learning_rate": 1.0405159256601623e-06,
+      "loss": 0.6602,
+      "step": 16846
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.4040865945744705,
+      "learning_rate": 1.040419396885826e-06,
+      "loss": 0.6536,
+      "step": 16847
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.8465753603918986,
+      "learning_rate": 1.0403228677342527e-06,
+      "loss": 0.6286,
+      "step": 16848
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.8638413179273825,
+      "learning_rate": 1.0402263382063421e-06,
+      "loss": 0.6628,
+      "step": 16849
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.071647709695875,
+      "learning_rate": 1.0401298083029953e-06,
+      "loss": 0.6693,
+      "step": 16850
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.135104193441353,
+      "learning_rate": 1.0400332780251137e-06,
+      "loss": 0.6562,
+      "step": 16851
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.9764113992256758,
+      "learning_rate": 1.039936747373598e-06,
+      "loss": 0.6081,
+      "step": 16852
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.036487214241292,
+      "learning_rate": 1.039840216349349e-06,
+      "loss": 0.6022,
+      "step": 16853
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.3880312819673533,
+      "learning_rate": 1.0397436849532679e-06,
+      "loss": 0.6289,
+      "step": 16854
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.395276063035004,
+      "learning_rate": 1.0396471531862554e-06,
+      "loss": 0.6107,
+      "step": 16855
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 5.695329658150383,
+      "learning_rate": 1.0395506210492125e-06,
+      "loss": 0.5833,
+      "step": 16856
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.9039239352067248,
+      "learning_rate": 1.03945408854304e-06,
+      "loss": 0.6178,
+      "step": 16857
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.3885379673690332,
+      "learning_rate": 1.0393575556686394e-06,
+      "loss": 0.5833,
+      "step": 16858
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.7203741123696816,
+      "learning_rate": 1.0392610224269108e-06,
+      "loss": 0.6439,
+      "step": 16859
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.6777559271010483,
+      "learning_rate": 1.0391644888187559e-06,
+      "loss": 0.6497,
+      "step": 16860
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.1841910598561585,
+      "learning_rate": 1.0390679548450754e-06,
+      "loss": 0.61,
+      "step": 16861
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.1327452340226136,
+      "learning_rate": 1.03897142050677e-06,
+      "loss": 0.5918,
+      "step": 16862
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 6.473037429079463,
+      "learning_rate": 1.038874885804741e-06,
+      "loss": 0.5749,
+      "step": 16863
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.1180108314384625,
+      "learning_rate": 1.0387783507398891e-06,
+      "loss": 0.6133,
+      "step": 16864
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.14282703861017,
+      "learning_rate": 1.0386818153131158e-06,
+      "loss": 0.6276,
+      "step": 16865
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.626832529036941,
+      "learning_rate": 1.0385852795253211e-06,
+      "loss": 0.6035,
+      "step": 16866
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.468196128621624,
+      "learning_rate": 1.0384887433774072e-06,
+      "loss": 0.6165,
+      "step": 16867
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.385336341412554,
+      "learning_rate": 1.0383922068702742e-06,
+      "loss": 0.6243,
+      "step": 16868
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 6.518745311362683,
+      "learning_rate": 1.038295670004823e-06,
+      "loss": 0.7253,
+      "step": 16869
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.595794793416118,
+      "learning_rate": 1.0381991327819555e-06,
+      "loss": 0.6517,
+      "step": 16870
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.327302818124801,
+      "learning_rate": 1.038102595202572e-06,
+      "loss": 0.6315,
+      "step": 16871
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.864622619111131,
+      "learning_rate": 1.0380060572675734e-06,
+      "loss": 0.6465,
+      "step": 16872
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.8697149342752524,
+      "learning_rate": 1.0379095189778608e-06,
+      "loss": 0.6191,
+      "step": 16873
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.935109975975967,
+      "learning_rate": 1.0378129803343357e-06,
+      "loss": 0.5951,
+      "step": 16874
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.717897118414862,
+      "learning_rate": 1.0377164413378984e-06,
+      "loss": 0.61,
+      "step": 16875
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.6887003137918732,
+      "learning_rate": 1.0376199019894502e-06,
+      "loss": 0.5905,
+      "step": 16876
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.3232395517461955,
+      "learning_rate": 1.0375233622898922e-06,
+      "loss": 0.5859,
+      "step": 16877
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.536613367417289,
+      "learning_rate": 1.0374268222401257e-06,
+      "loss": 0.6191,
+      "step": 16878
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.4562617557235753,
+      "learning_rate": 1.0373302818410508e-06,
+      "loss": 0.6406,
+      "step": 16879
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.7614115339185172,
+      "learning_rate": 1.0372337410935695e-06,
+      "loss": 0.5625,
+      "step": 16880
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.3207548200839225,
+      "learning_rate": 1.0371371999985823e-06,
+      "loss": 0.696,
+      "step": 16881
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 6.745829387297003,
+      "learning_rate": 1.0370406585569899e-06,
+      "loss": 0.6608,
+      "step": 16882
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.459179359641243,
+      "learning_rate": 1.036944116769694e-06,
+      "loss": 0.5859,
+      "step": 16883
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.7923002913373485,
+      "learning_rate": 1.0368475746375958e-06,
+      "loss": 0.6139,
+      "step": 16884
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.325384526677914,
+      "learning_rate": 1.0367510321615955e-06,
+      "loss": 0.5586,
+      "step": 16885
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.5203447966497903,
+      "learning_rate": 1.036654489342595e-06,
+      "loss": 0.5898,
+      "step": 16886
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.044791252459728,
+      "learning_rate": 1.0365579461814943e-06,
+      "loss": 0.6732,
+      "step": 16887
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.790810237550608,
+      "learning_rate": 1.0364614026791954e-06,
+      "loss": 0.6862,
+      "step": 16888
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.254415533328343,
+      "learning_rate": 1.036364858836599e-06,
+      "loss": 0.6712,
+      "step": 16889
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.6322385147089205,
+      "learning_rate": 1.0362683146546059e-06,
+      "loss": 0.6439,
+      "step": 16890
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.956128850478375,
+      "learning_rate": 1.036171770134118e-06,
+      "loss": 0.627,
+      "step": 16891
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.059931893321662,
+      "learning_rate": 1.0360752252760353e-06,
+      "loss": 0.694,
+      "step": 16892
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.6495638765664395,
+      "learning_rate": 1.0359786800812591e-06,
+      "loss": 0.6198,
+      "step": 16893
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.7291146911585407,
+      "learning_rate": 1.0358821345506914e-06,
+      "loss": 0.6654,
+      "step": 16894
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.595924551402929,
+      "learning_rate": 1.035785588685232e-06,
+      "loss": 0.6354,
+      "step": 16895
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.591699386948632,
+      "learning_rate": 1.0356890424857829e-06,
+      "loss": 0.6289,
+      "step": 16896
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.574264984428751,
+      "learning_rate": 1.0355924959532444e-06,
+      "loss": 0.6263,
+      "step": 16897
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.521292964190738,
+      "learning_rate": 1.0354959490885182e-06,
+      "loss": 0.6719,
+      "step": 16898
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.2317268403116177,
+      "learning_rate": 1.0353994018925054e-06,
+      "loss": 0.6875,
+      "step": 16899
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.63727486693808,
+      "learning_rate": 1.0353028543661065e-06,
+      "loss": 0.6341,
+      "step": 16900
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.7849971217273124,
+      "learning_rate": 1.0352063065102233e-06,
+      "loss": 0.584,
+      "step": 16901
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.375820447758978,
+      "learning_rate": 1.0351097583257563e-06,
+      "loss": 0.6289,
+      "step": 16902
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.9423738166019895,
+      "learning_rate": 1.0350132098136067e-06,
+      "loss": 0.6276,
+      "step": 16903
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.587760954863385,
+      "learning_rate": 1.0349166609746758e-06,
+      "loss": 0.6016,
+      "step": 16904
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.106410507306074,
+      "learning_rate": 1.0348201118098647e-06,
+      "loss": 0.6504,
+      "step": 16905
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.9787260095441273,
+      "learning_rate": 1.0347235623200742e-06,
+      "loss": 0.6621,
+      "step": 16906
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.3421683733743284,
+      "learning_rate": 1.034627012506206e-06,
+      "loss": 0.6621,
+      "step": 16907
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.63148219550678,
+      "learning_rate": 1.0345304623691606e-06,
+      "loss": 0.6647,
+      "step": 16908
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.227222648296845,
+      "learning_rate": 1.0344339119098393e-06,
+      "loss": 0.6296,
+      "step": 16909
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.803856408744923,
+      "learning_rate": 1.0343373611291432e-06,
+      "loss": 0.6426,
+      "step": 16910
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.3252775141379436,
+      "learning_rate": 1.034240810027974e-06,
+      "loss": 0.6549,
+      "step": 16911
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.1476286315965956,
+      "learning_rate": 1.0341442586072318e-06,
+      "loss": 0.6517,
+      "step": 16912
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.342360864501622,
+      "learning_rate": 1.034047706867818e-06,
+      "loss": 0.6315,
+      "step": 16913
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.3686264303463855,
+      "learning_rate": 1.0339511548106345e-06,
+      "loss": 0.6569,
+      "step": 16914
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.383742015239332,
+      "learning_rate": 1.0338546024365818e-06,
+      "loss": 0.6458,
+      "step": 16915
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.403902516941529,
+      "learning_rate": 1.0337580497465606e-06,
+      "loss": 0.6296,
+      "step": 16916
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.928045948441593,
+      "learning_rate": 1.033661496741473e-06,
+      "loss": 0.6315,
+      "step": 16917
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.956286664282401,
+      "learning_rate": 1.0335649434222196e-06,
+      "loss": 0.6549,
+      "step": 16918
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.7099256663336013,
+      "learning_rate": 1.0334683897897015e-06,
+      "loss": 0.6999,
+      "step": 16919
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.5799919071305224,
+      "learning_rate": 1.03337183584482e-06,
+      "loss": 0.6302,
+      "step": 16920
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.746335990017424,
+      "learning_rate": 1.0332752815884763e-06,
+      "loss": 0.6432,
+      "step": 16921
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.0203873408817477,
+      "learning_rate": 1.033178727021571e-06,
+      "loss": 0.6126,
+      "step": 16922
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.276576389073642,
+      "learning_rate": 1.0330821721450064e-06,
+      "loss": 0.6084,
+      "step": 16923
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.3958092069765593,
+      "learning_rate": 1.0329856169596826e-06,
+      "loss": 0.6439,
+      "step": 16924
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.2557450818829885,
+      "learning_rate": 1.032889061466501e-06,
+      "loss": 0.625,
+      "step": 16925
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.8138592441800476,
+      "learning_rate": 1.032792505666363e-06,
+      "loss": 0.6413,
+      "step": 16926
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.2377670908764,
+      "learning_rate": 1.0326959495601699e-06,
+      "loss": 0.6276,
+      "step": 16927
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 4.996683443409343,
+      "learning_rate": 1.0325993931488223e-06,
+      "loss": 0.6491,
+      "step": 16928
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.033095670463723,
+      "learning_rate": 1.0325028364332218e-06,
+      "loss": 0.6348,
+      "step": 16929
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 5.37890379229039,
+      "learning_rate": 1.032406279414269e-06,
+      "loss": 0.6393,
+      "step": 16930
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.0587714988480887,
+      "learning_rate": 1.0323097220928665e-06,
+      "loss": 0.609,
+      "step": 16931
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.870848905582098,
+      "learning_rate": 1.0322131644699137e-06,
+      "loss": 0.6634,
+      "step": 16932
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.2644115480494134,
+      "learning_rate": 1.0321166065463129e-06,
+      "loss": 0.6608,
+      "step": 16933
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.8621008009384044,
+      "learning_rate": 1.032020048322965e-06,
+      "loss": 0.6178,
+      "step": 16934
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.631854640501059,
+      "learning_rate": 1.031923489800771e-06,
+      "loss": 0.6276,
+      "step": 16935
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.7336011665859177,
+      "learning_rate": 1.0318269309806321e-06,
+      "loss": 0.6296,
+      "step": 16936
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.022436065437599,
+      "learning_rate": 1.0317303718634503e-06,
+      "loss": 0.6471,
+      "step": 16937
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.576371112325695,
+      "learning_rate": 1.0316338124501254e-06,
+      "loss": 0.6263,
+      "step": 16938
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.69136043643536,
+      "learning_rate": 1.0315372527415598e-06,
+      "loss": 0.6335,
+      "step": 16939
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.508859877332182,
+      "learning_rate": 1.0314406927386538e-06,
+      "loss": 0.6243,
+      "step": 16940
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.141714704748336,
+      "learning_rate": 1.0313441324423096e-06,
+      "loss": 0.6204,
+      "step": 16941
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.886585082964473,
+      "learning_rate": 1.0312475718534273e-06,
+      "loss": 0.6706,
+      "step": 16942
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.3342934096397627,
+      "learning_rate": 1.0311510109729087e-06,
+      "loss": 0.6296,
+      "step": 16943
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.1539401952336776,
+      "learning_rate": 1.0310544498016555e-06,
+      "loss": 0.6042,
+      "step": 16944
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.2137392285151805,
+      "learning_rate": 1.0309578883405679e-06,
+      "loss": 0.6523,
+      "step": 16945
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.1430680961427564,
+      "learning_rate": 1.0308613265905476e-06,
+      "loss": 0.6315,
+      "step": 16946
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.3096091617782952,
+      "learning_rate": 1.0307647645524961e-06,
+      "loss": 0.6986,
+      "step": 16947
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.8193556925369374,
+      "learning_rate": 1.0306682022273138e-06,
+      "loss": 0.5951,
+      "step": 16948
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 6.471368748115347,
+      "learning_rate": 1.030571639615903e-06,
+      "loss": 0.6263,
+      "step": 16949
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.165546528787982,
+      "learning_rate": 1.0304750767191644e-06,
+      "loss": 0.5993,
+      "step": 16950
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.7122828631331304,
+      "learning_rate": 1.0303785135379987e-06,
+      "loss": 0.6302,
+      "step": 16951
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.929622944754282,
+      "learning_rate": 1.030281950073308e-06,
+      "loss": 0.668,
+      "step": 16952
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.885184290853034,
+      "learning_rate": 1.0301853863259932e-06,
+      "loss": 0.6576,
+      "step": 16953
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.2622111170047163,
+      "learning_rate": 1.0300888222969553e-06,
+      "loss": 0.6413,
+      "step": 16954
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.013025497622891,
+      "learning_rate": 1.029992257987096e-06,
+      "loss": 0.6686,
+      "step": 16955
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.2390687374288483,
+      "learning_rate": 1.0298956933973159e-06,
+      "loss": 0.6289,
+      "step": 16956
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.571458793732338,
+      "learning_rate": 1.0297991285285175e-06,
+      "loss": 0.6712,
+      "step": 16957
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.7870080295711706,
+      "learning_rate": 1.0297025633816004e-06,
+      "loss": 0.6328,
+      "step": 16958
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.269729123902515,
+      "learning_rate": 1.0296059979574667e-06,
+      "loss": 0.6784,
+      "step": 16959
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.0289114903120464,
+      "learning_rate": 1.0295094322570181e-06,
+      "loss": 0.6283,
+      "step": 16960
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.155955313285144,
+      "learning_rate": 1.029412866281155e-06,
+      "loss": 0.6549,
+      "step": 16961
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.748958958076715,
+      "learning_rate": 1.0293163000307791e-06,
+      "loss": 0.6536,
+      "step": 16962
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.924767533611446,
+      "learning_rate": 1.0292197335067914e-06,
+      "loss": 0.6094,
+      "step": 16963
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.138244400654443,
+      "learning_rate": 1.0291231667100938e-06,
+      "loss": 0.6562,
+      "step": 16964
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.0473493295180787,
+      "learning_rate": 1.0290265996415868e-06,
+      "loss": 0.6048,
+      "step": 16965
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 7.916954519236495,
+      "learning_rate": 1.0289300323021717e-06,
+      "loss": 0.6576,
+      "step": 16966
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.748272753118663,
+      "learning_rate": 1.0288334646927504e-06,
+      "loss": 0.6055,
+      "step": 16967
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.624423801116614,
+      "learning_rate": 1.0287368968142238e-06,
+      "loss": 0.6556,
+      "step": 16968
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.534704487828142,
+      "learning_rate": 1.0286403286674933e-06,
+      "loss": 0.6855,
+      "step": 16969
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.913687908816576,
+      "learning_rate": 1.02854376025346e-06,
+      "loss": 0.5957,
+      "step": 16970
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.493494314884079,
+      "learning_rate": 1.0284471915730251e-06,
+      "loss": 0.64,
+      "step": 16971
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.030761023811398,
+      "learning_rate": 1.0283506226270898e-06,
+      "loss": 0.6712,
+      "step": 16972
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.103160413351532,
+      "learning_rate": 1.0282540534165562e-06,
+      "loss": 0.6393,
+      "step": 16973
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.7894548361082463,
+      "learning_rate": 1.028157483942325e-06,
+      "loss": 0.5944,
+      "step": 16974
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.704451461619044,
+      "learning_rate": 1.0280609142052971e-06,
+      "loss": 0.6237,
+      "step": 16975
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 4.649757037100336,
+      "learning_rate": 1.0279643442063745e-06,
+      "loss": 0.6784,
+      "step": 16976
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.237382350531514,
+      "learning_rate": 1.0278677739464583e-06,
+      "loss": 0.668,
+      "step": 16977
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.9009242251303857,
+      "learning_rate": 1.0277712034264495e-06,
+      "loss": 0.6393,
+      "step": 16978
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.7772656624080567,
+      "learning_rate": 1.0276746326472494e-06,
+      "loss": 0.6634,
+      "step": 16979
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.557062589984387,
+      "learning_rate": 1.0275780616097598e-06,
+      "loss": 0.6126,
+      "step": 16980
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.7830310392915756,
+      "learning_rate": 1.0274814903148818e-06,
+      "loss": 0.6243,
+      "step": 16981
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.308127196888093,
+      "learning_rate": 1.0273849187635164e-06,
+      "loss": 0.6087,
+      "step": 16982
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.232589411928265,
+      "learning_rate": 1.0272883469565653e-06,
+      "loss": 0.6751,
+      "step": 16983
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.8114845090372302,
+      "learning_rate": 1.0271917748949294e-06,
+      "loss": 0.6582,
+      "step": 16984
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.315144970813219,
+      "learning_rate": 1.0270952025795103e-06,
+      "loss": 0.6686,
+      "step": 16985
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.1228485360197045,
+      "learning_rate": 1.0269986300112094e-06,
+      "loss": 0.6992,
+      "step": 16986
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 8.963302604133256,
+      "learning_rate": 1.0269020571909282e-06,
+      "loss": 0.668,
+      "step": 16987
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.616484628215722,
+      "learning_rate": 1.0268054841195672e-06,
+      "loss": 0.6608,
+      "step": 16988
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.7266624539087263,
+      "learning_rate": 1.0267089107980285e-06,
+      "loss": 0.6517,
+      "step": 16989
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.758331338024889,
+      "learning_rate": 1.026612337227213e-06,
+      "loss": 0.6445,
+      "step": 16990
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 5.4219060661974305,
+      "learning_rate": 1.0265157634080223e-06,
+      "loss": 0.6315,
+      "step": 16991
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.866004844797063,
+      "learning_rate": 1.026419189341358e-06,
+      "loss": 0.6393,
+      "step": 16992
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.353460374987434,
+      "learning_rate": 1.0263226150281207e-06,
+      "loss": 0.638,
+      "step": 16993
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.984943994454884,
+      "learning_rate": 1.026226040469212e-06,
+      "loss": 0.61,
+      "step": 16994
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.8298685439601856,
+      "learning_rate": 1.0261294656655335e-06,
+      "loss": 0.6452,
+      "step": 16995
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 3.356274012499151,
+      "learning_rate": 1.026032890617986e-06,
+      "loss": 0.5749,
+      "step": 16996
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.982510016652706,
+      "learning_rate": 1.025936315327472e-06,
+      "loss": 0.6406,
+      "step": 16997
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1786592378131657,
+      "learning_rate": 1.0258397397948915e-06,
+      "loss": 0.6543,
+      "step": 16998
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.7592449686809255,
+      "learning_rate": 1.0257431640211467e-06,
+      "loss": 0.6042,
+      "step": 16999
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1912298226303157,
+      "learning_rate": 1.0256465880071388e-06,
+      "loss": 0.6543,
+      "step": 17000
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.9140604984488236,
+      "learning_rate": 1.0255500117537685e-06,
+      "loss": 0.6517,
+      "step": 17001
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.4456528726432127,
+      "learning_rate": 1.0254534352619379e-06,
+      "loss": 0.6986,
+      "step": 17002
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.915716736338164,
+      "learning_rate": 1.0253568585325484e-06,
+      "loss": 0.6576,
+      "step": 17003
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.8128014172338305,
+      "learning_rate": 1.025260281566501e-06,
+      "loss": 0.6224,
+      "step": 17004
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.5487706030611808,
+      "learning_rate": 1.025163704364697e-06,
+      "loss": 0.5983,
+      "step": 17005
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.9334467357329195,
+      "learning_rate": 1.0250671269280378e-06,
+      "loss": 0.6725,
+      "step": 17006
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.260989913662972,
+      "learning_rate": 1.0249705492574255e-06,
+      "loss": 0.6276,
+      "step": 17007
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.7675278001033896,
+      "learning_rate": 1.0248739713537602e-06,
+      "loss": 0.6322,
+      "step": 17008
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.982603285029481,
+      "learning_rate": 1.024777393217944e-06,
+      "loss": 0.612,
+      "step": 17009
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.6224919731429415,
+      "learning_rate": 1.0246808148508785e-06,
+      "loss": 0.6647,
+      "step": 17010
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.771904889760953,
+      "learning_rate": 1.0245842362534648e-06,
+      "loss": 0.5775,
+      "step": 17011
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.669237244123511,
+      "learning_rate": 1.024487657426604e-06,
+      "loss": 0.6602,
+      "step": 17012
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.403405177120064,
+      "learning_rate": 1.0243910783711977e-06,
+      "loss": 0.6797,
+      "step": 17013
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.949074749627071,
+      "learning_rate": 1.0242944990881475e-06,
+      "loss": 0.6185,
+      "step": 17014
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.182273666224407,
+      "learning_rate": 1.0241979195783546e-06,
+      "loss": 0.6745,
+      "step": 17015
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.784583981028072,
+      "learning_rate": 1.02410133984272e-06,
+      "loss": 0.6029,
+      "step": 17016
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1977899170539423,
+      "learning_rate": 1.0240047598821457e-06,
+      "loss": 0.6211,
+      "step": 17017
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.0891220924701144,
+      "learning_rate": 1.023908179697533e-06,
+      "loss": 0.6556,
+      "step": 17018
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.8840669075192067,
+      "learning_rate": 1.023811599289783e-06,
+      "loss": 0.6146,
+      "step": 17019
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.117501217478128,
+      "learning_rate": 1.0237150186597975e-06,
+      "loss": 0.6855,
+      "step": 17020
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.137367010642698,
+      "learning_rate": 1.0236184378084774e-06,
+      "loss": 0.6719,
+      "step": 17021
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.6465651143866573,
+      "learning_rate": 1.0235218567367242e-06,
+      "loss": 0.6504,
+      "step": 17022
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.556426746197336,
+      "learning_rate": 1.0234252754454397e-06,
+      "loss": 0.6693,
+      "step": 17023
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.072902659543418,
+      "learning_rate": 1.0233286939355248e-06,
+      "loss": 0.5999,
+      "step": 17024
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.918605703821626,
+      "learning_rate": 1.0232321122078812e-06,
+      "loss": 0.6667,
+      "step": 17025
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.80876348821034,
+      "learning_rate": 1.0231355302634106e-06,
+      "loss": 0.6257,
+      "step": 17026
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.26016906599224,
+      "learning_rate": 1.0230389481030134e-06,
+      "loss": 0.6497,
+      "step": 17027
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.7127502521366775,
+      "learning_rate": 1.0229423657275922e-06,
+      "loss": 0.6263,
+      "step": 17028
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.488701344900023,
+      "learning_rate": 1.0228457831380475e-06,
+      "loss": 0.6478,
+      "step": 17029
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.086166692004341,
+      "learning_rate": 1.0227492003352814e-06,
+      "loss": 0.6387,
+      "step": 17030
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.024867803214779,
+      "learning_rate": 1.022652617320195e-06,
+      "loss": 0.6341,
+      "step": 17031
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.3816875469333163,
+      "learning_rate": 1.0225560340936894e-06,
+      "loss": 0.6224,
+      "step": 17032
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.855919086459169,
+      "learning_rate": 1.0224594506566666e-06,
+      "loss": 0.7005,
+      "step": 17033
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.594411771620379,
+      "learning_rate": 1.0223628670100276e-06,
+      "loss": 0.625,
+      "step": 17034
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.4281876723783897,
+      "learning_rate": 1.022266283154674e-06,
+      "loss": 0.6712,
+      "step": 17035
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.719206366561483,
+      "learning_rate": 1.0221696990915073e-06,
+      "loss": 0.6478,
+      "step": 17036
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1591180227936553,
+      "learning_rate": 1.022073114821429e-06,
+      "loss": 0.6133,
+      "step": 17037
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.6828038557292015,
+      "learning_rate": 1.0219765303453401e-06,
+      "loss": 0.64,
+      "step": 17038
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.558791916104818,
+      "learning_rate": 1.0218799456641423e-06,
+      "loss": 0.6413,
+      "step": 17039
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.020345921468737,
+      "learning_rate": 1.0217833607787372e-06,
+      "loss": 0.6517,
+      "step": 17040
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.9682791297367346,
+      "learning_rate": 1.0216867756900257e-06,
+      "loss": 0.6374,
+      "step": 17041
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 6.177494134841801,
+      "learning_rate": 1.0215901903989103e-06,
+      "loss": 0.7253,
+      "step": 17042
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.2275426237907516,
+      "learning_rate": 1.0214936049062911e-06,
+      "loss": 0.6536,
+      "step": 17043
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.0608746315349475,
+      "learning_rate": 1.0213970192130705e-06,
+      "loss": 0.651,
+      "step": 17044
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.5266758579117394,
+      "learning_rate": 1.0213004333201493e-06,
+      "loss": 0.6582,
+      "step": 17045
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.7899543627943686,
+      "learning_rate": 1.0212038472284297e-06,
+      "loss": 0.6081,
+      "step": 17046
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.9630369287809293,
+      "learning_rate": 1.0211072609388125e-06,
+      "loss": 0.6335,
+      "step": 17047
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.020237390380146,
+      "learning_rate": 1.0210106744521994e-06,
+      "loss": 0.6322,
+      "step": 17048
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.7855164839397735,
+      "learning_rate": 1.0209140877694917e-06,
+      "loss": 0.6784,
+      "step": 17049
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.783287310962729,
+      "learning_rate": 1.0208175008915913e-06,
+      "loss": 0.6243,
+      "step": 17050
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.925294986147642,
+      "learning_rate": 1.020720913819399e-06,
+      "loss": 0.6484,
+      "step": 17051
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1182763698755753,
+      "learning_rate": 1.0206243265538166e-06,
+      "loss": 0.6387,
+      "step": 17052
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.886102917035699,
+      "learning_rate": 1.0205277390957458e-06,
+      "loss": 0.6582,
+      "step": 17053
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.254679222971205,
+      "learning_rate": 1.0204311514460874e-06,
+      "loss": 0.5931,
+      "step": 17054
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.2807271256847574,
+      "learning_rate": 1.0203345636057437e-06,
+      "loss": 0.6693,
+      "step": 17055
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 5.688865076968785,
+      "learning_rate": 1.0202379755756154e-06,
+      "loss": 0.6413,
+      "step": 17056
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.0893955371717334,
+      "learning_rate": 1.0201413873566043e-06,
+      "loss": 0.6315,
+      "step": 17057
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.9730078280418417,
+      "learning_rate": 1.0200447989496119e-06,
+      "loss": 0.6263,
+      "step": 17058
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1437465679856955,
+      "learning_rate": 1.0199482103555397e-06,
+      "loss": 0.5876,
+      "step": 17059
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 4.780038368078662,
+      "learning_rate": 1.0198516215752892e-06,
+      "loss": 0.6562,
+      "step": 17060
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.2497800804935633,
+      "learning_rate": 1.0197550326097615e-06,
+      "loss": 0.6335,
+      "step": 17061
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1320814012557627,
+      "learning_rate": 1.0196584434598583e-06,
+      "loss": 0.6589,
+      "step": 17062
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.802112431985338,
+      "learning_rate": 1.0195618541264813e-06,
+      "loss": 0.6217,
+      "step": 17063
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.1070905249083682,
+      "learning_rate": 1.0194652646105317e-06,
+      "loss": 0.6582,
+      "step": 17064
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.1296714809325636,
+      "learning_rate": 1.0193686749129112e-06,
+      "loss": 0.6302,
+      "step": 17065
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.9996738421921854,
+      "learning_rate": 1.019272085034521e-06,
+      "loss": 0.6387,
+      "step": 17066
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.9898155852426513,
+      "learning_rate": 1.0191754949762628e-06,
+      "loss": 0.6009,
+      "step": 17067
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.0816392106357884,
+      "learning_rate": 1.0190789047390378e-06,
+      "loss": 0.627,
+      "step": 17068
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.7084036782130827,
+      "learning_rate": 1.0189823143237482e-06,
+      "loss": 0.6569,
+      "step": 17069
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.0792883763558336,
+      "learning_rate": 1.0188857237312942e-06,
+      "loss": 0.6087,
+      "step": 17070
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.467390438940322,
+      "learning_rate": 1.0187891329625787e-06,
+      "loss": 0.6172,
+      "step": 17071
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.899305187094525,
+      "learning_rate": 1.018692542018502e-06,
+      "loss": 0.6576,
+      "step": 17072
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.371011974984921,
+      "learning_rate": 1.018595950899967e-06,
+      "loss": 0.6139,
+      "step": 17073
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.1107781759268134,
+      "learning_rate": 1.0184993596078738e-06,
+      "loss": 0.5895,
+      "step": 17074
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.2068301992480412,
+      "learning_rate": 1.0184027681431244e-06,
+      "loss": 0.6497,
+      "step": 17075
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.141634929362269,
+      "learning_rate": 1.0183061765066208e-06,
+      "loss": 0.6732,
+      "step": 17076
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.603887621724599,
+      "learning_rate": 1.0182095846992633e-06,
+      "loss": 0.6497,
+      "step": 17077
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.9353561835905424,
+      "learning_rate": 1.0181129927219547e-06,
+      "loss": 0.6139,
+      "step": 17078
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.3604935001017,
+      "learning_rate": 1.0180164005755958e-06,
+      "loss": 0.6126,
+      "step": 17079
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.175051392669207,
+      "learning_rate": 1.017919808261088e-06,
+      "loss": 0.6211,
+      "step": 17080
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.2173737727810154,
+      "learning_rate": 1.0178232157793335e-06,
+      "loss": 0.6797,
+      "step": 17081
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.754660082145916,
+      "learning_rate": 1.0177266231312329e-06,
+      "loss": 0.6139,
+      "step": 17082
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.1224042192297503,
+      "learning_rate": 1.0176300303176883e-06,
+      "loss": 0.6725,
+      "step": 17083
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.7873014095850466,
+      "learning_rate": 1.0175334373396012e-06,
+      "loss": 0.653,
+      "step": 17084
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.122407482870709,
+      "learning_rate": 1.0174368441978728e-06,
+      "loss": 0.638,
+      "step": 17085
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.767909782418691,
+      "learning_rate": 1.0173402508934051e-06,
+      "loss": 0.6195,
+      "step": 17086
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.2940569140899867,
+      "learning_rate": 1.017243657427099e-06,
+      "loss": 0.6478,
+      "step": 17087
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.9927909796971313,
+      "learning_rate": 1.0171470637998563e-06,
+      "loss": 0.638,
+      "step": 17088
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.1398778642809173,
+      "learning_rate": 1.017050470012579e-06,
+      "loss": 0.6178,
+      "step": 17089
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.1225425373307685,
+      "learning_rate": 1.0169538760661676e-06,
+      "loss": 0.5944,
+      "step": 17090
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.5140477741205864,
+      "learning_rate": 1.0168572819615242e-06,
+      "loss": 0.5781,
+      "step": 17091
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.283149582569681,
+      "learning_rate": 1.0167606876995505e-06,
+      "loss": 0.6426,
+      "step": 17092
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.598922858602128,
+      "learning_rate": 1.0166640932811478e-06,
+      "loss": 0.6589,
+      "step": 17093
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.3776676578199556,
+      "learning_rate": 1.0165674987072179e-06,
+      "loss": 0.666,
+      "step": 17094
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.3953947866530334,
+      "learning_rate": 1.0164709039786616e-06,
+      "loss": 0.5964,
+      "step": 17095
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.2416145827616827,
+      "learning_rate": 1.0163743090963811e-06,
+      "loss": 0.6426,
+      "step": 17096
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.7680289680606105,
+      "learning_rate": 1.016277714061278e-06,
+      "loss": 0.6771,
+      "step": 17097
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.5092086211796,
+      "learning_rate": 1.016181118874253e-06,
+      "loss": 0.6517,
+      "step": 17098
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.2711046803894934,
+      "learning_rate": 1.0160845235362085e-06,
+      "loss": 0.6452,
+      "step": 17099
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.137505240909917,
+      "learning_rate": 1.015987928048046e-06,
+      "loss": 0.6673,
+      "step": 17100
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.8804641175940526,
+      "learning_rate": 1.0158913324106661e-06,
+      "loss": 0.5833,
+      "step": 17101
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.3408417665400165,
+      "learning_rate": 1.0157947366249714e-06,
+      "loss": 0.6335,
+      "step": 17102
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.4912996736860222,
+      "learning_rate": 1.0156981406918632e-06,
+      "loss": 0.6908,
+      "step": 17103
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.6266118497427122,
+      "learning_rate": 1.0156015446122424e-06,
+      "loss": 0.653,
+      "step": 17104
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.769777758366648,
+      "learning_rate": 1.0155049483870113e-06,
+      "loss": 0.6777,
+      "step": 17105
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 5.181082323911349,
+      "learning_rate": 1.0154083520170712e-06,
+      "loss": 0.6888,
+      "step": 17106
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.030459228317862,
+      "learning_rate": 1.0153117555033233e-06,
+      "loss": 0.6322,
+      "step": 17107
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.240220701185133,
+      "learning_rate": 1.0152151588466695e-06,
+      "loss": 0.61,
+      "step": 17108
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.088880940772439,
+      "learning_rate": 1.0151185620480115e-06,
+      "loss": 0.6712,
+      "step": 17109
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.094656947978499,
+      "learning_rate": 1.0150219651082506e-06,
+      "loss": 0.6758,
+      "step": 17110
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.7458714825625834,
+      "learning_rate": 1.014925368028288e-06,
+      "loss": 0.7168,
+      "step": 17111
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.873536999147027,
+      "learning_rate": 1.0148287708090261e-06,
+      "loss": 0.6198,
+      "step": 17112
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.833417382091207,
+      "learning_rate": 1.0147321734513658e-06,
+      "loss": 0.6003,
+      "step": 17113
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.811097474094833,
+      "learning_rate": 1.0146355759562087e-06,
+      "loss": 0.6732,
+      "step": 17114
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.2153218903694953,
+      "learning_rate": 1.0145389783244567e-06,
+      "loss": 0.7018,
+      "step": 17115
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.5279761527808695,
+      "learning_rate": 1.014442380557011e-06,
+      "loss": 0.6725,
+      "step": 17116
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.6405443223053187,
+      "learning_rate": 1.014345782654773e-06,
+      "loss": 0.679,
+      "step": 17117
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.1878369265264945,
+      "learning_rate": 1.014249184618645e-06,
+      "loss": 0.6478,
+      "step": 17118
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.6849572313238697,
+      "learning_rate": 1.014152586449528e-06,
+      "loss": 0.6432,
+      "step": 17119
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 8.875611032533346,
+      "learning_rate": 1.0140559881483235e-06,
+      "loss": 0.6582,
+      "step": 17120
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.532050732774491,
+      "learning_rate": 1.0139593897159333e-06,
+      "loss": 0.6097,
+      "step": 17121
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.425873976024409,
+      "learning_rate": 1.013862791153259e-06,
+      "loss": 0.6628,
+      "step": 17122
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.1053470228056295,
+      "learning_rate": 1.0137661924612017e-06,
+      "loss": 0.6758,
+      "step": 17123
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.5904796542305424,
+      "learning_rate": 1.0136695936406636e-06,
+      "loss": 0.6081,
+      "step": 17124
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.964485795815236,
+      "learning_rate": 1.0135729946925458e-06,
+      "loss": 0.6797,
+      "step": 17125
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.388192683331944,
+      "learning_rate": 1.0134763956177504e-06,
+      "loss": 0.6296,
+      "step": 17126
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.703472685347129,
+      "learning_rate": 1.0133797964171782e-06,
+      "loss": 0.6413,
+      "step": 17127
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.3535575936438637,
+      "learning_rate": 1.013283197091731e-06,
+      "loss": 0.6081,
+      "step": 17128
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.5620913083067887,
+      "learning_rate": 1.0131865976423113e-06,
+      "loss": 0.6263,
+      "step": 17129
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.9626912070472526,
+      "learning_rate": 1.0130899980698193e-06,
+      "loss": 0.6243,
+      "step": 17130
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 4.309238761533303,
+      "learning_rate": 1.0129933983751575e-06,
+      "loss": 0.6562,
+      "step": 17131
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.4009842473399767,
+      "learning_rate": 1.0128967985592271e-06,
+      "loss": 0.6139,
+      "step": 17132
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.2120952417503785,
+      "learning_rate": 1.0128001986229296e-06,
+      "loss": 0.6543,
+      "step": 17133
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.691611420198526,
+      "learning_rate": 1.0127035985671669e-06,
+      "loss": 0.6543,
+      "step": 17134
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.976085675559478,
+      "learning_rate": 1.0126069983928401e-06,
+      "loss": 0.6283,
+      "step": 17135
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.829652063846915,
+      "learning_rate": 1.0125103981008514e-06,
+      "loss": 0.6354,
+      "step": 17136
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.971975164035998,
+      "learning_rate": 1.012413797692102e-06,
+      "loss": 0.6738,
+      "step": 17137
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.7459641470557603,
+      "learning_rate": 1.012317197167493e-06,
+      "loss": 0.6348,
+      "step": 17138
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.75624871253105,
+      "learning_rate": 1.0122205965279272e-06,
+      "loss": 0.651,
+      "step": 17139
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.8347652143842104,
+      "learning_rate": 1.0121239957743053e-06,
+      "loss": 0.612,
+      "step": 17140
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.37632466374376,
+      "learning_rate": 1.012027394907529e-06,
+      "loss": 0.6191,
+      "step": 17141
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.2496724374097505,
+      "learning_rate": 1.0119307939285e-06,
+      "loss": 0.6849,
+      "step": 17142
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.109133133568469,
+      "learning_rate": 1.0118341928381198e-06,
+      "loss": 0.6732,
+      "step": 17143
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.9995532734988353,
+      "learning_rate": 1.01173759163729e-06,
+      "loss": 0.5944,
+      "step": 17144
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.2505056753308956,
+      "learning_rate": 1.0116409903269126e-06,
+      "loss": 0.651,
+      "step": 17145
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.823945251834524,
+      "learning_rate": 1.011544388907888e-06,
+      "loss": 0.5915,
+      "step": 17146
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.390129084027486,
+      "learning_rate": 1.0114477873811193e-06,
+      "loss": 0.6348,
+      "step": 17147
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.154175525169476,
+      "learning_rate": 1.011351185747507e-06,
+      "loss": 0.6361,
+      "step": 17148
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.814960769335168,
+      "learning_rate": 1.0112545840079534e-06,
+      "loss": 0.6172,
+      "step": 17149
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.4168870826076345,
+      "learning_rate": 1.0111579821633596e-06,
+      "loss": 0.6198,
+      "step": 17150
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.944835799960231,
+      "learning_rate": 1.0110613802146274e-06,
+      "loss": 0.6497,
+      "step": 17151
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5055727900749005,
+      "learning_rate": 1.0109647781626581e-06,
+      "loss": 0.6325,
+      "step": 17152
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.7863529944401684,
+      "learning_rate": 1.010868176008354e-06,
+      "loss": 0.5804,
+      "step": 17153
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.1332042529974737,
+      "learning_rate": 1.010771573752616e-06,
+      "loss": 0.6146,
+      "step": 17154
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.163917659376779,
+      "learning_rate": 1.0106749713963463e-06,
+      "loss": 0.6035,
+      "step": 17155
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5226096533688493,
+      "learning_rate": 1.0105783689404454e-06,
+      "loss": 0.6178,
+      "step": 17156
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.3344054650924715,
+      "learning_rate": 1.0104817663858161e-06,
+      "loss": 0.6016,
+      "step": 17157
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.829197716370603,
+      "learning_rate": 1.0103851637333596e-06,
+      "loss": 0.6172,
+      "step": 17158
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.430394639073579,
+      "learning_rate": 1.0102885609839775e-06,
+      "loss": 0.6159,
+      "step": 17159
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.009496565270386,
+      "learning_rate": 1.0101919581385713e-06,
+      "loss": 0.6243,
+      "step": 17160
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.287616219171595,
+      "learning_rate": 1.0100953551980425e-06,
+      "loss": 0.6224,
+      "step": 17161
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.8042515711603966,
+      "learning_rate": 1.009998752163293e-06,
+      "loss": 0.6322,
+      "step": 17162
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.122849195037774,
+      "learning_rate": 1.0099021490352242e-06,
+      "loss": 0.6009,
+      "step": 17163
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.7647931414111255,
+      "learning_rate": 1.0098055458147376e-06,
+      "loss": 0.6562,
+      "step": 17164
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.7935937354722036,
+      "learning_rate": 1.009708942502735e-06,
+      "loss": 0.6523,
+      "step": 17165
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.7066616248324973,
+      "learning_rate": 1.0096123391001185e-06,
+      "loss": 0.6582,
+      "step": 17166
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.004589714788471,
+      "learning_rate": 1.0095157356077885e-06,
+      "loss": 0.6471,
+      "step": 17167
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.796904079372568,
+      "learning_rate": 1.0094191320266478e-06,
+      "loss": 0.6257,
+      "step": 17168
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.9507199345217736,
+      "learning_rate": 1.0093225283575974e-06,
+      "loss": 0.6582,
+      "step": 17169
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.90293776934589,
+      "learning_rate": 1.009225924601539e-06,
+      "loss": 0.6263,
+      "step": 17170
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.896620028195126,
+      "learning_rate": 1.009129320759374e-06,
+      "loss": 0.6328,
+      "step": 17171
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.015511117306315,
+      "learning_rate": 1.0090327168320046e-06,
+      "loss": 0.6042,
+      "step": 17172
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5616289279565665,
+      "learning_rate": 1.0089361128203318e-06,
+      "loss": 0.6478,
+      "step": 17173
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.163651832519545,
+      "learning_rate": 1.0088395087252575e-06,
+      "loss": 0.6055,
+      "step": 17174
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.709578549889298,
+      "learning_rate": 1.0087429045476834e-06,
+      "loss": 0.6211,
+      "step": 17175
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5589182602632965,
+      "learning_rate": 1.008646300288511e-06,
+      "loss": 0.6387,
+      "step": 17176
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.882221088607357,
+      "learning_rate": 1.0085496959486417e-06,
+      "loss": 0.6706,
+      "step": 17177
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.249662462719627,
+      "learning_rate": 1.0084530915289775e-06,
+      "loss": 0.6322,
+      "step": 17178
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.346101960198177,
+      "learning_rate": 1.0083564870304199e-06,
+      "loss": 0.7311,
+      "step": 17179
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.0853205613070296,
+      "learning_rate": 1.0082598824538702e-06,
+      "loss": 0.6139,
+      "step": 17180
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.266003475039164,
+      "learning_rate": 1.0081632778002304e-06,
+      "loss": 0.6172,
+      "step": 17181
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.3540834209146504,
+      "learning_rate": 1.0080666730704025e-06,
+      "loss": 0.6719,
+      "step": 17182
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.62577084703678,
+      "learning_rate": 1.0079700682652869e-06,
+      "loss": 0.6608,
+      "step": 17183
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 5.283132057559506,
+      "learning_rate": 1.0078734633857863e-06,
+      "loss": 0.61,
+      "step": 17184
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5829779179400454,
+      "learning_rate": 1.0077768584328019e-06,
+      "loss": 0.6348,
+      "step": 17185
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.023599968709655,
+      "learning_rate": 1.0076802534072352e-06,
+      "loss": 0.6302,
+      "step": 17186
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.8130146564594454,
+      "learning_rate": 1.0075836483099885e-06,
+      "loss": 0.6602,
+      "step": 17187
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.7567467256964466,
+      "learning_rate": 1.0074870431419627e-06,
+      "loss": 0.5739,
+      "step": 17188
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.6457976450610103,
+      "learning_rate": 1.0073904379040595e-06,
+      "loss": 0.6029,
+      "step": 17189
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.6535079544982123,
+      "learning_rate": 1.0072938325971808e-06,
+      "loss": 0.6289,
+      "step": 17190
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.4831086105183915,
+      "learning_rate": 1.007197227222228e-06,
+      "loss": 0.6549,
+      "step": 17191
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.9577123226867883,
+      "learning_rate": 1.0071006217801033e-06,
+      "loss": 0.6706,
+      "step": 17192
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.632685275547029,
+      "learning_rate": 1.0070040162717074e-06,
+      "loss": 0.6133,
+      "step": 17193
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.5346070150997457,
+      "learning_rate": 1.0069074106979425e-06,
+      "loss": 0.6589,
+      "step": 17194
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 7.056909915235414,
+      "learning_rate": 1.0068108050597105e-06,
+      "loss": 0.6162,
+      "step": 17195
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.0620933631216474,
+      "learning_rate": 1.006714199357912e-06,
+      "loss": 0.6289,
+      "step": 17196
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 3.3716040564234273,
+      "learning_rate": 1.0066175935934497e-06,
+      "loss": 0.6426,
+      "step": 17197
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 4.393890924307363,
+      "learning_rate": 1.0065209877672249e-06,
+      "loss": 0.6855,
+      "step": 17198
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.197635015351404,
+      "learning_rate": 1.0064243818801389e-06,
+      "loss": 0.6784,
+      "step": 17199
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.8157857101035355,
+      "learning_rate": 1.0063277759330937e-06,
+      "loss": 0.6556,
+      "step": 17200
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.462105436389611,
+      "learning_rate": 1.0062311699269906e-06,
+      "loss": 0.6432,
+      "step": 17201
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 7.00282053812332,
+      "learning_rate": 1.0061345638627318e-06,
+      "loss": 0.6868,
+      "step": 17202
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.80960231338598,
+      "learning_rate": 1.0060379577412181e-06,
+      "loss": 0.623,
+      "step": 17203
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.9419895094848383,
+      "learning_rate": 1.0059413515633518e-06,
+      "loss": 0.6146,
+      "step": 17204
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.8914758200966233,
+      "learning_rate": 1.0058447453300347e-06,
+      "loss": 0.5859,
+      "step": 17205
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.104985956869868,
+      "learning_rate": 1.0057481390421676e-06,
+      "loss": 0.6504,
+      "step": 17206
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.3270699307001745,
+      "learning_rate": 1.0056515327006526e-06,
+      "loss": 0.6797,
+      "step": 17207
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.6580069371638704,
+      "learning_rate": 1.005554926306392e-06,
+      "loss": 0.6602,
+      "step": 17208
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.051408161100162,
+      "learning_rate": 1.005458319860286e-06,
+      "loss": 0.6198,
+      "step": 17209
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.022065701854874,
+      "learning_rate": 1.0053617133632374e-06,
+      "loss": 0.679,
+      "step": 17210
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.6853661476418025,
+      "learning_rate": 1.0052651068161476e-06,
+      "loss": 0.6309,
+      "step": 17211
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.843517873088034,
+      "learning_rate": 1.0051685002199177e-06,
+      "loss": 0.6523,
+      "step": 17212
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.912744327178719,
+      "learning_rate": 1.0050718935754502e-06,
+      "loss": 0.6549,
+      "step": 17213
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.8250334549595024,
+      "learning_rate": 1.0049752868836456e-06,
+      "loss": 0.5885,
+      "step": 17214
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.79143804738883,
+      "learning_rate": 1.0048786801454068e-06,
+      "loss": 0.6823,
+      "step": 17215
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.833485890494847,
+      "learning_rate": 1.0047820733616346e-06,
+      "loss": 0.6217,
+      "step": 17216
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.8429670256201045,
+      "learning_rate": 1.0046854665332306e-06,
+      "loss": 0.6097,
+      "step": 17217
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.187661787058169,
+      "learning_rate": 1.0045888596610974e-06,
+      "loss": 0.5801,
+      "step": 17218
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.346193918754595,
+      "learning_rate": 1.0044922527461358e-06,
+      "loss": 0.7201,
+      "step": 17219
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.710968718404482,
+      "learning_rate": 1.0043956457892471e-06,
+      "loss": 0.623,
+      "step": 17220
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.9005235639857356,
+      "learning_rate": 1.0042990387913341e-06,
+      "loss": 0.651,
+      "step": 17221
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.779536315364121,
+      "learning_rate": 1.0042024317532974e-06,
+      "loss": 0.6283,
+      "step": 17222
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.452272706654731,
+      "learning_rate": 1.0041058246760388e-06,
+      "loss": 0.6426,
+      "step": 17223
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.426295710859241,
+      "learning_rate": 1.0040092175604607e-06,
+      "loss": 0.6784,
+      "step": 17224
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.828666497327293,
+      "learning_rate": 1.003912610407464e-06,
+      "loss": 0.6198,
+      "step": 17225
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.3921448435749286,
+      "learning_rate": 1.0038160032179506e-06,
+      "loss": 0.6042,
+      "step": 17226
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.965572957642672,
+      "learning_rate": 1.003719395992822e-06,
+      "loss": 0.5703,
+      "step": 17227
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.838864022203711,
+      "learning_rate": 1.00362278873298e-06,
+      "loss": 0.6406,
+      "step": 17228
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.9831270615692285,
+      "learning_rate": 1.0035261814393263e-06,
+      "loss": 0.6536,
+      "step": 17229
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.067226619250042,
+      "learning_rate": 1.003429574112762e-06,
+      "loss": 0.681,
+      "step": 17230
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.255524516415049,
+      "learning_rate": 1.0033329667541896e-06,
+      "loss": 0.6484,
+      "step": 17231
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.537202986015353,
+      "learning_rate": 1.0032363593645105e-06,
+      "loss": 0.6315,
+      "step": 17232
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.690925357096082,
+      "learning_rate": 1.0031397519446258e-06,
+      "loss": 0.6289,
+      "step": 17233
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.9321483059685747,
+      "learning_rate": 1.0030431444954376e-06,
+      "loss": 0.6009,
+      "step": 17234
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.675676040395417,
+      "learning_rate": 1.0029465370178475e-06,
+      "loss": 0.6133,
+      "step": 17235
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.0695168045877272,
+      "learning_rate": 1.002849929512757e-06,
+      "loss": 0.6198,
+      "step": 17236
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.204261142618547,
+      "learning_rate": 1.0027533219810682e-06,
+      "loss": 0.6348,
+      "step": 17237
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.173765110309366,
+      "learning_rate": 1.0026567144236821e-06,
+      "loss": 0.6732,
+      "step": 17238
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.352899532278141,
+      "learning_rate": 1.0025601068415006e-06,
+      "loss": 0.6647,
+      "step": 17239
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.5467500477963263,
+      "learning_rate": 1.0024634992354256e-06,
+      "loss": 0.6354,
+      "step": 17240
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.1473697826401845,
+      "learning_rate": 1.0023668916063585e-06,
+      "loss": 0.6582,
+      "step": 17241
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.614412872483421,
+      "learning_rate": 1.0022702839552012e-06,
+      "loss": 0.6061,
+      "step": 17242
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 5.746336158732258,
+      "learning_rate": 1.0021736762828545e-06,
+      "loss": 0.6003,
+      "step": 17243
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.536387857070599,
+      "learning_rate": 1.0020770685902212e-06,
+      "loss": 0.6283,
+      "step": 17244
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.4341644112280436,
+      "learning_rate": 1.0019804608782022e-06,
+      "loss": 0.6393,
+      "step": 17245
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.1254357194537774,
+      "learning_rate": 1.0018838531476995e-06,
+      "loss": 0.6497,
+      "step": 17246
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.7048415598324245,
+      "learning_rate": 1.0017872453996148e-06,
+      "loss": 0.6387,
+      "step": 17247
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.072894228444092,
+      "learning_rate": 1.0016906376348496e-06,
+      "loss": 0.6165,
+      "step": 17248
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.9491256655283933,
+      "learning_rate": 1.001594029854305e-06,
+      "loss": 0.6419,
+      "step": 17249
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.083475575094535,
+      "learning_rate": 1.0014974220588836e-06,
+      "loss": 0.6048,
+      "step": 17250
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.7208805066418638,
+      "learning_rate": 1.0014008142494866e-06,
+      "loss": 0.6764,
+      "step": 17251
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.32665185856043,
+      "learning_rate": 1.0013042064270154e-06,
+      "loss": 0.6784,
+      "step": 17252
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.7078796649841514,
+      "learning_rate": 1.0012075985923721e-06,
+      "loss": 0.6517,
+      "step": 17253
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.6550700423285893,
+      "learning_rate": 1.0011109907464585e-06,
+      "loss": 0.6908,
+      "step": 17254
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.542784337608103,
+      "learning_rate": 1.0010143828901756e-06,
+      "loss": 0.6178,
+      "step": 17255
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.201422965846114,
+      "learning_rate": 1.0009177750244253e-06,
+      "loss": 0.6406,
+      "step": 17256
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.1448644087565683,
+      "learning_rate": 1.0008211671501095e-06,
+      "loss": 0.6497,
+      "step": 17257
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.137764692585806,
+      "learning_rate": 1.0007245592681298e-06,
+      "loss": 0.6842,
+      "step": 17258
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.271255251513789,
+      "learning_rate": 1.0006279513793874e-06,
+      "loss": 0.6296,
+      "step": 17259
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.9589464499138773,
+      "learning_rate": 1.0005313434847841e-06,
+      "loss": 0.5918,
+      "step": 17260
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.220987686230116,
+      "learning_rate": 1.0004347355852226e-06,
+      "loss": 0.6986,
+      "step": 17261
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.655961241236927,
+      "learning_rate": 1.0003381276816027e-06,
+      "loss": 0.6836,
+      "step": 17262
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 3.364984091016474,
+      "learning_rate": 1.0002415197748277e-06,
+      "loss": 0.6224,
+      "step": 17263
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 4.5536218342661465,
+      "learning_rate": 1.0001449118657983e-06,
+      "loss": 0.6396,
+      "step": 17264
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.8495832364689377,
+      "learning_rate": 1.0000483039554165e-06,
+      "loss": 0.6654,
+      "step": 17265
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.395417155554365,
+      "learning_rate": 9.999516960445839e-07,
+      "loss": 0.6296,
+      "step": 17266
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.140224175040271,
+      "learning_rate": 9.998550881342019e-07,
+      "loss": 0.6247,
+      "step": 17267
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.844456804950263,
+      "learning_rate": 9.997584802251724e-07,
+      "loss": 0.651,
+      "step": 17268
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.870427840025103,
+      "learning_rate": 9.996618723183973e-07,
+      "loss": 0.6139,
+      "step": 17269
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.9864718207819205,
+      "learning_rate": 9.995652644147776e-07,
+      "loss": 0.6771,
+      "step": 17270
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.5457313381252717,
+      "learning_rate": 9.994686565152156e-07,
+      "loss": 0.6393,
+      "step": 17271
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.315858049502274,
+      "learning_rate": 9.993720486206125e-07,
+      "loss": 0.6654,
+      "step": 17272
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.760908639459377,
+      "learning_rate": 9.992754407318705e-07,
+      "loss": 0.6439,
+      "step": 17273
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.747000294950304,
+      "learning_rate": 9.991788328498906e-07,
+      "loss": 0.6582,
+      "step": 17274
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.259178230187191,
+      "learning_rate": 9.990822249755746e-07,
+      "loss": 0.651,
+      "step": 17275
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.6637273867415394,
+      "learning_rate": 9.989856171098245e-07,
+      "loss": 0.6471,
+      "step": 17276
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.3500869254066705,
+      "learning_rate": 9.988890092535417e-07,
+      "loss": 0.584,
+      "step": 17277
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.2506307532713894,
+      "learning_rate": 9.987924014076276e-07,
+      "loss": 0.6667,
+      "step": 17278
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.584675672693785,
+      "learning_rate": 9.986957935729845e-07,
+      "loss": 0.5911,
+      "step": 17279
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.378924222098828,
+      "learning_rate": 9.985991857505135e-07,
+      "loss": 0.5931,
+      "step": 17280
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.335773321660843,
+      "learning_rate": 9.985025779411165e-07,
+      "loss": 0.6836,
+      "step": 17281
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.8232171815730447,
+      "learning_rate": 9.98405970145695e-07,
+      "loss": 0.6348,
+      "step": 17282
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.837921491112488,
+      "learning_rate": 9.983093623651505e-07,
+      "loss": 0.6484,
+      "step": 17283
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.2610787962928556,
+      "learning_rate": 9.982127546003851e-07,
+      "loss": 0.6087,
+      "step": 17284
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.32294888534835,
+      "learning_rate": 9.981161468523004e-07,
+      "loss": 0.627,
+      "step": 17285
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.2079006014972693,
+      "learning_rate": 9.980195391217979e-07,
+      "loss": 0.6048,
+      "step": 17286
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.007766056423373,
+      "learning_rate": 9.97922931409779e-07,
+      "loss": 0.6445,
+      "step": 17287
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.3915659409035275,
+      "learning_rate": 9.978263237171454e-07,
+      "loss": 0.6387,
+      "step": 17288
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.0952651244194054,
+      "learning_rate": 9.97729716044799e-07,
+      "loss": 0.6094,
+      "step": 17289
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.452417464934383,
+      "learning_rate": 9.976331083936414e-07,
+      "loss": 0.6221,
+      "step": 17290
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.4781533734600947,
+      "learning_rate": 9.975365007645741e-07,
+      "loss": 0.6204,
+      "step": 17291
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.338990185146652,
+      "learning_rate": 9.974398931584995e-07,
+      "loss": 0.6465,
+      "step": 17292
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.9627362629400054,
+      "learning_rate": 9.97343285576318e-07,
+      "loss": 0.5755,
+      "step": 17293
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.585920512255009,
+      "learning_rate": 9.97246678018932e-07,
+      "loss": 0.6191,
+      "step": 17294
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.365940267845096,
+      "learning_rate": 9.971500704872429e-07,
+      "loss": 0.6458,
+      "step": 17295
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.74425879005676,
+      "learning_rate": 9.970534629821524e-07,
+      "loss": 0.6536,
+      "step": 17296
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.4117056685613085,
+      "learning_rate": 9.969568555045625e-07,
+      "loss": 0.6745,
+      "step": 17297
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.420030241377768,
+      "learning_rate": 9.96860248055374e-07,
+      "loss": 0.6784,
+      "step": 17298
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.366403874485703,
+      "learning_rate": 9.967636406354898e-07,
+      "loss": 0.6615,
+      "step": 17299
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.983330414568246,
+      "learning_rate": 9.966670332458103e-07,
+      "loss": 0.6439,
+      "step": 17300
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.5102768159091267,
+      "learning_rate": 9.965704258872379e-07,
+      "loss": 0.6569,
+      "step": 17301
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.417103224663362,
+      "learning_rate": 9.964738185606738e-07,
+      "loss": 0.6686,
+      "step": 17302
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.759486841281788,
+      "learning_rate": 9.9637721126702e-07,
+      "loss": 0.6253,
+      "step": 17303
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.179723891537662,
+      "learning_rate": 9.962806040071778e-07,
+      "loss": 0.638,
+      "step": 17304
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.526938078441234,
+      "learning_rate": 9.961839967820495e-07,
+      "loss": 0.6348,
+      "step": 17305
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.7194891245516004,
+      "learning_rate": 9.960873895925362e-07,
+      "loss": 0.6393,
+      "step": 17306
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.7155129351732783,
+      "learning_rate": 9.959907824395395e-07,
+      "loss": 0.6712,
+      "step": 17307
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.0927739629685895,
+      "learning_rate": 9.95894175323961e-07,
+      "loss": 0.6719,
+      "step": 17308
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.954406680973517,
+      "learning_rate": 9.957975682467027e-07,
+      "loss": 0.651,
+      "step": 17309
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.7295363347958213,
+      "learning_rate": 9.95700961208666e-07,
+      "loss": 0.6471,
+      "step": 17310
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.9665502120271543,
+      "learning_rate": 9.956043542107528e-07,
+      "loss": 0.6302,
+      "step": 17311
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.233318094015883,
+      "learning_rate": 9.955077472538648e-07,
+      "loss": 0.6523,
+      "step": 17312
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.7268229232388,
+      "learning_rate": 9.954111403389027e-07,
+      "loss": 0.6491,
+      "step": 17313
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.03693072331686,
+      "learning_rate": 9.953145334667693e-07,
+      "loss": 0.627,
+      "step": 17314
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.5406821638668102,
+      "learning_rate": 9.952179266383654e-07,
+      "loss": 0.6191,
+      "step": 17315
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.951799943041009,
+      "learning_rate": 9.951213198545933e-07,
+      "loss": 0.627,
+      "step": 17316
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.9179141094673873,
+      "learning_rate": 9.950247131163541e-07,
+      "loss": 0.5918,
+      "step": 17317
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.0826115388545587,
+      "learning_rate": 9.9492810642455e-07,
+      "loss": 0.6556,
+      "step": 17318
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.485731187401998,
+      "learning_rate": 9.948314997800826e-07,
+      "loss": 0.5996,
+      "step": 17319
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.778754452548987,
+      "learning_rate": 9.947348931838526e-07,
+      "loss": 0.6647,
+      "step": 17320
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.364411036436592,
+      "learning_rate": 9.946382866367625e-07,
+      "loss": 0.625,
+      "step": 17321
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 6.261521444256794,
+      "learning_rate": 9.94541680139714e-07,
+      "loss": 0.6517,
+      "step": 17322
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.4314886122175485,
+      "learning_rate": 9.94445073693608e-07,
+      "loss": 0.5957,
+      "step": 17323
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.343701379580414,
+      "learning_rate": 9.94348467299347e-07,
+      "loss": 0.6445,
+      "step": 17324
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 5.995683694853375,
+      "learning_rate": 9.942518609578327e-07,
+      "loss": 0.6745,
+      "step": 17325
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.2198777924170026,
+      "learning_rate": 9.941552546699656e-07,
+      "loss": 0.6732,
+      "step": 17326
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 4.223291198436136,
+      "learning_rate": 9.940586484366483e-07,
+      "loss": 0.5716,
+      "step": 17327
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.0440251936781832,
+      "learning_rate": 9.939620422587818e-07,
+      "loss": 0.6458,
+      "step": 17328
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.7024838025837266,
+      "learning_rate": 9.938654361372686e-07,
+      "loss": 0.6374,
+      "step": 17329
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.80706855031585,
+      "learning_rate": 9.937688300730093e-07,
+      "loss": 0.6439,
+      "step": 17330
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.5823484919351065,
+      "learning_rate": 9.936722240669062e-07,
+      "loss": 0.6074,
+      "step": 17331
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.1673195698692624,
+      "learning_rate": 9.935756181198615e-07,
+      "loss": 0.6647,
+      "step": 17332
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.246309180670043,
+      "learning_rate": 9.934790122327754e-07,
+      "loss": 0.5996,
+      "step": 17333
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.7535052714546038,
+      "learning_rate": 9.933824064065502e-07,
+      "loss": 0.6419,
+      "step": 17334
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.711013765243277,
+      "learning_rate": 9.93285800642088e-07,
+      "loss": 0.6276,
+      "step": 17335
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.9539014243200934,
+      "learning_rate": 9.931891949402896e-07,
+      "loss": 0.6387,
+      "step": 17336
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.118526561224825,
+      "learning_rate": 9.930925893020574e-07,
+      "loss": 0.6204,
+      "step": 17337
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.321337173183871,
+      "learning_rate": 9.929959837282925e-07,
+      "loss": 0.6608,
+      "step": 17338
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.4952101767623285,
+      "learning_rate": 9.928993782198969e-07,
+      "loss": 0.6647,
+      "step": 17339
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.585883371721984,
+      "learning_rate": 9.928027727777719e-07,
+      "loss": 0.6276,
+      "step": 17340
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.3689784639617266,
+      "learning_rate": 9.927061674028193e-07,
+      "loss": 0.6185,
+      "step": 17341
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.400115534738745,
+      "learning_rate": 9.926095620959407e-07,
+      "loss": 0.6341,
+      "step": 17342
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.0874339191055786,
+      "learning_rate": 9.925129568580374e-07,
+      "loss": 0.5983,
+      "step": 17343
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.732669778319713,
+      "learning_rate": 9.924163516900114e-07,
+      "loss": 0.666,
+      "step": 17344
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.577974265245599,
+      "learning_rate": 9.923197465927647e-07,
+      "loss": 0.6549,
+      "step": 17345
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.761877609253576,
+      "learning_rate": 9.922231415671982e-07,
+      "loss": 0.6569,
+      "step": 17346
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.4179118524604135,
+      "learning_rate": 9.921265366142138e-07,
+      "loss": 0.6221,
+      "step": 17347
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.759159570510402,
+      "learning_rate": 9.920299317347132e-07,
+      "loss": 0.6276,
+      "step": 17348
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.186806934818881,
+      "learning_rate": 9.919333269295977e-07,
+      "loss": 0.6758,
+      "step": 17349
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.1093917294953513,
+      "learning_rate": 9.918367221997695e-07,
+      "loss": 0.627,
+      "step": 17350
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.910742793953512,
+      "learning_rate": 9.917401175461297e-07,
+      "loss": 0.6556,
+      "step": 17351
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.143119323645674,
+      "learning_rate": 9.916435129695804e-07,
+      "loss": 0.6113,
+      "step": 17352
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.167836027716115,
+      "learning_rate": 9.915469084710228e-07,
+      "loss": 0.6374,
+      "step": 17353
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.798266749275704,
+      "learning_rate": 9.914503040513582e-07,
+      "loss": 0.666,
+      "step": 17354
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.592213376995982,
+      "learning_rate": 9.91353699711489e-07,
+      "loss": 0.6758,
+      "step": 17355
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.4066304034986823,
+      "learning_rate": 9.912570954523167e-07,
+      "loss": 0.6029,
+      "step": 17356
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.139385727437995,
+      "learning_rate": 9.911604912747424e-07,
+      "loss": 0.6478,
+      "step": 17357
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.810242389199682,
+      "learning_rate": 9.91063887179668e-07,
+      "loss": 0.6654,
+      "step": 17358
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.4660876133125313,
+      "learning_rate": 9.909672831679958e-07,
+      "loss": 0.6719,
+      "step": 17359
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.7615917087470305,
+      "learning_rate": 9.90870679240626e-07,
+      "loss": 0.5911,
+      "step": 17360
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.368795537608442,
+      "learning_rate": 9.907740753984612e-07,
+      "loss": 0.6868,
+      "step": 17361
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.9014104207199742,
+      "learning_rate": 9.906774716424025e-07,
+      "loss": 0.6048,
+      "step": 17362
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.650978330377212,
+      "learning_rate": 9.90580867973352e-07,
+      "loss": 0.6309,
+      "step": 17363
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.535797971110275,
+      "learning_rate": 9.904842643922112e-07,
+      "loss": 0.681,
+      "step": 17364
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.18693768728951,
+      "learning_rate": 9.903876608998819e-07,
+      "loss": 0.6178,
+      "step": 17365
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.022140240803048,
+      "learning_rate": 9.902910574972648e-07,
+      "loss": 0.6048,
+      "step": 17366
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.6403833293012138,
+      "learning_rate": 9.901944541852625e-07,
+      "loss": 0.6556,
+      "step": 17367
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.202022565481712,
+      "learning_rate": 9.90097850964776e-07,
+      "loss": 0.6471,
+      "step": 17368
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.59650957277905,
+      "learning_rate": 9.900012478367072e-07,
+      "loss": 0.6706,
+      "step": 17369
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.060045523826881,
+      "learning_rate": 9.899046448019574e-07,
+      "loss": 0.627,
+      "step": 17370
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.6817634898819485,
+      "learning_rate": 9.898080418614288e-07,
+      "loss": 0.6087,
+      "step": 17371
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.362347622768725,
+      "learning_rate": 9.897114390160228e-07,
+      "loss": 0.681,
+      "step": 17372
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.1129461463767583,
+      "learning_rate": 9.896148362666405e-07,
+      "loss": 0.6328,
+      "step": 17373
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.67221745001429,
+      "learning_rate": 9.89518233614184e-07,
+      "loss": 0.6497,
+      "step": 17374
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.813164234011203,
+      "learning_rate": 9.894216310595545e-07,
+      "loss": 0.5918,
+      "step": 17375
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.322570891578026,
+      "learning_rate": 9.893250286036539e-07,
+      "loss": 0.6953,
+      "step": 17376
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.2598945262600205,
+      "learning_rate": 9.89228426247384e-07,
+      "loss": 0.6419,
+      "step": 17377
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.35444807093282,
+      "learning_rate": 9.891318239916464e-07,
+      "loss": 0.6602,
+      "step": 17378
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.251136133897727,
+      "learning_rate": 9.890352218373418e-07,
+      "loss": 0.6628,
+      "step": 17379
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.1331812715517056,
+      "learning_rate": 9.88938619785373e-07,
+      "loss": 0.6491,
+      "step": 17380
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.0241978522690993,
+      "learning_rate": 9.888420178366405e-07,
+      "loss": 0.5885,
+      "step": 17381
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.272807722330589,
+      "learning_rate": 9.887454159920467e-07,
+      "loss": 0.6178,
+      "step": 17382
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.0450502430397877,
+      "learning_rate": 9.886488142524929e-07,
+      "loss": 0.6419,
+      "step": 17383
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 4.726805545377388,
+      "learning_rate": 9.885522126188808e-07,
+      "loss": 0.6706,
+      "step": 17384
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.072424218523567,
+      "learning_rate": 9.88455611092112e-07,
+      "loss": 0.6289,
+      "step": 17385
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.060415261736962,
+      "learning_rate": 9.883590096730877e-07,
+      "loss": 0.6185,
+      "step": 17386
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.777056315671761,
+      "learning_rate": 9.882624083627098e-07,
+      "loss": 0.6452,
+      "step": 17387
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.719879621056617,
+      "learning_rate": 9.881658071618804e-07,
+      "loss": 0.6608,
+      "step": 17388
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.294683902619622,
+      "learning_rate": 9.880692060715e-07,
+      "loss": 0.6361,
+      "step": 17389
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.5055183276236836,
+      "learning_rate": 9.87972605092471e-07,
+      "loss": 0.6224,
+      "step": 17390
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.456624997204455,
+      "learning_rate": 9.878760042256946e-07,
+      "loss": 0.64,
+      "step": 17391
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.916627990938017,
+      "learning_rate": 9.877794034720729e-07,
+      "loss": 0.64,
+      "step": 17392
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.037138289209211,
+      "learning_rate": 9.87682802832507e-07,
+      "loss": 0.6471,
+      "step": 17393
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.036984189930057,
+      "learning_rate": 9.875862023078982e-07,
+      "loss": 0.612,
+      "step": 17394
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.906154543405576,
+      "learning_rate": 9.874896018991487e-07,
+      "loss": 0.6296,
+      "step": 17395
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 5.065978527139109,
+      "learning_rate": 9.873930016071598e-07,
+      "loss": 0.6803,
+      "step": 17396
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.3154755109171297,
+      "learning_rate": 9.87296401432833e-07,
+      "loss": 0.6784,
+      "step": 17397
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 3.8065582831005633,
+      "learning_rate": 9.871998013770708e-07,
+      "loss": 0.6426,
+      "step": 17398
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.9251757441934707,
+      "learning_rate": 9.871032014407732e-07,
+      "loss": 0.6621,
+      "step": 17399
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.2272642614863267,
+      "learning_rate": 9.870066016248426e-07,
+      "loss": 0.6107,
+      "step": 17400
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.069685598222167,
+      "learning_rate": 9.869100019301807e-07,
+      "loss": 0.6185,
+      "step": 17401
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.1133565095705613,
+      "learning_rate": 9.868134023576888e-07,
+      "loss": 0.6081,
+      "step": 17402
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.056725711198001,
+      "learning_rate": 9.867168029082685e-07,
+      "loss": 0.5872,
+      "step": 17403
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.319940714477502,
+      "learning_rate": 9.866202035828217e-07,
+      "loss": 0.6589,
+      "step": 17404
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.501406441013278,
+      "learning_rate": 9.8652360438225e-07,
+      "loss": 0.6813,
+      "step": 17405
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.646011405849257,
+      "learning_rate": 9.864270053074543e-07,
+      "loss": 0.6914,
+      "step": 17406
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.469524801048799,
+      "learning_rate": 9.863304063593363e-07,
+      "loss": 0.6159,
+      "step": 17407
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.2844510600710883,
+      "learning_rate": 9.862338075387984e-07,
+      "loss": 0.7077,
+      "step": 17408
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.1604759421577615,
+      "learning_rate": 9.861372088467412e-07,
+      "loss": 0.6243,
+      "step": 17409
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.9032547669454143,
+      "learning_rate": 9.860406102840666e-07,
+      "loss": 0.6309,
+      "step": 17410
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.747720990749747,
+      "learning_rate": 9.859440118516766e-07,
+      "loss": 0.6237,
+      "step": 17411
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.939242946878792,
+      "learning_rate": 9.858474135504722e-07,
+      "loss": 0.6914,
+      "step": 17412
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.1944593917429827,
+      "learning_rate": 9.857508153813551e-07,
+      "loss": 0.6504,
+      "step": 17413
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.421183358838909,
+      "learning_rate": 9.85654217345227e-07,
+      "loss": 0.6237,
+      "step": 17414
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.811727461132674,
+      "learning_rate": 9.855576194429892e-07,
+      "loss": 0.6243,
+      "step": 17415
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.930557149096399,
+      "learning_rate": 9.854610216755434e-07,
+      "loss": 0.571,
+      "step": 17416
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.336102320186337,
+      "learning_rate": 9.853644240437912e-07,
+      "loss": 0.6667,
+      "step": 17417
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.573679562152484,
+      "learning_rate": 9.852678265486343e-07,
+      "loss": 0.6211,
+      "step": 17418
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.992051105806668,
+      "learning_rate": 9.851712291909742e-07,
+      "loss": 0.6289,
+      "step": 17419
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.4459706893323694,
+      "learning_rate": 9.850746319717118e-07,
+      "loss": 0.6536,
+      "step": 17420
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.930301498059312,
+      "learning_rate": 9.849780348917493e-07,
+      "loss": 0.6725,
+      "step": 17421
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.1620630005428367,
+      "learning_rate": 9.848814379519884e-07,
+      "loss": 0.679,
+      "step": 17422
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.9183406048997633,
+      "learning_rate": 9.847848411533302e-07,
+      "loss": 0.6341,
+      "step": 17423
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.8230394021390253,
+      "learning_rate": 9.846882444966766e-07,
+      "loss": 0.6615,
+      "step": 17424
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.6413190488377736,
+      "learning_rate": 9.845916479829292e-07,
+      "loss": 0.6009,
+      "step": 17425
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.362864444556243,
+      "learning_rate": 9.844950516129886e-07,
+      "loss": 0.6042,
+      "step": 17426
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.9288565413567245,
+      "learning_rate": 9.843984553877578e-07,
+      "loss": 0.6445,
+      "step": 17427
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.905145185780454,
+      "learning_rate": 9.84301859308137e-07,
+      "loss": 0.6107,
+      "step": 17428
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.191674898384126,
+      "learning_rate": 9.842052633750285e-07,
+      "loss": 0.6003,
+      "step": 17429
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.496751673574792,
+      "learning_rate": 9.841086675893336e-07,
+      "loss": 0.6146,
+      "step": 17430
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.04797733868172,
+      "learning_rate": 9.840120719519544e-07,
+      "loss": 0.6406,
+      "step": 17431
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.161502511428071,
+      "learning_rate": 9.839154764637914e-07,
+      "loss": 0.6328,
+      "step": 17432
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.535112045212668,
+      "learning_rate": 9.83818881125747e-07,
+      "loss": 0.6341,
+      "step": 17433
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.18815066014578,
+      "learning_rate": 9.83722285938722e-07,
+      "loss": 0.6673,
+      "step": 17434
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.524736672063215,
+      "learning_rate": 9.836256909036188e-07,
+      "loss": 0.681,
+      "step": 17435
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.4155074443210705,
+      "learning_rate": 9.835290960213381e-07,
+      "loss": 0.6497,
+      "step": 17436
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.681510995882524,
+      "learning_rate": 9.834325012927823e-07,
+      "loss": 0.5833,
+      "step": 17437
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 5.3109167193099225,
+      "learning_rate": 9.833359067188523e-07,
+      "loss": 0.6432,
+      "step": 17438
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.4167227745592115,
+      "learning_rate": 9.832393123004494e-07,
+      "loss": 0.625,
+      "step": 17439
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.3446655856888716,
+      "learning_rate": 9.831427180384757e-07,
+      "loss": 0.6289,
+      "step": 17440
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.5335080999615154,
+      "learning_rate": 9.830461239338325e-07,
+      "loss": 0.6449,
+      "step": 17441
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.453962231765395,
+      "learning_rate": 9.829495299874212e-07,
+      "loss": 0.6335,
+      "step": 17442
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.795892134040347,
+      "learning_rate": 9.828529362001436e-07,
+      "loss": 0.6367,
+      "step": 17443
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.314786260033133,
+      "learning_rate": 9.827563425729013e-07,
+      "loss": 0.6406,
+      "step": 17444
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.4647075313544984,
+      "learning_rate": 9.826597491065952e-07,
+      "loss": 0.6113,
+      "step": 17445
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.8900313384525442,
+      "learning_rate": 9.825631558021273e-07,
+      "loss": 0.6341,
+      "step": 17446
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.974207542059112,
+      "learning_rate": 9.824665626603987e-07,
+      "loss": 0.6725,
+      "step": 17447
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.429354781923542,
+      "learning_rate": 9.823699696823116e-07,
+      "loss": 0.6628,
+      "step": 17448
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.412236019985001,
+      "learning_rate": 9.82273376868767e-07,
+      "loss": 0.6048,
+      "step": 17449
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 6.059029600999812,
+      "learning_rate": 9.821767842206666e-07,
+      "loss": 0.6341,
+      "step": 17450
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.408834575747848,
+      "learning_rate": 9.820801917389123e-07,
+      "loss": 0.7012,
+      "step": 17451
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.6188322413747254,
+      "learning_rate": 9.819835994244044e-07,
+      "loss": 0.6549,
+      "step": 17452
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.530730342935476,
+      "learning_rate": 9.818870072780454e-07,
+      "loss": 0.6562,
+      "step": 17453
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.148884704531288,
+      "learning_rate": 9.817904153007366e-07,
+      "loss": 0.6276,
+      "step": 17454
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.6615171548237018,
+      "learning_rate": 9.816938234933793e-07,
+      "loss": 0.6693,
+      "step": 17455
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.5499695795889363,
+      "learning_rate": 9.815972318568755e-07,
+      "loss": 0.6686,
+      "step": 17456
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.5903768106029994,
+      "learning_rate": 9.81500640392126e-07,
+      "loss": 0.6165,
+      "step": 17457
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.995821197228488,
+      "learning_rate": 9.814040491000331e-07,
+      "loss": 0.6302,
+      "step": 17458
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.375074766500156,
+      "learning_rate": 9.813074579814978e-07,
+      "loss": 0.6673,
+      "step": 17459
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.338942793661134,
+      "learning_rate": 9.812108670374212e-07,
+      "loss": 0.5788,
+      "step": 17460
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.0260062241929595,
+      "learning_rate": 9.811142762687057e-07,
+      "loss": 0.6094,
+      "step": 17461
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.136933443975528,
+      "learning_rate": 9.81017685676252e-07,
+      "loss": 0.6169,
+      "step": 17462
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.7580479396573527,
+      "learning_rate": 9.80921095260962e-07,
+      "loss": 0.6432,
+      "step": 17463
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.4768804185083213,
+      "learning_rate": 9.808245050237375e-07,
+      "loss": 0.6393,
+      "step": 17464
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 3.3115228169305424,
+      "learning_rate": 9.807279149654793e-07,
+      "loss": 0.6009,
+      "step": 17465
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.426736309278423,
+      "learning_rate": 9.80631325087089e-07,
+      "loss": 0.6667,
+      "step": 17466
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.8870452241539235,
+      "learning_rate": 9.805347353894684e-07,
+      "loss": 0.5872,
+      "step": 17467
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.8897924760772677,
+      "learning_rate": 9.804381458735186e-07,
+      "loss": 0.6042,
+      "step": 17468
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.046647819875917,
+      "learning_rate": 9.803415565401416e-07,
+      "loss": 0.6536,
+      "step": 17469
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.23249185623194,
+      "learning_rate": 9.802449673902384e-07,
+      "loss": 0.666,
+      "step": 17470
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.359449250217625,
+      "learning_rate": 9.80148378424711e-07,
+      "loss": 0.6582,
+      "step": 17471
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.452055045636468,
+      "learning_rate": 9.800517896444604e-07,
+      "loss": 0.6126,
+      "step": 17472
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.410774990895538,
+      "learning_rate": 9.79955201050388e-07,
+      "loss": 0.6413,
+      "step": 17473
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.442921550199329,
+      "learning_rate": 9.798586126433958e-07,
+      "loss": 0.7051,
+      "step": 17474
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.8228060815215996,
+      "learning_rate": 9.797620244243847e-07,
+      "loss": 0.638,
+      "step": 17475
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.7802573618408366,
+      "learning_rate": 9.796654363942562e-07,
+      "loss": 0.6107,
+      "step": 17476
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.488928971299095,
+      "learning_rate": 9.795688485539127e-07,
+      "loss": 0.6227,
+      "step": 17477
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.136929266456842,
+      "learning_rate": 9.794722609042545e-07,
+      "loss": 0.6126,
+      "step": 17478
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.950814181594509,
+      "learning_rate": 9.793756734461833e-07,
+      "loss": 0.6348,
+      "step": 17479
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.454600600253988,
+      "learning_rate": 9.79279086180601e-07,
+      "loss": 0.6536,
+      "step": 17480
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.173632657989819,
+      "learning_rate": 9.791824991084088e-07,
+      "loss": 0.6107,
+      "step": 17481
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.569468477573112,
+      "learning_rate": 9.790859122305082e-07,
+      "loss": 0.6465,
+      "step": 17482
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.2539369848606636,
+      "learning_rate": 9.789893255478006e-07,
+      "loss": 0.6354,
+      "step": 17483
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 6.6919652161257295,
+      "learning_rate": 9.788927390611876e-07,
+      "loss": 0.6654,
+      "step": 17484
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.103323014150516,
+      "learning_rate": 9.787961527715706e-07,
+      "loss": 0.6296,
+      "step": 17485
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.7174708403501775,
+      "learning_rate": 9.786995666798506e-07,
+      "loss": 0.5967,
+      "step": 17486
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.936106147355737,
+      "learning_rate": 9.786029807869294e-07,
+      "loss": 0.6042,
+      "step": 17487
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.412365351758187,
+      "learning_rate": 9.78506395093709e-07,
+      "loss": 0.6445,
+      "step": 17488
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.332410241920782,
+      "learning_rate": 9.784098096010897e-07,
+      "loss": 0.5723,
+      "step": 17489
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.673028478768332,
+      "learning_rate": 9.78313224309974e-07,
+      "loss": 0.6152,
+      "step": 17490
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.8946880645101944,
+      "learning_rate": 9.782166392212631e-07,
+      "loss": 0.6354,
+      "step": 17491
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.8525945309320138,
+      "learning_rate": 9.781200543358578e-07,
+      "loss": 0.6432,
+      "step": 17492
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.451557408655703,
+      "learning_rate": 9.7802346965466e-07,
+      "loss": 0.6061,
+      "step": 17493
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.089037225593564,
+      "learning_rate": 9.77926885178571e-07,
+      "loss": 0.6758,
+      "step": 17494
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.61726164314525,
+      "learning_rate": 9.778303009084928e-07,
+      "loss": 0.6536,
+      "step": 17495
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.776572708877954,
+      "learning_rate": 9.777337168453258e-07,
+      "loss": 0.6276,
+      "step": 17496
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.98601963204349,
+      "learning_rate": 9.776371329899727e-07,
+      "loss": 0.6038,
+      "step": 17497
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.8613197974757822,
+      "learning_rate": 9.775405493433336e-07,
+      "loss": 0.7207,
+      "step": 17498
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.194452760184603,
+      "learning_rate": 9.774439659063107e-07,
+      "loss": 0.6517,
+      "step": 17499
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.652929186472461,
+      "learning_rate": 9.77347382679805e-07,
+      "loss": 0.6533,
+      "step": 17500
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.7851048496035746,
+      "learning_rate": 9.772507996647187e-07,
+      "loss": 0.6699,
+      "step": 17501
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.185015315210823,
+      "learning_rate": 9.771542168619522e-07,
+      "loss": 0.612,
+      "step": 17502
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 5.76216690234867,
+      "learning_rate": 9.77057634272408e-07,
+      "loss": 0.6693,
+      "step": 17503
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.6425544901699687,
+      "learning_rate": 9.769610518969867e-07,
+      "loss": 0.612,
+      "step": 17504
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.124103032896323,
+      "learning_rate": 9.768644697365897e-07,
+      "loss": 0.6328,
+      "step": 17505
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.015883861898645,
+      "learning_rate": 9.767678877921187e-07,
+      "loss": 0.5951,
+      "step": 17506
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.578040756275895,
+      "learning_rate": 9.76671306064475e-07,
+      "loss": 0.6074,
+      "step": 17507
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.43595895075806,
+      "learning_rate": 9.765747245545602e-07,
+      "loss": 0.6608,
+      "step": 17508
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.3196087538798142,
+      "learning_rate": 9.764781432632757e-07,
+      "loss": 0.6289,
+      "step": 17509
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.9999314956254066,
+      "learning_rate": 9.763815621915225e-07,
+      "loss": 0.6673,
+      "step": 17510
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.240930444252206,
+      "learning_rate": 9.762849813402027e-07,
+      "loss": 0.6699,
+      "step": 17511
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.1047737568740477,
+      "learning_rate": 9.76188400710217e-07,
+      "loss": 0.6328,
+      "step": 17512
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.2425767166445034,
+      "learning_rate": 9.76091820302467e-07,
+      "loss": 0.6471,
+      "step": 17513
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.7825560525025104,
+      "learning_rate": 9.759952401178542e-07,
+      "loss": 0.6296,
+      "step": 17514
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.3424217832116327,
+      "learning_rate": 9.758986601572799e-07,
+      "loss": 0.6328,
+      "step": 17515
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.386683678454218,
+      "learning_rate": 9.758020804216456e-07,
+      "loss": 0.6315,
+      "step": 17516
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.892719224859221,
+      "learning_rate": 9.757055009118529e-07,
+      "loss": 0.5898,
+      "step": 17517
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.9417648412610937,
+      "learning_rate": 9.756089216288024e-07,
+      "loss": 0.6341,
+      "step": 17518
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.5621419684391915,
+      "learning_rate": 9.75512342573396e-07,
+      "loss": 0.6549,
+      "step": 17519
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.368351286481359,
+      "learning_rate": 9.754157637465354e-07,
+      "loss": 0.6836,
+      "step": 17520
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.4592220685943484,
+      "learning_rate": 9.753191851491215e-07,
+      "loss": 0.6393,
+      "step": 17521
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.3845155478031215,
+      "learning_rate": 9.752226067820558e-07,
+      "loss": 0.6608,
+      "step": 17522
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.874136260139411,
+      "learning_rate": 9.751260286462397e-07,
+      "loss": 0.6445,
+      "step": 17523
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.771340903753417,
+      "learning_rate": 9.75029450742575e-07,
+      "loss": 0.6549,
+      "step": 17524
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.3312549939969855,
+      "learning_rate": 9.749328730719623e-07,
+      "loss": 0.6211,
+      "step": 17525
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.634258636449962,
+      "learning_rate": 9.74836295635303e-07,
+      "loss": 0.6367,
+      "step": 17526
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.9821629066669395,
+      "learning_rate": 9.747397184334992e-07,
+      "loss": 0.5983,
+      "step": 17527
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.5310493819656363,
+      "learning_rate": 9.746431414674517e-07,
+      "loss": 0.6406,
+      "step": 17528
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 7.420864831907942,
+      "learning_rate": 9.745465647380618e-07,
+      "loss": 0.6784,
+      "step": 17529
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.7535534183379085,
+      "learning_rate": 9.744499882462316e-07,
+      "loss": 0.5853,
+      "step": 17530
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.234881813846411,
+      "learning_rate": 9.743534119928616e-07,
+      "loss": 0.6354,
+      "step": 17531
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.342361678753801,
+      "learning_rate": 9.742568359788532e-07,
+      "loss": 0.6862,
+      "step": 17532
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 4.304423624746413,
+      "learning_rate": 9.741602602051084e-07,
+      "loss": 0.6152,
+      "step": 17533
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.258456328262849,
+      "learning_rate": 9.74063684672528e-07,
+      "loss": 0.6641,
+      "step": 17534
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.177200819377178,
+      "learning_rate": 9.739671093820137e-07,
+      "loss": 0.623,
+      "step": 17535
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.8901584180917546,
+      "learning_rate": 9.738705343344664e-07,
+      "loss": 0.6387,
+      "step": 17536
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.299375413135025,
+      "learning_rate": 9.73773959530788e-07,
+      "loss": 0.6829,
+      "step": 17537
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.070332685104839,
+      "learning_rate": 9.736773849718797e-07,
+      "loss": 0.6309,
+      "step": 17538
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.485822137436113,
+      "learning_rate": 9.735808106586424e-07,
+      "loss": 0.6296,
+      "step": 17539
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.14101824701917,
+      "learning_rate": 9.734842365919776e-07,
+      "loss": 0.6478,
+      "step": 17540
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.9897958972784866,
+      "learning_rate": 9.73387662772787e-07,
+      "loss": 0.6361,
+      "step": 17541
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.0913743915855707,
+      "learning_rate": 9.732910892019714e-07,
+      "loss": 0.6263,
+      "step": 17542
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.647058756272464,
+      "learning_rate": 9.731945158804327e-07,
+      "loss": 0.6771,
+      "step": 17543
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.878671287079614,
+      "learning_rate": 9.73097942809072e-07,
+      "loss": 0.6562,
+      "step": 17544
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.8984629904643255,
+      "learning_rate": 9.730013699887905e-07,
+      "loss": 0.623,
+      "step": 17545
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.4585240324672903,
+      "learning_rate": 9.729047974204898e-07,
+      "loss": 0.6097,
+      "step": 17546
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.9890698925644648,
+      "learning_rate": 9.728082251050705e-07,
+      "loss": 0.6312,
+      "step": 17547
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.065144338453511,
+      "learning_rate": 9.727116530434348e-07,
+      "loss": 0.6348,
+      "step": 17548
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.8457048614587994,
+      "learning_rate": 9.726150812364835e-07,
+      "loss": 0.6074,
+      "step": 17549
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.2849807570022653,
+      "learning_rate": 9.725185096851185e-07,
+      "loss": 0.6686,
+      "step": 17550
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.939533720890301,
+      "learning_rate": 9.724219383902403e-07,
+      "loss": 0.6217,
+      "step": 17551
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.121447765002271,
+      "learning_rate": 9.723253673527505e-07,
+      "loss": 0.6999,
+      "step": 17552
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.7354319195029246,
+      "learning_rate": 9.722287965735504e-07,
+      "loss": 0.6074,
+      "step": 17553
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.8905572645176716,
+      "learning_rate": 9.721322260535418e-07,
+      "loss": 0.6471,
+      "step": 17554
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.3641939126086986,
+      "learning_rate": 9.720356557936254e-07,
+      "loss": 0.6449,
+      "step": 17555
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.9780706495869333,
+      "learning_rate": 9.719390857947028e-07,
+      "loss": 0.6178,
+      "step": 17556
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.794671023181731,
+      "learning_rate": 9.718425160576753e-07,
+      "loss": 0.6185,
+      "step": 17557
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.18337956061762,
+      "learning_rate": 9.717459465834439e-07,
+      "loss": 0.6387,
+      "step": 17558
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.454401670935251,
+      "learning_rate": 9.7164937737291e-07,
+      "loss": 0.6732,
+      "step": 17559
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.1948680049778844,
+      "learning_rate": 9.71552808426975e-07,
+      "loss": 0.623,
+      "step": 17560
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.8479177751303086,
+      "learning_rate": 9.714562397465402e-07,
+      "loss": 0.6094,
+      "step": 17561
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.049574617303882,
+      "learning_rate": 9.713596713325067e-07,
+      "loss": 0.6602,
+      "step": 17562
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.570112066087917,
+      "learning_rate": 9.712631031857764e-07,
+      "loss": 0.6348,
+      "step": 17563
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.080158020765884,
+      "learning_rate": 9.711665353072495e-07,
+      "loss": 0.6217,
+      "step": 17564
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.1795189224467646,
+      "learning_rate": 9.710699676978284e-07,
+      "loss": 0.6497,
+      "step": 17565
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.7532886380725117,
+      "learning_rate": 9.709734003584133e-07,
+      "loss": 0.5885,
+      "step": 17566
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.123207873329007,
+      "learning_rate": 9.708768332899064e-07,
+      "loss": 0.6224,
+      "step": 17567
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.001413380685775,
+      "learning_rate": 9.707802664932085e-07,
+      "loss": 0.6829,
+      "step": 17568
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.021877181999723,
+      "learning_rate": 9.70683699969221e-07,
+      "loss": 0.6243,
+      "step": 17569
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.1942754165817036,
+      "learning_rate": 9.705871337188452e-07,
+      "loss": 0.5612,
+      "step": 17570
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.2304826142743486,
+      "learning_rate": 9.704905677429822e-07,
+      "loss": 0.5869,
+      "step": 17571
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.074432899410848,
+      "learning_rate": 9.703940020425332e-07,
+      "loss": 0.6439,
+      "step": 17572
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.4742271641976705,
+      "learning_rate": 9.702974366183996e-07,
+      "loss": 0.6784,
+      "step": 17573
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.147265738820831,
+      "learning_rate": 9.702008714714826e-07,
+      "loss": 0.6224,
+      "step": 17574
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.8292319258288785,
+      "learning_rate": 9.701043066026836e-07,
+      "loss": 0.6374,
+      "step": 17575
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.106026934611492,
+      "learning_rate": 9.700077420129039e-07,
+      "loss": 0.6243,
+      "step": 17576
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.537721386693679,
+      "learning_rate": 9.699111777030448e-07,
+      "loss": 0.6139,
+      "step": 17577
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.4791577941632137,
+      "learning_rate": 9.698146136740072e-07,
+      "loss": 0.6562,
+      "step": 17578
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.1266944855609586,
+      "learning_rate": 9.69718049926692e-07,
+      "loss": 0.6243,
+      "step": 17579
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.947561635418824,
+      "learning_rate": 9.696214864620012e-07,
+      "loss": 0.6862,
+      "step": 17580
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.634245656497065,
+      "learning_rate": 9.695249232808358e-07,
+      "loss": 0.6621,
+      "step": 17581
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.5194405195995713,
+      "learning_rate": 9.69428360384097e-07,
+      "loss": 0.612,
+      "step": 17582
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 6.458677002213588,
+      "learning_rate": 9.693317977726863e-07,
+      "loss": 0.6641,
+      "step": 17583
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.1772197214735645,
+      "learning_rate": 9.692352354475042e-07,
+      "loss": 0.6628,
+      "step": 17584
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.5781154341706722,
+      "learning_rate": 9.691386734094523e-07,
+      "loss": 0.6602,
+      "step": 17585
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.5687277949070406,
+      "learning_rate": 9.690421116594322e-07,
+      "loss": 0.6582,
+      "step": 17586
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.071250940545849,
+      "learning_rate": 9.689455501983446e-07,
+      "loss": 0.6647,
+      "step": 17587
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.027411510735821,
+      "learning_rate": 9.688489890270912e-07,
+      "loss": 0.6009,
+      "step": 17588
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 7.420333039892992,
+      "learning_rate": 9.687524281465724e-07,
+      "loss": 0.6302,
+      "step": 17589
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.7096818693639184,
+      "learning_rate": 9.686558675576908e-07,
+      "loss": 0.6191,
+      "step": 17590
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.305961134533792,
+      "learning_rate": 9.685593072613463e-07,
+      "loss": 0.6022,
+      "step": 17591
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.8966955018129044,
+      "learning_rate": 9.684627472584403e-07,
+      "loss": 0.6582,
+      "step": 17592
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.498298324541048,
+      "learning_rate": 9.683661875498747e-07,
+      "loss": 0.6699,
+      "step": 17593
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.6024784893867725,
+      "learning_rate": 9.682696281365498e-07,
+      "loss": 0.6608,
+      "step": 17594
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.075145772533956,
+      "learning_rate": 9.681730690193675e-07,
+      "loss": 0.6654,
+      "step": 17595
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.9372989694666587,
+      "learning_rate": 9.680765101992294e-07,
+      "loss": 0.5983,
+      "step": 17596
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.6105850567763236,
+      "learning_rate": 9.679799516770352e-07,
+      "loss": 0.6361,
+      "step": 17597
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 5.6232439678590564,
+      "learning_rate": 9.67883393453687e-07,
+      "loss": 0.6247,
+      "step": 17598
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 3.830865742917568,
+      "learning_rate": 9.677868355300864e-07,
+      "loss": 0.6087,
+      "step": 17599
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 4.325183982289669,
+      "learning_rate": 9.676902779071336e-07,
+      "loss": 0.6022,
+      "step": 17600
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.0095322591069866,
+      "learning_rate": 9.675937205857306e-07,
+      "loss": 0.5996,
+      "step": 17601
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.7810406374048697,
+      "learning_rate": 9.674971635667782e-07,
+      "loss": 0.653,
+      "step": 17602
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.148379923616934,
+      "learning_rate": 9.674006068511778e-07,
+      "loss": 0.625,
+      "step": 17603
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.6602236398769543,
+      "learning_rate": 9.673040504398304e-07,
+      "loss": 0.6159,
+      "step": 17604
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.026646714204084,
+      "learning_rate": 9.672074943336369e-07,
+      "loss": 0.6439,
+      "step": 17605
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.337397427883007,
+      "learning_rate": 9.67110938533499e-07,
+      "loss": 0.6263,
+      "step": 17606
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.9757081493013757,
+      "learning_rate": 9.670143830403175e-07,
+      "loss": 0.6061,
+      "step": 17607
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.78089794237827,
+      "learning_rate": 9.669178278549936e-07,
+      "loss": 0.6048,
+      "step": 17608
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.158642810363513,
+      "learning_rate": 9.668212729784286e-07,
+      "loss": 0.681,
+      "step": 17609
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.2269177855312754,
+      "learning_rate": 9.66724718411524e-07,
+      "loss": 0.6628,
+      "step": 17610
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.142447232723732,
+      "learning_rate": 9.666281641551801e-07,
+      "loss": 0.6523,
+      "step": 17611
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.8848150317220407,
+      "learning_rate": 9.665316102102987e-07,
+      "loss": 0.5742,
+      "step": 17612
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.089067915333971,
+      "learning_rate": 9.664350565777805e-07,
+      "loss": 0.5898,
+      "step": 17613
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.1636488292589733,
+      "learning_rate": 9.663385032585271e-07,
+      "loss": 0.6126,
+      "step": 17614
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.465127194756448,
+      "learning_rate": 9.662419502534391e-07,
+      "loss": 0.6764,
+      "step": 17615
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.759702962220651,
+      "learning_rate": 9.661453975634186e-07,
+      "loss": 0.5905,
+      "step": 17616
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.600304498379421,
+      "learning_rate": 9.660488451893659e-07,
+      "loss": 0.6439,
+      "step": 17617
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.286387786545267,
+      "learning_rate": 9.659522931321818e-07,
+      "loss": 0.6986,
+      "step": 17618
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 6.807232660475792,
+      "learning_rate": 9.658557413927683e-07,
+      "loss": 0.6549,
+      "step": 17619
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.0369580497876356,
+      "learning_rate": 9.657591899720262e-07,
+      "loss": 0.6159,
+      "step": 17620
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.393719743060412,
+      "learning_rate": 9.656626388708565e-07,
+      "loss": 0.681,
+      "step": 17621
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.352345341379561,
+      "learning_rate": 9.655660880901604e-07,
+      "loss": 0.6517,
+      "step": 17622
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.117841661194097,
+      "learning_rate": 9.654695376308397e-07,
+      "loss": 0.6204,
+      "step": 17623
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.743244078034846,
+      "learning_rate": 9.65372987493794e-07,
+      "loss": 0.6445,
+      "step": 17624
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.249499502652806,
+      "learning_rate": 9.652764376799257e-07,
+      "loss": 0.6152,
+      "step": 17625
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.267725972519825,
+      "learning_rate": 9.651798881901354e-07,
+      "loss": 0.6628,
+      "step": 17626
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.2900519057589195,
+      "learning_rate": 9.650833390253243e-07,
+      "loss": 0.6517,
+      "step": 17627
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.2181327990027024,
+      "learning_rate": 9.649867901863932e-07,
+      "loss": 0.6172,
+      "step": 17628
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.1272524161459163,
+      "learning_rate": 9.64890241674244e-07,
+      "loss": 0.623,
+      "step": 17629
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.499546474639508,
+      "learning_rate": 9.647936934897768e-07,
+      "loss": 0.6732,
+      "step": 17630
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.375379021216879,
+      "learning_rate": 9.646971456338936e-07,
+      "loss": 0.6328,
+      "step": 17631
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.882268852678586,
+      "learning_rate": 9.646005981074945e-07,
+      "loss": 0.6478,
+      "step": 17632
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.9666417758282004,
+      "learning_rate": 9.645040509114817e-07,
+      "loss": 0.5996,
+      "step": 17633
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.626683303648528,
+      "learning_rate": 9.644075040467555e-07,
+      "loss": 0.6387,
+      "step": 17634
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.229233194157166,
+      "learning_rate": 9.643109575142173e-07,
+      "loss": 0.6432,
+      "step": 17635
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.962659952247064,
+      "learning_rate": 9.642144113147681e-07,
+      "loss": 0.653,
+      "step": 17636
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.638435434212089,
+      "learning_rate": 9.641178654493088e-07,
+      "loss": 0.6556,
+      "step": 17637
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.195201942921678,
+      "learning_rate": 9.640213199187408e-07,
+      "loss": 0.6478,
+      "step": 17638
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.095100188100678,
+      "learning_rate": 9.639247747239649e-07,
+      "loss": 0.6562,
+      "step": 17639
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.3556807937381357,
+      "learning_rate": 9.638282298658822e-07,
+      "loss": 0.7064,
+      "step": 17640
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.6214648298235192,
+      "learning_rate": 9.637316853453938e-07,
+      "loss": 0.6335,
+      "step": 17641
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.313779631851879,
+      "learning_rate": 9.636351411634008e-07,
+      "loss": 0.6178,
+      "step": 17642
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.918935104568814,
+      "learning_rate": 9.635385973208047e-07,
+      "loss": 0.7135,
+      "step": 17643
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.0606800663716776,
+      "learning_rate": 9.634420538185059e-07,
+      "loss": 0.6204,
+      "step": 17644
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.951903203264599,
+      "learning_rate": 9.633455106574054e-07,
+      "loss": 0.6302,
+      "step": 17645
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.093785371479615,
+      "learning_rate": 9.632489678384044e-07,
+      "loss": 0.6178,
+      "step": 17646
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.003741744258961,
+      "learning_rate": 9.631524253624042e-07,
+      "loss": 0.6276,
+      "step": 17647
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.7302586212190563,
+      "learning_rate": 9.630558832303058e-07,
+      "loss": 0.6064,
+      "step": 17648
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.1170411939158065,
+      "learning_rate": 9.629593414430102e-07,
+      "loss": 0.6621,
+      "step": 17649
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.9483576941421337,
+      "learning_rate": 9.62862800001418e-07,
+      "loss": 0.651,
+      "step": 17650
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.041050625484223,
+      "learning_rate": 9.627662589064306e-07,
+      "loss": 0.7285,
+      "step": 17651
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.834806545635652,
+      "learning_rate": 9.626697181589493e-07,
+      "loss": 0.6491,
+      "step": 17652
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.440048501088676,
+      "learning_rate": 9.625731777598744e-07,
+      "loss": 0.6406,
+      "step": 17653
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.5294714327031382,
+      "learning_rate": 9.624766377101077e-07,
+      "loss": 0.6322,
+      "step": 17654
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.637121763909491,
+      "learning_rate": 9.623800980105495e-07,
+      "loss": 0.6595,
+      "step": 17655
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.06179136340959,
+      "learning_rate": 9.622835586621017e-07,
+      "loss": 0.6243,
+      "step": 17656
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.069170436255303,
+      "learning_rate": 9.621870196656647e-07,
+      "loss": 0.6406,
+      "step": 17657
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.141104837490692,
+      "learning_rate": 9.620904810221393e-07,
+      "loss": 0.5964,
+      "step": 17658
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.4992447495633487,
+      "learning_rate": 9.619939427324267e-07,
+      "loss": 0.6595,
+      "step": 17659
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.601358709503619,
+      "learning_rate": 9.618974047974281e-07,
+      "loss": 0.6335,
+      "step": 17660
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.563387837358582,
+      "learning_rate": 9.618008672180444e-07,
+      "loss": 0.6061,
+      "step": 17661
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.8540892366008346,
+      "learning_rate": 9.61704329995177e-07,
+      "loss": 0.6204,
+      "step": 17662
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.8529711878954114,
+      "learning_rate": 9.61607793129726e-07,
+      "loss": 0.6797,
+      "step": 17663
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.9012980965492376,
+      "learning_rate": 9.61511256622593e-07,
+      "loss": 0.666,
+      "step": 17664
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.649116634841249,
+      "learning_rate": 9.614147204746788e-07,
+      "loss": 0.6589,
+      "step": 17665
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 4.192293400200361,
+      "learning_rate": 9.613181846868843e-07,
+      "loss": 0.627,
+      "step": 17666
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.4871830592082427,
+      "learning_rate": 9.612216492601108e-07,
+      "loss": 0.6569,
+      "step": 17667
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.7522567525716672,
+      "learning_rate": 9.61125114195259e-07,
+      "loss": 0.6328,
+      "step": 17668
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.2835102455111707,
+      "learning_rate": 9.6102857949323e-07,
+      "loss": 0.6022,
+      "step": 17669
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.6872313512894834,
+      "learning_rate": 9.60932045154925e-07,
+      "loss": 0.6608,
+      "step": 17670
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.357758093844444,
+      "learning_rate": 9.608355111812442e-07,
+      "loss": 0.6829,
+      "step": 17671
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.837123589793168,
+      "learning_rate": 9.607389775730892e-07,
+      "loss": 0.6087,
+      "step": 17672
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.0203462361995754,
+      "learning_rate": 9.606424443313607e-07,
+      "loss": 0.6576,
+      "step": 17673
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.0420334333446304,
+      "learning_rate": 9.605459114569598e-07,
+      "loss": 0.557,
+      "step": 17674
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.0530698581585844,
+      "learning_rate": 9.604493789507874e-07,
+      "loss": 0.6725,
+      "step": 17675
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.517119436999365,
+      "learning_rate": 9.603528468137448e-07,
+      "loss": 0.6641,
+      "step": 17676
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.1123064562282714,
+      "learning_rate": 9.60256315046732e-07,
+      "loss": 0.6719,
+      "step": 17677
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.843801195559219,
+      "learning_rate": 9.60159783650651e-07,
+      "loss": 0.6387,
+      "step": 17678
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.053601948684102,
+      "learning_rate": 9.600632526264019e-07,
+      "loss": 0.64,
+      "step": 17679
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.7493307257386967,
+      "learning_rate": 9.599667219748862e-07,
+      "loss": 0.6432,
+      "step": 17680
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.247753621858947,
+      "learning_rate": 9.598701916970044e-07,
+      "loss": 0.6543,
+      "step": 17681
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.051797973223644,
+      "learning_rate": 9.597736617936582e-07,
+      "loss": 0.6042,
+      "step": 17682
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.131845974729202,
+      "learning_rate": 9.596771322657479e-07,
+      "loss": 0.6504,
+      "step": 17683
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.76029961804081,
+      "learning_rate": 9.595806031141738e-07,
+      "loss": 0.6367,
+      "step": 17684
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.266661934104959,
+      "learning_rate": 9.594840743398378e-07,
+      "loss": 0.6074,
+      "step": 17685
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.2319251260727198,
+      "learning_rate": 9.593875459436408e-07,
+      "loss": 0.6621,
+      "step": 17686
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.7593880904630277,
+      "learning_rate": 9.59291017926483e-07,
+      "loss": 0.5931,
+      "step": 17687
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.689116854775534,
+      "learning_rate": 9.59194490289266e-07,
+      "loss": 0.5951,
+      "step": 17688
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 7.285647996430094,
+      "learning_rate": 9.59097963032891e-07,
+      "loss": 0.724,
+      "step": 17689
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.8123589090284185,
+      "learning_rate": 9.590014361582573e-07,
+      "loss": 0.6569,
+      "step": 17690
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.9042496591593525,
+      "learning_rate": 9.589049096662674e-07,
+      "loss": 0.623,
+      "step": 17691
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.1618651290119715,
+      "learning_rate": 9.588083835578214e-07,
+      "loss": 0.6562,
+      "step": 17692
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.012831752326475,
+      "learning_rate": 9.587118578338205e-07,
+      "loss": 0.6406,
+      "step": 17693
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.368534030655689,
+      "learning_rate": 9.586153324951652e-07,
+      "loss": 0.6895,
+      "step": 17694
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.841963129158022,
+      "learning_rate": 9.585188075427568e-07,
+      "loss": 0.6315,
+      "step": 17695
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.38000343442709,
+      "learning_rate": 9.584222829774965e-07,
+      "loss": 0.6543,
+      "step": 17696
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.907538379199676,
+      "learning_rate": 9.583257588002844e-07,
+      "loss": 0.7161,
+      "step": 17697
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.5207896994199483,
+      "learning_rate": 9.582292350120214e-07,
+      "loss": 0.6569,
+      "step": 17698
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.3607139061172253,
+      "learning_rate": 9.58132711613609e-07,
+      "loss": 0.6934,
+      "step": 17699
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.376779767741697,
+      "learning_rate": 9.580361886059474e-07,
+      "loss": 0.6536,
+      "step": 17700
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.0120588283228713,
+      "learning_rate": 9.579396659899382e-07,
+      "loss": 0.6217,
+      "step": 17701
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.575641281845368,
+      "learning_rate": 9.578431437664819e-07,
+      "loss": 0.6257,
+      "step": 17702
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.9644766649792516,
+      "learning_rate": 9.577466219364785e-07,
+      "loss": 0.599,
+      "step": 17703
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.3341160618553216,
+      "learning_rate": 9.576501005008302e-07,
+      "loss": 0.6309,
+      "step": 17704
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.450260144515322,
+      "learning_rate": 9.57553579460437e-07,
+      "loss": 0.6322,
+      "step": 17705
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.295493632736048,
+      "learning_rate": 9.574570588162e-07,
+      "loss": 0.6374,
+      "step": 17706
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.366294865714473,
+      "learning_rate": 9.573605385690201e-07,
+      "loss": 0.6706,
+      "step": 17707
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.189527884563679,
+      "learning_rate": 9.57264018719798e-07,
+      "loss": 0.666,
+      "step": 17708
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.572093323704653,
+      "learning_rate": 9.571674992694351e-07,
+      "loss": 0.5911,
+      "step": 17709
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.466011738674901,
+      "learning_rate": 9.570709802188313e-07,
+      "loss": 0.6198,
+      "step": 17710
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.8794592051735513,
+      "learning_rate": 9.569744615688878e-07,
+      "loss": 0.6419,
+      "step": 17711
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.8057487430129546,
+      "learning_rate": 9.568779433205054e-07,
+      "loss": 0.5918,
+      "step": 17712
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.915781043574451,
+      "learning_rate": 9.567814254745849e-07,
+      "loss": 0.668,
+      "step": 17713
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.448428054153503,
+      "learning_rate": 9.566849080320276e-07,
+      "loss": 0.6589,
+      "step": 17714
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.0389501708086923,
+      "learning_rate": 9.565883909937339e-07,
+      "loss": 0.6536,
+      "step": 17715
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.7492859927163296,
+      "learning_rate": 9.56491874360604e-07,
+      "loss": 0.6257,
+      "step": 17716
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.975912493600467,
+      "learning_rate": 9.563953581335396e-07,
+      "loss": 0.612,
+      "step": 17717
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.164447030473554,
+      "learning_rate": 9.562988423134412e-07,
+      "loss": 0.6029,
+      "step": 17718
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.099400730016192,
+      "learning_rate": 9.562023269012094e-07,
+      "loss": 0.6667,
+      "step": 17719
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.2987597311295365,
+      "learning_rate": 9.561058118977455e-07,
+      "loss": 0.6797,
+      "step": 17720
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.845491361791202,
+      "learning_rate": 9.560092973039495e-07,
+      "loss": 0.666,
+      "step": 17721
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 6.643431572798844,
+      "learning_rate": 9.559127831207233e-07,
+      "loss": 0.6322,
+      "step": 17722
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.6037818263995924,
+      "learning_rate": 9.558162693489666e-07,
+      "loss": 0.6133,
+      "step": 17723
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.017053028051504,
+      "learning_rate": 9.557197559895803e-07,
+      "loss": 0.6204,
+      "step": 17724
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.317755690651298,
+      "learning_rate": 9.556232430434659e-07,
+      "loss": 0.5999,
+      "step": 17725
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.007476100116577,
+      "learning_rate": 9.555267305115232e-07,
+      "loss": 0.6217,
+      "step": 17726
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.5982140800163,
+      "learning_rate": 9.554302183946537e-07,
+      "loss": 0.6009,
+      "step": 17727
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 4.960104793424752,
+      "learning_rate": 9.553337066937577e-07,
+      "loss": 0.6292,
+      "step": 17728
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.993505888396005,
+      "learning_rate": 9.552371954097368e-07,
+      "loss": 0.6289,
+      "step": 17729
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.972491778012177,
+      "learning_rate": 9.551406845434906e-07,
+      "loss": 0.6523,
+      "step": 17730
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.222647623580929,
+      "learning_rate": 9.550441740959205e-07,
+      "loss": 0.6615,
+      "step": 17731
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.9311389468514,
+      "learning_rate": 9.54947664067927e-07,
+      "loss": 0.7064,
+      "step": 17732
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.9502926624027825,
+      "learning_rate": 9.54851154460411e-07,
+      "loss": 0.6549,
+      "step": 17733
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 5.306484414441007,
+      "learning_rate": 9.54754645274273e-07,
+      "loss": 0.6608,
+      "step": 17734
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.5835169835367844,
+      "learning_rate": 9.546581365104143e-07,
+      "loss": 0.6738,
+      "step": 17735
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.052731705151504,
+      "learning_rate": 9.54561628169735e-07,
+      "loss": 0.6491,
+      "step": 17736
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.952112064736356,
+      "learning_rate": 9.54465120253136e-07,
+      "loss": 0.6152,
+      "step": 17737
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.351726185450151,
+      "learning_rate": 9.543686127615182e-07,
+      "loss": 0.6921,
+      "step": 17738
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.4511820678713194,
+      "learning_rate": 9.54272105695782e-07,
+      "loss": 0.6595,
+      "step": 17739
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.2349257024757323,
+      "learning_rate": 9.54175599056828e-07,
+      "loss": 0.6198,
+      "step": 17740
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.7669546690496563,
+      "learning_rate": 9.540790928455577e-07,
+      "loss": 0.6576,
+      "step": 17741
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.132191704301948,
+      "learning_rate": 9.539825870628714e-07,
+      "loss": 0.6549,
+      "step": 17742
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.8772122348282774,
+      "learning_rate": 9.538860817096693e-07,
+      "loss": 0.7077,
+      "step": 17743
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.064278289145504,
+      "learning_rate": 9.537895767868525e-07,
+      "loss": 0.6178,
+      "step": 17744
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.182273478151772,
+      "learning_rate": 9.536930722953218e-07,
+      "loss": 0.6276,
+      "step": 17745
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.354841387878707,
+      "learning_rate": 9.535965682359777e-07,
+      "loss": 0.5846,
+      "step": 17746
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.9586224785838016,
+      "learning_rate": 9.535000646097207e-07,
+      "loss": 0.6706,
+      "step": 17747
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.0096345591430516,
+      "learning_rate": 9.534035614174523e-07,
+      "loss": 0.6016,
+      "step": 17748
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.676344519500544,
+      "learning_rate": 9.533070586600723e-07,
+      "loss": 0.6081,
+      "step": 17749
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.776061542900787,
+      "learning_rate": 9.532105563384814e-07,
+      "loss": 0.64,
+      "step": 17750
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.668383334292478,
+      "learning_rate": 9.531140544535805e-07,
+      "loss": 0.6426,
+      "step": 17751
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.042107397540998,
+      "learning_rate": 9.530175530062705e-07,
+      "loss": 0.6641,
+      "step": 17752
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.859766706229997,
+      "learning_rate": 9.529210519974516e-07,
+      "loss": 0.6126,
+      "step": 17753
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.9547815074192365,
+      "learning_rate": 9.528245514280249e-07,
+      "loss": 0.6686,
+      "step": 17754
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.8882214571007885,
+      "learning_rate": 9.527280512988913e-07,
+      "loss": 0.6283,
+      "step": 17755
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.5907832636827073,
+      "learning_rate": 9.526315516109502e-07,
+      "loss": 0.5742,
+      "step": 17756
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.016262501919679,
+      "learning_rate": 9.525350523651035e-07,
+      "loss": 0.6165,
+      "step": 17757
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.645683629356357,
+      "learning_rate": 9.52438553562251e-07,
+      "loss": 0.6634,
+      "step": 17758
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.809781581811608,
+      "learning_rate": 9.523420552032939e-07,
+      "loss": 0.64,
+      "step": 17759
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.270344880954026,
+      "learning_rate": 9.522455572891323e-07,
+      "loss": 0.653,
+      "step": 17760
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.1132507101368443,
+      "learning_rate": 9.521490598206674e-07,
+      "loss": 0.6393,
+      "step": 17761
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.4004535313018693,
+      "learning_rate": 9.520525627987997e-07,
+      "loss": 0.6393,
+      "step": 17762
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.546635188669679,
+      "learning_rate": 9.519560662244297e-07,
+      "loss": 0.6354,
+      "step": 17763
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.8694821917372595,
+      "learning_rate": 9.518595700984577e-07,
+      "loss": 0.6341,
+      "step": 17764
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.867619698896037,
+      "learning_rate": 9.517630744217847e-07,
+      "loss": 0.6419,
+      "step": 17765
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.5104362553389783,
+      "learning_rate": 9.516665791953111e-07,
+      "loss": 0.6497,
+      "step": 17766
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 7.145952585975903,
+      "learning_rate": 9.515700844199377e-07,
+      "loss": 0.6374,
+      "step": 17767
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.7612156810185544,
+      "learning_rate": 9.514735900965653e-07,
+      "loss": 0.6387,
+      "step": 17768
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.1951885350970315,
+      "learning_rate": 9.513770962260937e-07,
+      "loss": 0.6816,
+      "step": 17769
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.405518722243147,
+      "learning_rate": 9.512806028094241e-07,
+      "loss": 0.6061,
+      "step": 17770
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.050806283219642,
+      "learning_rate": 9.511841098474569e-07,
+      "loss": 0.6667,
+      "step": 17771
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.3253533563406656,
+      "learning_rate": 9.510876173410925e-07,
+      "loss": 0.6003,
+      "step": 17772
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.098123418623327,
+      "learning_rate": 9.509911252912321e-07,
+      "loss": 0.6452,
+      "step": 17773
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.6563009117982,
+      "learning_rate": 9.508946336987756e-07,
+      "loss": 0.6348,
+      "step": 17774
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.406808080048502,
+      "learning_rate": 9.507981425646242e-07,
+      "loss": 0.6673,
+      "step": 17775
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.401229749257686,
+      "learning_rate": 9.507016518896779e-07,
+      "loss": 0.6764,
+      "step": 17776
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.156126028979888,
+      "learning_rate": 9.506051616748373e-07,
+      "loss": 0.6123,
+      "step": 17777
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.854037409216599,
+      "learning_rate": 9.505086719210031e-07,
+      "loss": 0.6432,
+      "step": 17778
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.149375324337243,
+      "learning_rate": 9.504121826290759e-07,
+      "loss": 0.6426,
+      "step": 17779
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 8.545947619721105,
+      "learning_rate": 9.503156937999562e-07,
+      "loss": 0.638,
+      "step": 17780
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.2228744017072284,
+      "learning_rate": 9.502192054345449e-07,
+      "loss": 0.6628,
+      "step": 17781
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.02609203710027,
+      "learning_rate": 9.501227175337417e-07,
+      "loss": 0.6146,
+      "step": 17782
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.5288294400796576,
+      "learning_rate": 9.500262300984474e-07,
+      "loss": 0.681,
+      "step": 17783
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.8571190814781784,
+      "learning_rate": 9.499297431295631e-07,
+      "loss": 0.627,
+      "step": 17784
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 5.266238505743333,
+      "learning_rate": 9.498332566279887e-07,
+      "loss": 0.6712,
+      "step": 17785
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.6288669002711487,
+      "learning_rate": 9.497367705946251e-07,
+      "loss": 0.6322,
+      "step": 17786
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.0288325970921166,
+      "learning_rate": 9.496402850303724e-07,
+      "loss": 0.6126,
+      "step": 17787
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.7496152954460173,
+      "learning_rate": 9.49543799936132e-07,
+      "loss": 0.6064,
+      "step": 17788
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 6.131984269037891,
+      "learning_rate": 9.494473153128034e-07,
+      "loss": 0.6771,
+      "step": 17789
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.662014682722589,
+      "learning_rate": 9.493508311612874e-07,
+      "loss": 0.6055,
+      "step": 17790
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.1911257208618564,
+      "learning_rate": 9.492543474824845e-07,
+      "loss": 0.6914,
+      "step": 17791
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.992644942234791,
+      "learning_rate": 9.491578642772954e-07,
+      "loss": 0.651,
+      "step": 17792
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 4.953613929939214,
+      "learning_rate": 9.490613815466206e-07,
+      "loss": 0.6654,
+      "step": 17793
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.1392708253047346,
+      "learning_rate": 9.4896489929136e-07,
+      "loss": 0.6029,
+      "step": 17794
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.989116609013228,
+      "learning_rate": 9.48868417512415e-07,
+      "loss": 0.6387,
+      "step": 17795
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.3920392836235282,
+      "learning_rate": 9.487719362106853e-07,
+      "loss": 0.625,
+      "step": 17796
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.32543515705019,
+      "learning_rate": 9.486754553870718e-07,
+      "loss": 0.653,
+      "step": 17797
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.280465073624446,
+      "learning_rate": 9.485789750424745e-07,
+      "loss": 0.6562,
+      "step": 17798
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.4797212363672294,
+      "learning_rate": 9.484824951777946e-07,
+      "loss": 0.6445,
+      "step": 17799
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.326373495615128,
+      "learning_rate": 9.483860157939317e-07,
+      "loss": 0.6523,
+      "step": 17800
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.074735003077626,
+      "learning_rate": 9.482895368917872e-07,
+      "loss": 0.6367,
+      "step": 17801
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.8121289151440254,
+      "learning_rate": 9.481930584722608e-07,
+      "loss": 0.5964,
+      "step": 17802
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.6064312149840703,
+      "learning_rate": 9.480965805362529e-07,
+      "loss": 0.679,
+      "step": 17803
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.09470713927055,
+      "learning_rate": 9.480001030846646e-07,
+      "loss": 0.6263,
+      "step": 17804
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.603185572018974,
+      "learning_rate": 9.479036261183956e-07,
+      "loss": 0.6217,
+      "step": 17805
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.599281636176589,
+      "learning_rate": 9.478071496383466e-07,
+      "loss": 0.6374,
+      "step": 17806
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.863498728990495,
+      "learning_rate": 9.477106736454184e-07,
+      "loss": 0.597,
+      "step": 17807
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.91806984115593,
+      "learning_rate": 9.476141981405112e-07,
+      "loss": 0.6126,
+      "step": 17808
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.9187601556522975,
+      "learning_rate": 9.47517723124525e-07,
+      "loss": 0.612,
+      "step": 17809
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.352073925767498,
+      "learning_rate": 9.474212485983608e-07,
+      "loss": 0.6992,
+      "step": 17810
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.946485371609603,
+      "learning_rate": 9.473247745629185e-07,
+      "loss": 0.6003,
+      "step": 17811
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.1195677559472443,
+      "learning_rate": 9.472283010190989e-07,
+      "loss": 0.64,
+      "step": 17812
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.314849235999905,
+      "learning_rate": 9.47131827967802e-07,
+      "loss": 0.6569,
+      "step": 17813
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.9791385957615786,
+      "learning_rate": 9.470353554099288e-07,
+      "loss": 0.6217,
+      "step": 17814
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.4669929475689187,
+      "learning_rate": 9.469388833463792e-07,
+      "loss": 0.6849,
+      "step": 17815
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.3629916742209884,
+      "learning_rate": 9.468424117780534e-07,
+      "loss": 0.6582,
+      "step": 17816
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.3894852430743203,
+      "learning_rate": 9.467459407058521e-07,
+      "loss": 0.6452,
+      "step": 17817
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.1593581137130244,
+      "learning_rate": 9.46649470130676e-07,
+      "loss": 0.6074,
+      "step": 17818
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.117741762883233,
+      "learning_rate": 9.465530000534247e-07,
+      "loss": 0.6367,
+      "step": 17819
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.745684923314711,
+      "learning_rate": 9.464565304749993e-07,
+      "loss": 0.6452,
+      "step": 17820
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.1498376457538813,
+      "learning_rate": 9.463600613963e-07,
+      "loss": 0.6341,
+      "step": 17821
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.4910987604095345,
+      "learning_rate": 9.462635928182266e-07,
+      "loss": 0.6328,
+      "step": 17822
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.414887217638438,
+      "learning_rate": 9.4616712474168e-07,
+      "loss": 0.6289,
+      "step": 17823
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.3537943653713214,
+      "learning_rate": 9.460706571675603e-07,
+      "loss": 0.6341,
+      "step": 17824
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.4668329759720633,
+      "learning_rate": 9.45974190096768e-07,
+      "loss": 0.6393,
+      "step": 17825
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.0584078452599512,
+      "learning_rate": 9.458777235302032e-07,
+      "loss": 0.6465,
+      "step": 17826
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.798016504745578,
+      "learning_rate": 9.457812574687663e-07,
+      "loss": 0.6159,
+      "step": 17827
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.04191622351149,
+      "learning_rate": 9.456847919133584e-07,
+      "loss": 0.6536,
+      "step": 17828
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.898288051426951,
+      "learning_rate": 9.455883268648787e-07,
+      "loss": 0.6217,
+      "step": 17829
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.019952540974176,
+      "learning_rate": 9.454918623242278e-07,
+      "loss": 0.6113,
+      "step": 17830
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.450857650604771,
+      "learning_rate": 9.453953982923064e-07,
+      "loss": 0.6543,
+      "step": 17831
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.370418530536745,
+      "learning_rate": 9.452989347700143e-07,
+      "loss": 0.6621,
+      "step": 17832
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.9237517286448353,
+      "learning_rate": 9.452024717582524e-07,
+      "loss": 0.6257,
+      "step": 17833
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.2248221163066018,
+      "learning_rate": 9.451060092579208e-07,
+      "loss": 0.6276,
+      "step": 17834
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.8994214821921207,
+      "learning_rate": 9.450095472699193e-07,
+      "loss": 0.6087,
+      "step": 17835
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.8879810677550006,
+      "learning_rate": 9.449130857951488e-07,
+      "loss": 0.6315,
+      "step": 17836
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.3773317371503904,
+      "learning_rate": 9.44816624834509e-07,
+      "loss": 0.6393,
+      "step": 17837
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.5293178748777114,
+      "learning_rate": 9.447201643889007e-07,
+      "loss": 0.6562,
+      "step": 17838
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 7.260834330108559,
+      "learning_rate": 9.44623704459224e-07,
+      "loss": 0.6608,
+      "step": 17839
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.5814216867871713,
+      "learning_rate": 9.445272450463791e-07,
+      "loss": 0.668,
+      "step": 17840
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.6806712497237934,
+      "learning_rate": 9.444307861512667e-07,
+      "loss": 0.653,
+      "step": 17841
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.181598328854309,
+      "learning_rate": 9.443343277747864e-07,
+      "loss": 0.653,
+      "step": 17842
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.714309715911518,
+      "learning_rate": 9.442378699178387e-07,
+      "loss": 0.6341,
+      "step": 17843
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.043618896149813,
+      "learning_rate": 9.441414125813241e-07,
+      "loss": 0.6276,
+      "step": 17844
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.3442691846574415,
+      "learning_rate": 9.440449557661423e-07,
+      "loss": 0.6654,
+      "step": 17845
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.2171128351455827,
+      "learning_rate": 9.439484994731941e-07,
+      "loss": 0.6738,
+      "step": 17846
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.040257802820499,
+      "learning_rate": 9.438520437033793e-07,
+      "loss": 0.6406,
+      "step": 17847
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.9047497910442712,
+      "learning_rate": 9.437555884575989e-07,
+      "loss": 0.6185,
+      "step": 17848
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.442966994224173,
+      "learning_rate": 9.43659133736752e-07,
+      "loss": 0.6471,
+      "step": 17849
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.459294529941251,
+      "learning_rate": 9.435626795417396e-07,
+      "loss": 0.6908,
+      "step": 17850
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 4.807502357995917,
+      "learning_rate": 9.434662258734616e-07,
+      "loss": 0.5775,
+      "step": 17851
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.043287840208821,
+      "learning_rate": 9.433697727328184e-07,
+      "loss": 0.6654,
+      "step": 17852
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.6243038203433824,
+      "learning_rate": 9.4327332012071e-07,
+      "loss": 0.5977,
+      "step": 17853
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.857696175442868,
+      "learning_rate": 9.43176868038037e-07,
+      "loss": 0.6719,
+      "step": 17854
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.189832288205719,
+      "learning_rate": 9.430804164856994e-07,
+      "loss": 0.5954,
+      "step": 17855
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.867602072372314,
+      "learning_rate": 9.429839654645968e-07,
+      "loss": 0.6862,
+      "step": 17856
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.9493073255880544,
+      "learning_rate": 9.428875149756303e-07,
+      "loss": 0.6204,
+      "step": 17857
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 6.3602049797995575,
+      "learning_rate": 9.427910650196993e-07,
+      "loss": 0.6452,
+      "step": 17858
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.4720687941810016,
+      "learning_rate": 9.426946155977046e-07,
+      "loss": 0.6406,
+      "step": 17859
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.350326622593579,
+      "learning_rate": 9.42598166710546e-07,
+      "loss": 0.6452,
+      "step": 17860
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.212699621642072,
+      "learning_rate": 9.425017183591241e-07,
+      "loss": 0.6315,
+      "step": 17861
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.641475066912086,
+      "learning_rate": 9.424052705443383e-07,
+      "loss": 0.6243,
+      "step": 17862
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.3489624548839294,
+      "learning_rate": 9.423088232670896e-07,
+      "loss": 0.6719,
+      "step": 17863
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.5806683180414356,
+      "learning_rate": 9.422123765282775e-07,
+      "loss": 0.6348,
+      "step": 17864
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.1670665318602658,
+      "learning_rate": 9.421159303288024e-07,
+      "loss": 0.668,
+      "step": 17865
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.836204979538136,
+      "learning_rate": 9.420194846695644e-07,
+      "loss": 0.6178,
+      "step": 17866
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.4078881700576047,
+      "learning_rate": 9.419230395514642e-07,
+      "loss": 0.6178,
+      "step": 17867
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 5.9091739263688465,
+      "learning_rate": 9.418265949754009e-07,
+      "loss": 0.7129,
+      "step": 17868
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.8711964734025006,
+      "learning_rate": 9.417301509422751e-07,
+      "loss": 0.6074,
+      "step": 17869
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.548767581774667,
+      "learning_rate": 9.416337074529872e-07,
+      "loss": 0.6172,
+      "step": 17870
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.454311975433809,
+      "learning_rate": 9.415372645084368e-07,
+      "loss": 0.6283,
+      "step": 17871
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.346210842905487,
+      "learning_rate": 9.414408221095244e-07,
+      "loss": 0.6569,
+      "step": 17872
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.1409386527160676,
+      "learning_rate": 9.4134438025715e-07,
+      "loss": 0.6745,
+      "step": 17873
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.3366902214123035,
+      "learning_rate": 9.412479389522141e-07,
+      "loss": 0.6465,
+      "step": 17874
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.03690933445413,
+      "learning_rate": 9.411514981956158e-07,
+      "loss": 0.6367,
+      "step": 17875
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.185252267316281,
+      "learning_rate": 9.41055057988256e-07,
+      "loss": 0.6517,
+      "step": 17876
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.544962603956109,
+      "learning_rate": 9.409586183310344e-07,
+      "loss": 0.6478,
+      "step": 17877
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.2147660235974715,
+      "learning_rate": 9.408621792248514e-07,
+      "loss": 0.6641,
+      "step": 17878
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.7465404013659644,
+      "learning_rate": 9.407657406706068e-07,
+      "loss": 0.6497,
+      "step": 17879
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.138148305906151,
+      "learning_rate": 9.406693026692009e-07,
+      "loss": 0.6725,
+      "step": 17880
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 5.298741715973607,
+      "learning_rate": 9.405728652215339e-07,
+      "loss": 0.6178,
+      "step": 17881
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.0495495145165745,
+      "learning_rate": 9.404764283285052e-07,
+      "loss": 0.6471,
+      "step": 17882
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.999007003285316,
+      "learning_rate": 9.403799919910152e-07,
+      "loss": 0.6738,
+      "step": 17883
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.9917105053433732,
+      "learning_rate": 9.402835562099641e-07,
+      "loss": 0.651,
+      "step": 17884
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.7956645511841494,
+      "learning_rate": 9.401871209862519e-07,
+      "loss": 0.6523,
+      "step": 17885
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.173530918092411,
+      "learning_rate": 9.400906863207786e-07,
+      "loss": 0.6712,
+      "step": 17886
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.8783403129960545,
+      "learning_rate": 9.399942522144445e-07,
+      "loss": 0.6276,
+      "step": 17887
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 9.236081394600406,
+      "learning_rate": 9.39897818668149e-07,
+      "loss": 0.6068,
+      "step": 17888
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.880758800163883,
+      "learning_rate": 9.398013856827927e-07,
+      "loss": 0.6458,
+      "step": 17889
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.1605068571666126,
+      "learning_rate": 9.39704953259275e-07,
+      "loss": 0.6335,
+      "step": 17890
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.5628727065459427,
+      "learning_rate": 9.396085213984967e-07,
+      "loss": 0.6816,
+      "step": 17891
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.780433640533812,
+      "learning_rate": 9.39512090101357e-07,
+      "loss": 0.6497,
+      "step": 17892
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.096387524329257,
+      "learning_rate": 9.394156593687564e-07,
+      "loss": 0.6458,
+      "step": 17893
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.999331498902498,
+      "learning_rate": 9.393192292015954e-07,
+      "loss": 0.6725,
+      "step": 17894
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.360588192701973,
+      "learning_rate": 9.39222799600773e-07,
+      "loss": 0.638,
+      "step": 17895
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 7.447104745616933,
+      "learning_rate": 9.391263705671893e-07,
+      "loss": 0.6628,
+      "step": 17896
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.72096616092581,
+      "learning_rate": 9.39029942101745e-07,
+      "loss": 0.6328,
+      "step": 17897
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.9246699186474054,
+      "learning_rate": 9.389335142053393e-07,
+      "loss": 0.6367,
+      "step": 17898
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.513640552709621,
+      "learning_rate": 9.388370868788728e-07,
+      "loss": 0.6413,
+      "step": 17899
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.387108195403713,
+      "learning_rate": 9.387406601232454e-07,
+      "loss": 0.6348,
+      "step": 17900
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.599091466220337,
+      "learning_rate": 9.386442339393563e-07,
+      "loss": 0.6413,
+      "step": 17901
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.858426466796927,
+      "learning_rate": 9.385478083281064e-07,
+      "loss": 0.6562,
+      "step": 17902
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.358051652479251,
+      "learning_rate": 9.384513832903948e-07,
+      "loss": 0.6048,
+      "step": 17903
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.4430207421473686,
+      "learning_rate": 9.383549588271219e-07,
+      "loss": 0.6608,
+      "step": 17904
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.801908507917144,
+      "learning_rate": 9.382585349391879e-07,
+      "loss": 0.6432,
+      "step": 17905
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.8966989068488678,
+      "learning_rate": 9.381621116274922e-07,
+      "loss": 0.6816,
+      "step": 17906
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.2554673108673353,
+      "learning_rate": 9.380656888929356e-07,
+      "loss": 0.6165,
+      "step": 17907
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.6996982694004332,
+      "learning_rate": 9.37969266736417e-07,
+      "loss": 0.5918,
+      "step": 17908
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.6103893594571552,
+      "learning_rate": 9.378728451588365e-07,
+      "loss": 0.6178,
+      "step": 17909
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.925123329395304,
+      "learning_rate": 9.377764241610945e-07,
+      "loss": 0.6419,
+      "step": 17910
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.297159666355628,
+      "learning_rate": 9.376800037440904e-07,
+      "loss": 0.6413,
+      "step": 17911
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.8693104719577254,
+      "learning_rate": 9.375835839087246e-07,
+      "loss": 0.6328,
+      "step": 17912
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.376844866592482,
+      "learning_rate": 9.374871646558965e-07,
+      "loss": 0.6725,
+      "step": 17913
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.960753929944141,
+      "learning_rate": 9.373907459865066e-07,
+      "loss": 0.6387,
+      "step": 17914
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.568143900105626,
+      "learning_rate": 9.37294327901454e-07,
+      "loss": 0.6367,
+      "step": 17915
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.448951292601405,
+      "learning_rate": 9.371979104016392e-07,
+      "loss": 0.6224,
+      "step": 17916
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.11184533659313,
+      "learning_rate": 9.371014934879616e-07,
+      "loss": 0.668,
+      "step": 17917
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.3748625629391262,
+      "learning_rate": 9.370050771613216e-07,
+      "loss": 0.6406,
+      "step": 17918
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 6.218726292212582,
+      "learning_rate": 9.369086614226186e-07,
+      "loss": 0.6823,
+      "step": 17919
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.943685072766547,
+      "learning_rate": 9.36812246272753e-07,
+      "loss": 0.6283,
+      "step": 17920
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.29196695929045,
+      "learning_rate": 9.367158317126242e-07,
+      "loss": 0.6243,
+      "step": 17921
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.2593622232678228,
+      "learning_rate": 9.366194177431318e-07,
+      "loss": 0.6725,
+      "step": 17922
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.8438918683884307,
+      "learning_rate": 9.365230043651762e-07,
+      "loss": 0.5977,
+      "step": 17923
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.0494944598802696,
+      "learning_rate": 9.364265915796568e-07,
+      "loss": 0.612,
+      "step": 17924
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.0053848794495948,
+      "learning_rate": 9.36330179387474e-07,
+      "loss": 0.6426,
+      "step": 17925
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.635943441632832,
+      "learning_rate": 9.362337677895267e-07,
+      "loss": 0.6497,
+      "step": 17926
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.8731789456862042,
+      "learning_rate": 9.361373567867162e-07,
+      "loss": 0.6029,
+      "step": 17927
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.284774882854601,
+      "learning_rate": 9.360409463799405e-07,
+      "loss": 0.6204,
+      "step": 17928
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 4.167554068706813,
+      "learning_rate": 9.359445365701008e-07,
+      "loss": 0.6706,
+      "step": 17929
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.2422295723675942,
+      "learning_rate": 9.358481273580962e-07,
+      "loss": 0.6432,
+      "step": 17930
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.235961242552443,
+      "learning_rate": 9.357517187448269e-07,
+      "loss": 0.612,
+      "step": 17931
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.6824816031766345,
+      "learning_rate": 9.356553107311921e-07,
+      "loss": 0.6634,
+      "step": 17932
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.7324169553038757,
+      "learning_rate": 9.355589033180925e-07,
+      "loss": 0.6081,
+      "step": 17933
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.8005414737376215,
+      "learning_rate": 9.354624965064271e-07,
+      "loss": 0.6387,
+      "step": 17934
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.8706160393006965,
+      "learning_rate": 9.353660902970958e-07,
+      "loss": 0.6387,
+      "step": 17935
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.5320562298717237,
+      "learning_rate": 9.352696846909988e-07,
+      "loss": 0.7012,
+      "step": 17936
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.3929699293137516,
+      "learning_rate": 9.351732796890352e-07,
+      "loss": 0.6732,
+      "step": 17937
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.259898658805264,
+      "learning_rate": 9.350768752921051e-07,
+      "loss": 0.6549,
+      "step": 17938
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.262679274296862,
+      "learning_rate": 9.349804715011084e-07,
+      "loss": 0.6751,
+      "step": 17939
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.8926628065112086,
+      "learning_rate": 9.348840683169452e-07,
+      "loss": 0.6283,
+      "step": 17940
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.702144126721415,
+      "learning_rate": 9.347876657405142e-07,
+      "loss": 0.6595,
+      "step": 17941
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.164960573064039,
+      "learning_rate": 9.346912637727157e-07,
+      "loss": 0.6224,
+      "step": 17942
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.2736300943077765,
+      "learning_rate": 9.345948624144494e-07,
+      "loss": 0.6576,
+      "step": 17943
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.3819446142289338,
+      "learning_rate": 9.344984616666153e-07,
+      "loss": 0.6367,
+      "step": 17944
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.154448235733686,
+      "learning_rate": 9.344020615301125e-07,
+      "loss": 0.6309,
+      "step": 17945
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.1449646396881135,
+      "learning_rate": 9.343056620058412e-07,
+      "loss": 0.6439,
+      "step": 17946
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.152660914726504,
+      "learning_rate": 9.342092630947014e-07,
+      "loss": 0.6517,
+      "step": 17947
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.313531763633607,
+      "learning_rate": 9.341128647975918e-07,
+      "loss": 0.5957,
+      "step": 17948
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.2266285023645405,
+      "learning_rate": 9.340164671154127e-07,
+      "loss": 0.6523,
+      "step": 17949
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.9941357875333656,
+      "learning_rate": 9.339200700490639e-07,
+      "loss": 0.6367,
+      "step": 17950
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.081114350543267,
+      "learning_rate": 9.338236735994448e-07,
+      "loss": 0.6582,
+      "step": 17951
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.9381118504682657,
+      "learning_rate": 9.337272777674554e-07,
+      "loss": 0.6908,
+      "step": 17952
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.3869472512690875,
+      "learning_rate": 9.336308825539954e-07,
+      "loss": 0.6406,
+      "step": 17953
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.6342751720018214,
+      "learning_rate": 9.335344879599638e-07,
+      "loss": 0.6432,
+      "step": 17954
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.980417496182122,
+      "learning_rate": 9.33438093986261e-07,
+      "loss": 0.6029,
+      "step": 17955
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 6.846777874794096,
+      "learning_rate": 9.333417006337859e-07,
+      "loss": 0.6452,
+      "step": 17956
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.577459667922083,
+      "learning_rate": 9.33245307903439e-07,
+      "loss": 0.638,
+      "step": 17957
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.471219156056992,
+      "learning_rate": 9.331489157961193e-07,
+      "loss": 0.6882,
+      "step": 17958
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.2791512801997236,
+      "learning_rate": 9.330525243127267e-07,
+      "loss": 0.6862,
+      "step": 17959
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.609263640272435,
+      "learning_rate": 9.329561334541613e-07,
+      "loss": 0.6589,
+      "step": 17960
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.6544122761779336,
+      "learning_rate": 9.328597432213218e-07,
+      "loss": 0.6491,
+      "step": 17961
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.7071211269195015,
+      "learning_rate": 9.327633536151083e-07,
+      "loss": 0.6302,
+      "step": 17962
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.2685628263368756,
+      "learning_rate": 9.326669646364203e-07,
+      "loss": 0.6419,
+      "step": 17963
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.1086654715195268,
+      "learning_rate": 9.325705762861576e-07,
+      "loss": 0.6302,
+      "step": 17964
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.7191869747324495,
+      "learning_rate": 9.324741885652196e-07,
+      "loss": 0.6217,
+      "step": 17965
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.5857885598944086,
+      "learning_rate": 9.323778014745063e-07,
+      "loss": 0.6204,
+      "step": 17966
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.003493136819434,
+      "learning_rate": 9.322814150149165e-07,
+      "loss": 0.6576,
+      "step": 17967
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.129641355386616,
+      "learning_rate": 9.321850291873505e-07,
+      "loss": 0.6589,
+      "step": 17968
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.8794877160673105,
+      "learning_rate": 9.320886439927074e-07,
+      "loss": 0.6478,
+      "step": 17969
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.940416052150813,
+      "learning_rate": 9.31992259431887e-07,
+      "loss": 0.6289,
+      "step": 17970
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.2243418664620713,
+      "learning_rate": 9.318958755057891e-07,
+      "loss": 0.6855,
+      "step": 17971
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.657784684431376,
+      "learning_rate": 9.317994922153127e-07,
+      "loss": 0.6204,
+      "step": 17972
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.135419653212132,
+      "learning_rate": 9.317031095613581e-07,
+      "loss": 0.6322,
+      "step": 17973
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.858286084602687,
+      "learning_rate": 9.316067275448242e-07,
+      "loss": 0.6178,
+      "step": 17974
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.386941543302486,
+      "learning_rate": 9.315103461666107e-07,
+      "loss": 0.6439,
+      "step": 17975
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.4155622881101615,
+      "learning_rate": 9.314139654276173e-07,
+      "loss": 0.627,
+      "step": 17976
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.3195469758906926,
+      "learning_rate": 9.313175853287433e-07,
+      "loss": 0.5807,
+      "step": 17977
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.3401202060399107,
+      "learning_rate": 9.312212058708884e-07,
+      "loss": 0.6556,
+      "step": 17978
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.507279740327574,
+      "learning_rate": 9.31124827054952e-07,
+      "loss": 0.5996,
+      "step": 17979
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.9206732053731463,
+      "learning_rate": 9.310284488818341e-07,
+      "loss": 0.6471,
+      "step": 17980
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.162020746988836,
+      "learning_rate": 9.309320713524332e-07,
+      "loss": 0.6439,
+      "step": 17981
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.497321624638129,
+      "learning_rate": 9.308356944676498e-07,
+      "loss": 0.6165,
+      "step": 17982
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.6640400478407438,
+      "learning_rate": 9.307393182283826e-07,
+      "loss": 0.6634,
+      "step": 17983
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 5.627770530854225,
+      "learning_rate": 9.306429426355317e-07,
+      "loss": 0.6914,
+      "step": 17984
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.7148762472613392,
+      "learning_rate": 9.305465676899963e-07,
+      "loss": 0.5931,
+      "step": 17985
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.352004606790111,
+      "learning_rate": 9.304501933926762e-07,
+      "loss": 0.6491,
+      "step": 17986
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.9678014163915427,
+      "learning_rate": 9.303538197444705e-07,
+      "loss": 0.6003,
+      "step": 17987
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.778852105139559,
+      "learning_rate": 9.302574467462784e-07,
+      "loss": 0.6387,
+      "step": 17988
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.2254309576883635,
+      "learning_rate": 9.30161074399e-07,
+      "loss": 0.6165,
+      "step": 17989
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.4895795441831954,
+      "learning_rate": 9.300647027035342e-07,
+      "loss": 0.6576,
+      "step": 17990
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.806108763250136,
+      "learning_rate": 9.299683316607811e-07,
+      "loss": 0.6419,
+      "step": 17991
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 4.913244696576081,
+      "learning_rate": 9.298719612716394e-07,
+      "loss": 0.6208,
+      "step": 17992
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.334315766659914,
+      "learning_rate": 9.297755915370093e-07,
+      "loss": 0.5905,
+      "step": 17993
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.8563435055068394,
+      "learning_rate": 9.296792224577894e-07,
+      "loss": 0.6426,
+      "step": 17994
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.321481426874746,
+      "learning_rate": 9.295828540348798e-07,
+      "loss": 0.638,
+      "step": 17995
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.898467313533734,
+      "learning_rate": 9.294864862691795e-07,
+      "loss": 0.623,
+      "step": 17996
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.3124182990517186,
+      "learning_rate": 9.293901191615882e-07,
+      "loss": 0.6615,
+      "step": 17997
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.297376188451019,
+      "learning_rate": 9.29293752713005e-07,
+      "loss": 0.651,
+      "step": 17998
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.8189223503535326,
+      "learning_rate": 9.291973869243294e-07,
+      "loss": 0.6045,
+      "step": 17999
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.402453303602096,
+      "learning_rate": 9.291010217964614e-07,
+      "loss": 0.6654,
+      "step": 18000
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.9168477435806874,
+      "learning_rate": 9.290046573302992e-07,
+      "loss": 0.679,
+      "step": 18001
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 3.065307798660695,
+      "learning_rate": 9.289082935267431e-07,
+      "loss": 0.6243,
+      "step": 18002
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.241746510004309,
+      "learning_rate": 9.28811930386692e-07,
+      "loss": 0.6738,
+      "step": 18003
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.036108866630069,
+      "learning_rate": 9.287155679110455e-07,
+      "loss": 0.6087,
+      "step": 18004
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.5438684780584278,
+      "learning_rate": 9.28619206100703e-07,
+      "loss": 0.6426,
+      "step": 18005
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.1767122204471425,
+      "learning_rate": 9.285228449565641e-07,
+      "loss": 0.6126,
+      "step": 18006
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.454021456959534,
+      "learning_rate": 9.284264844795272e-07,
+      "loss": 0.6927,
+      "step": 18007
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.8591056868427787,
+      "learning_rate": 9.283301246704926e-07,
+      "loss": 0.6536,
+      "step": 18008
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.790307462584372,
+      "learning_rate": 9.28233765530359e-07,
+      "loss": 0.6393,
+      "step": 18009
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.292315391623623,
+      "learning_rate": 9.281374070600263e-07,
+      "loss": 0.6107,
+      "step": 18010
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.9419869460210433,
+      "learning_rate": 9.280410492603933e-07,
+      "loss": 0.571,
+      "step": 18011
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.2284458665410796,
+      "learning_rate": 9.279446921323598e-07,
+      "loss": 0.6003,
+      "step": 18012
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.794436196841769,
+      "learning_rate": 9.27848335676825e-07,
+      "loss": 0.6302,
+      "step": 18013
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.834054942308517,
+      "learning_rate": 9.277519798946877e-07,
+      "loss": 0.6068,
+      "step": 18014
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.283476705006508,
+      "learning_rate": 9.276556247868474e-07,
+      "loss": 0.6895,
+      "step": 18015
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.835622343599892,
+      "learning_rate": 9.275592703542038e-07,
+      "loss": 0.6465,
+      "step": 18016
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.787025415921265,
+      "learning_rate": 9.274629165976559e-07,
+      "loss": 0.6647,
+      "step": 18017
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.5561792648640687,
+      "learning_rate": 9.273665635181032e-07,
+      "loss": 0.6354,
+      "step": 18018
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.343225161130771,
+      "learning_rate": 9.272702111164449e-07,
+      "loss": 0.638,
+      "step": 18019
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.590874806146387,
+      "learning_rate": 9.271738593935798e-07,
+      "loss": 0.5983,
+      "step": 18020
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.7854330881626215,
+      "learning_rate": 9.270775083504076e-07,
+      "loss": 0.6823,
+      "step": 18021
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.3298819016287915,
+      "learning_rate": 9.269811579878274e-07,
+      "loss": 0.6178,
+      "step": 18022
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.285146187095742,
+      "learning_rate": 9.268848083067386e-07,
+      "loss": 0.6243,
+      "step": 18023
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.558716905422219,
+      "learning_rate": 9.267884593080403e-07,
+      "loss": 0.6576,
+      "step": 18024
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.554157902063901,
+      "learning_rate": 9.266921109926317e-07,
+      "loss": 0.6569,
+      "step": 18025
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.3461645005393095,
+      "learning_rate": 9.265957633614125e-07,
+      "loss": 0.6335,
+      "step": 18026
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.198368347609128,
+      "learning_rate": 9.264994164152812e-07,
+      "loss": 0.6126,
+      "step": 18027
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.631275926792149,
+      "learning_rate": 9.264030701551374e-07,
+      "loss": 0.6628,
+      "step": 18028
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.7973821333159785,
+      "learning_rate": 9.263067245818803e-07,
+      "loss": 0.6367,
+      "step": 18029
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.256748107921519,
+      "learning_rate": 9.262103796964089e-07,
+      "loss": 0.6882,
+      "step": 18030
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.400116745182738,
+      "learning_rate": 9.261140354996227e-07,
+      "loss": 0.6432,
+      "step": 18031
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.187619435489262,
+      "learning_rate": 9.260176919924207e-07,
+      "loss": 0.6243,
+      "step": 18032
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.1375429820026985,
+      "learning_rate": 9.259213491757023e-07,
+      "loss": 0.668,
+      "step": 18033
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.394875009228196,
+      "learning_rate": 9.258250070503665e-07,
+      "loss": 0.6419,
+      "step": 18034
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.456723966506128,
+      "learning_rate": 9.257286656173121e-07,
+      "loss": 0.6243,
+      "step": 18035
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.023989539618392,
+      "learning_rate": 9.256323248774387e-07,
+      "loss": 0.6413,
+      "step": 18036
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.312432145453305,
+      "learning_rate": 9.255359848316456e-07,
+      "loss": 0.666,
+      "step": 18037
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.987040490416904,
+      "learning_rate": 9.254396454808316e-07,
+      "loss": 0.7005,
+      "step": 18038
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.4231980089066054,
+      "learning_rate": 9.253433068258963e-07,
+      "loss": 0.6432,
+      "step": 18039
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 6.449758879997457,
+      "learning_rate": 9.252469688677383e-07,
+      "loss": 0.5911,
+      "step": 18040
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.485759839984833,
+      "learning_rate": 9.251506316072568e-07,
+      "loss": 0.6406,
+      "step": 18041
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.9487025746671565,
+      "learning_rate": 9.250542950453513e-07,
+      "loss": 0.6562,
+      "step": 18042
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.040826307637409,
+      "learning_rate": 9.249579591829204e-07,
+      "loss": 0.64,
+      "step": 18043
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.8754474390835725,
+      "learning_rate": 9.248616240208637e-07,
+      "loss": 0.6901,
+      "step": 18044
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.0290449534018244,
+      "learning_rate": 9.2476528956008e-07,
+      "loss": 0.5892,
+      "step": 18045
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.195195528049991,
+      "learning_rate": 9.24668955801469e-07,
+      "loss": 0.6497,
+      "step": 18046
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.41974518632869,
+      "learning_rate": 9.245726227459288e-07,
+      "loss": 0.6094,
+      "step": 18047
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.244353602469189,
+      "learning_rate": 9.24476290394359e-07,
+      "loss": 0.6706,
+      "step": 18048
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.5534971826449513,
+      "learning_rate": 9.243799587476586e-07,
+      "loss": 0.638,
+      "step": 18049
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.6944871461432722,
+      "learning_rate": 9.24283627806727e-07,
+      "loss": 0.6439,
+      "step": 18050
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.7942152281693735,
+      "learning_rate": 9.241872975724627e-07,
+      "loss": 0.6628,
+      "step": 18051
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.411337117430386,
+      "learning_rate": 9.240909680457655e-07,
+      "loss": 0.6309,
+      "step": 18052
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.835921040513171,
+      "learning_rate": 9.239946392275338e-07,
+      "loss": 0.5977,
+      "step": 18053
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 4.606073950359294,
+      "learning_rate": 9.238983111186666e-07,
+      "loss": 0.6517,
+      "step": 18054
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.806482225157005,
+      "learning_rate": 9.238019837200636e-07,
+      "loss": 0.6322,
+      "step": 18055
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.3977386381364383,
+      "learning_rate": 9.237056570326231e-07,
+      "loss": 0.6465,
+      "step": 18056
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.9600789207673834,
+      "learning_rate": 9.236093310572446e-07,
+      "loss": 0.6458,
+      "step": 18057
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.5440801140776035,
+      "learning_rate": 9.235130057948267e-07,
+      "loss": 0.6465,
+      "step": 18058
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 5.084828143781226,
+      "learning_rate": 9.234166812462693e-07,
+      "loss": 0.6107,
+      "step": 18059
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.7875494361233195,
+      "learning_rate": 9.233203574124702e-07,
+      "loss": 0.6497,
+      "step": 18060
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.3293363348092293,
+      "learning_rate": 9.232240342943293e-07,
+      "loss": 0.6562,
+      "step": 18061
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.9500748285821716,
+      "learning_rate": 9.231277118927449e-07,
+      "loss": 0.653,
+      "step": 18062
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.3997727234270303,
+      "learning_rate": 9.230313902086167e-07,
+      "loss": 0.6439,
+      "step": 18063
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.6985666422922927,
+      "learning_rate": 9.229350692428432e-07,
+      "loss": 0.599,
+      "step": 18064
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.271351495468433,
+      "learning_rate": 9.228387489963236e-07,
+      "loss": 0.6201,
+      "step": 18065
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.222094516013259,
+      "learning_rate": 9.227424294699571e-07,
+      "loss": 0.5833,
+      "step": 18066
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.1857829139387084,
+      "learning_rate": 9.226461106646418e-07,
+      "loss": 0.6048,
+      "step": 18067
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.409097702130692,
+      "learning_rate": 9.225497925812775e-07,
+      "loss": 0.6016,
+      "step": 18068
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.9279052027519845,
+      "learning_rate": 9.224534752207626e-07,
+      "loss": 0.6094,
+      "step": 18069
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.6157743143745575,
+      "learning_rate": 9.223571585839963e-07,
+      "loss": 0.6042,
+      "step": 18070
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.255708176434571,
+      "learning_rate": 9.222608426718776e-07,
+      "loss": 0.6484,
+      "step": 18071
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.101784906143681,
+      "learning_rate": 9.221645274853056e-07,
+      "loss": 0.6351,
+      "step": 18072
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.1833687985422006,
+      "learning_rate": 9.220682130251786e-07,
+      "loss": 0.6159,
+      "step": 18073
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.149633906044712,
+      "learning_rate": 9.219718992923959e-07,
+      "loss": 0.6582,
+      "step": 18074
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.5323546863918276,
+      "learning_rate": 9.218755862878563e-07,
+      "loss": 0.5758,
+      "step": 18075
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.584234949970057,
+      "learning_rate": 9.21779274012459e-07,
+      "loss": 0.6035,
+      "step": 18076
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.475165377837814,
+      "learning_rate": 9.216829624671023e-07,
+      "loss": 0.6517,
+      "step": 18077
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.202628077765428,
+      "learning_rate": 9.215866516526857e-07,
+      "loss": 0.6257,
+      "step": 18078
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.982071610648805,
+      "learning_rate": 9.214903415701082e-07,
+      "loss": 0.6536,
+      "step": 18079
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.3767023552644617,
+      "learning_rate": 9.213940322202677e-07,
+      "loss": 0.5859,
+      "step": 18080
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.276115188225346,
+      "learning_rate": 9.212977236040636e-07,
+      "loss": 0.6647,
+      "step": 18081
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.9692475065215307,
+      "learning_rate": 9.21201415722395e-07,
+      "loss": 0.6725,
+      "step": 18082
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.121839105491323,
+      "learning_rate": 9.211051085761606e-07,
+      "loss": 0.6348,
+      "step": 18083
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.0828693093895407,
+      "learning_rate": 9.210088021662592e-07,
+      "loss": 0.6576,
+      "step": 18084
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.361864343008302,
+      "learning_rate": 9.209124964935898e-07,
+      "loss": 0.6185,
+      "step": 18085
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.156013802551538,
+      "learning_rate": 9.208161915590508e-07,
+      "loss": 0.5967,
+      "step": 18086
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.4116929097419906,
+      "learning_rate": 9.207198873635413e-07,
+      "loss": 0.6673,
+      "step": 18087
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.2938241493108973,
+      "learning_rate": 9.206235839079602e-07,
+      "loss": 0.6198,
+      "step": 18088
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.090473543895346,
+      "learning_rate": 9.205272811932061e-07,
+      "loss": 0.6009,
+      "step": 18089
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.304391302460514,
+      "learning_rate": 9.204309792201779e-07,
+      "loss": 0.6146,
+      "step": 18090
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.437747530029953,
+      "learning_rate": 9.203346779897743e-07,
+      "loss": 0.666,
+      "step": 18091
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.11777780087901,
+      "learning_rate": 9.202383775028947e-07,
+      "loss": 0.6393,
+      "step": 18092
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.842866237233455,
+      "learning_rate": 9.201420777604372e-07,
+      "loss": 0.5794,
+      "step": 18093
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.1756350874005226,
+      "learning_rate": 9.200457787633004e-07,
+      "loss": 0.6289,
+      "step": 18094
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.900272923030467,
+      "learning_rate": 9.199494805123837e-07,
+      "loss": 0.6751,
+      "step": 18095
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.577942150165335,
+      "learning_rate": 9.198531830085854e-07,
+      "loss": 0.6003,
+      "step": 18096
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.038124141170987,
+      "learning_rate": 9.197568862528046e-07,
+      "loss": 0.6204,
+      "step": 18097
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.5737864916180584,
+      "learning_rate": 9.196605902459396e-07,
+      "loss": 0.653,
+      "step": 18098
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.536081426376318,
+      "learning_rate": 9.195642949888899e-07,
+      "loss": 0.6068,
+      "step": 18099
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.338119805673534,
+      "learning_rate": 9.194680004825536e-07,
+      "loss": 0.5768,
+      "step": 18100
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.353644032924196,
+      "learning_rate": 9.193717067278293e-07,
+      "loss": 0.6185,
+      "step": 18101
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.289447331863877,
+      "learning_rate": 9.192754137256161e-07,
+      "loss": 0.6836,
+      "step": 18102
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.247450216646051,
+      "learning_rate": 9.191791214768125e-07,
+      "loss": 0.5944,
+      "step": 18103
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.456305846015594,
+      "learning_rate": 9.190828299823173e-07,
+      "loss": 0.6484,
+      "step": 18104
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.319535245853474,
+      "learning_rate": 9.189865392430297e-07,
+      "loss": 0.6738,
+      "step": 18105
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.987303812912631,
+      "learning_rate": 9.188902492598476e-07,
+      "loss": 0.6458,
+      "step": 18106
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.3234308052936163,
+      "learning_rate": 9.187939600336698e-07,
+      "loss": 0.6615,
+      "step": 18107
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.3824335313595437,
+      "learning_rate": 9.186976715653953e-07,
+      "loss": 0.6667,
+      "step": 18108
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.314077228820131,
+      "learning_rate": 9.186013838559226e-07,
+      "loss": 0.61,
+      "step": 18109
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.598522306658079,
+      "learning_rate": 9.185050969061504e-07,
+      "loss": 0.6641,
+      "step": 18110
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.707597025464148,
+      "learning_rate": 9.184088107169771e-07,
+      "loss": 0.6022,
+      "step": 18111
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.118504224443284,
+      "learning_rate": 9.183125252893022e-07,
+      "loss": 0.653,
+      "step": 18112
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.211124553787781,
+      "learning_rate": 9.182162406240231e-07,
+      "loss": 0.6152,
+      "step": 18113
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.893839408759137,
+      "learning_rate": 9.181199567220395e-07,
+      "loss": 0.6465,
+      "step": 18114
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.2693068104817264,
+      "learning_rate": 9.180236735842492e-07,
+      "loss": 0.6836,
+      "step": 18115
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.771089442246397,
+      "learning_rate": 9.179273912115514e-07,
+      "loss": 0.6458,
+      "step": 18116
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.602721098292529,
+      "learning_rate": 9.178311096048443e-07,
+      "loss": 0.64,
+      "step": 18117
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.164301917450453,
+      "learning_rate": 9.177348287650273e-07,
+      "loss": 0.6419,
+      "step": 18118
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 6.488285155826243,
+      "learning_rate": 9.176385486929981e-07,
+      "loss": 0.6686,
+      "step": 18119
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.426274496421779,
+      "learning_rate": 9.175422693896553e-07,
+      "loss": 0.6758,
+      "step": 18120
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.7456599006762348,
+      "learning_rate": 9.174459908558982e-07,
+      "loss": 0.681,
+      "step": 18121
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.900765787659097,
+      "learning_rate": 9.173497130926245e-07,
+      "loss": 0.6667,
+      "step": 18122
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.8494846789310806,
+      "learning_rate": 9.172534361007337e-07,
+      "loss": 0.6523,
+      "step": 18123
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.100763448607279,
+      "learning_rate": 9.171571598811236e-07,
+      "loss": 0.6406,
+      "step": 18124
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.8232119490901284,
+      "learning_rate": 9.170608844346935e-07,
+      "loss": 0.6055,
+      "step": 18125
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.587911868924834,
+      "learning_rate": 9.16964609762341e-07,
+      "loss": 0.6419,
+      "step": 18126
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.13627365525406,
+      "learning_rate": 9.168683358649653e-07,
+      "loss": 0.6426,
+      "step": 18127
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.367491518039062,
+      "learning_rate": 9.167720627434648e-07,
+      "loss": 0.6302,
+      "step": 18128
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.0931317387437325,
+      "learning_rate": 9.166757903987381e-07,
+      "loss": 0.6771,
+      "step": 18129
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.924271670126176,
+      "learning_rate": 9.165795188316833e-07,
+      "loss": 0.6335,
+      "step": 18130
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.020027402081508,
+      "learning_rate": 9.164832480431996e-07,
+      "loss": 0.6322,
+      "step": 18131
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.562092251016536,
+      "learning_rate": 9.163869780341852e-07,
+      "loss": 0.6289,
+      "step": 18132
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.185989782403527,
+      "learning_rate": 9.162907088055382e-07,
+      "loss": 0.6172,
+      "step": 18133
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.18674756645961,
+      "learning_rate": 9.161944403581575e-07,
+      "loss": 0.6257,
+      "step": 18134
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 5.788869352507617,
+      "learning_rate": 9.160981726929414e-07,
+      "loss": 0.61,
+      "step": 18135
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 4.77737002447882,
+      "learning_rate": 9.160019058107885e-07,
+      "loss": 0.5944,
+      "step": 18136
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 3.904818864654025,
+      "learning_rate": 9.159056397125974e-07,
+      "loss": 0.6296,
+      "step": 18137
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.817070019309717,
+      "learning_rate": 9.158093743992667e-07,
+      "loss": 0.6061,
+      "step": 18138
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.8597253107620135,
+      "learning_rate": 9.15713109871694e-07,
+      "loss": 0.6126,
+      "step": 18139
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.337682796214984,
+      "learning_rate": 9.156168461307785e-07,
+      "loss": 0.6615,
+      "step": 18140
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.622732825150123,
+      "learning_rate": 9.155205831774183e-07,
+      "loss": 0.6423,
+      "step": 18141
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.329592685537281,
+      "learning_rate": 9.154243210125122e-07,
+      "loss": 0.6237,
+      "step": 18142
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.923177347317991,
+      "learning_rate": 9.153280596369582e-07,
+      "loss": 0.6536,
+      "step": 18143
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.8004037282501497,
+      "learning_rate": 9.15231799051655e-07,
+      "loss": 0.653,
+      "step": 18144
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.8290465635874837,
+      "learning_rate": 9.151355392575012e-07,
+      "loss": 0.599,
+      "step": 18145
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.3425841918446375,
+      "learning_rate": 9.150392802553945e-07,
+      "loss": 0.6419,
+      "step": 18146
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.416536044510755,
+      "learning_rate": 9.149430220462337e-07,
+      "loss": 0.5964,
+      "step": 18147
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.054075125724332,
+      "learning_rate": 9.148467646309175e-07,
+      "loss": 0.6133,
+      "step": 18148
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.0053941372707556,
+      "learning_rate": 9.147505080103436e-07,
+      "loss": 0.6243,
+      "step": 18149
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.644271665627855,
+      "learning_rate": 9.14654252185411e-07,
+      "loss": 0.6188,
+      "step": 18150
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.3604381988282395,
+      "learning_rate": 9.145579971570177e-07,
+      "loss": 0.6452,
+      "step": 18151
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.321865098532106,
+      "learning_rate": 9.144617429260625e-07,
+      "loss": 0.6341,
+      "step": 18152
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.693569344972487,
+      "learning_rate": 9.143654894934433e-07,
+      "loss": 0.6647,
+      "step": 18153
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.4457999363389575,
+      "learning_rate": 9.142692368600582e-07,
+      "loss": 0.6569,
+      "step": 18154
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.643255321809023,
+      "learning_rate": 9.141729850268062e-07,
+      "loss": 0.6426,
+      "step": 18155
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.115093180476964,
+      "learning_rate": 9.140767339945851e-07,
+      "loss": 0.5957,
+      "step": 18156
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.090527448017933,
+      "learning_rate": 9.139804837642935e-07,
+      "loss": 0.6673,
+      "step": 18157
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.667164766847176,
+      "learning_rate": 9.138842343368299e-07,
+      "loss": 0.6165,
+      "step": 18158
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 7.128525209553047,
+      "learning_rate": 9.137879857130923e-07,
+      "loss": 0.6178,
+      "step": 18159
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.8852586912674747,
+      "learning_rate": 9.136917378939788e-07,
+      "loss": 0.6543,
+      "step": 18160
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.481129712284533,
+      "learning_rate": 9.135954908803882e-07,
+      "loss": 0.6126,
+      "step": 18161
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.7743357282892798,
+      "learning_rate": 9.134992446732182e-07,
+      "loss": 0.6374,
+      "step": 18162
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.7360800423710994,
+      "learning_rate": 9.134029992733677e-07,
+      "loss": 0.6445,
+      "step": 18163
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.2022180831310956,
+      "learning_rate": 9.133067546817344e-07,
+      "loss": 0.6068,
+      "step": 18164
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.178119365361533,
+      "learning_rate": 9.132105108992173e-07,
+      "loss": 0.571,
+      "step": 18165
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.639202277281453,
+      "learning_rate": 9.131142679267139e-07,
+      "loss": 0.6152,
+      "step": 18166
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.5987556872988793,
+      "learning_rate": 9.130180257651227e-07,
+      "loss": 0.6133,
+      "step": 18167
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.1258757705675695,
+      "learning_rate": 9.129217844153418e-07,
+      "loss": 0.6348,
+      "step": 18168
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.908474067960074,
+      "learning_rate": 9.128255438782697e-07,
+      "loss": 0.6849,
+      "step": 18169
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 5.789175521778792,
+      "learning_rate": 9.127293041548044e-07,
+      "loss": 0.7012,
+      "step": 18170
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.7318750119191995,
+      "learning_rate": 9.126330652458447e-07,
+      "loss": 0.6471,
+      "step": 18171
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.26274793616103,
+      "learning_rate": 9.125368271522881e-07,
+      "loss": 0.6204,
+      "step": 18172
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.004910380569,
+      "learning_rate": 9.124405898750328e-07,
+      "loss": 0.6136,
+      "step": 18173
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.636443585673921,
+      "learning_rate": 9.123443534149774e-07,
+      "loss": 0.6432,
+      "step": 18174
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.125385836932116,
+      "learning_rate": 9.122481177730197e-07,
+      "loss": 0.6621,
+      "step": 18175
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.8713513844104495,
+      "learning_rate": 9.121518829500583e-07,
+      "loss": 0.681,
+      "step": 18176
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.215554464610303,
+      "learning_rate": 9.12055648946991e-07,
+      "loss": 0.7135,
+      "step": 18177
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.7573237018756735,
+      "learning_rate": 9.119594157647163e-07,
+      "loss": 0.6185,
+      "step": 18178
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.142019555930571,
+      "learning_rate": 9.118631834041319e-07,
+      "loss": 0.668,
+      "step": 18179
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.9007288060918444,
+      "learning_rate": 9.117669518661365e-07,
+      "loss": 0.5638,
+      "step": 18180
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.3637255217504616,
+      "learning_rate": 9.116707211516276e-07,
+      "loss": 0.5745,
+      "step": 18181
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.5831745815936995,
+      "learning_rate": 9.115744912615039e-07,
+      "loss": 0.625,
+      "step": 18182
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.619856969729984,
+      "learning_rate": 9.114782621966632e-07,
+      "loss": 0.6751,
+      "step": 18183
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.462447573823886,
+      "learning_rate": 9.113820339580037e-07,
+      "loss": 0.6432,
+      "step": 18184
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.2524602129399325,
+      "learning_rate": 9.11285806546424e-07,
+      "loss": 0.6335,
+      "step": 18185
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.701709670679784,
+      "learning_rate": 9.111895799628211e-07,
+      "loss": 0.64,
+      "step": 18186
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.2207303736701736,
+      "learning_rate": 9.110933542080938e-07,
+      "loss": 0.6517,
+      "step": 18187
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.406136711411864,
+      "learning_rate": 9.109971292831401e-07,
+      "loss": 0.6152,
+      "step": 18188
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.081087191467893,
+      "learning_rate": 9.109009051888582e-07,
+      "loss": 0.6439,
+      "step": 18189
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.678648088870373,
+      "learning_rate": 9.108046819261459e-07,
+      "loss": 0.5983,
+      "step": 18190
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 4.802124775078641,
+      "learning_rate": 9.107084594959017e-07,
+      "loss": 0.6406,
+      "step": 18191
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.558880195831095,
+      "learning_rate": 9.10612237899023e-07,
+      "loss": 0.6445,
+      "step": 18192
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.675063016071835,
+      "learning_rate": 9.105160171364084e-07,
+      "loss": 0.6406,
+      "step": 18193
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.348777888472594,
+      "learning_rate": 9.104197972089553e-07,
+      "loss": 0.6087,
+      "step": 18194
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.1344093629295453,
+      "learning_rate": 9.103235781175626e-07,
+      "loss": 0.6882,
+      "step": 18195
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.1443358381636792,
+      "learning_rate": 9.102273598631275e-07,
+      "loss": 0.6497,
+      "step": 18196
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.0421079517166234,
+      "learning_rate": 9.101311424465489e-07,
+      "loss": 0.625,
+      "step": 18197
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.380269227505805,
+      "learning_rate": 9.100349258687242e-07,
+      "loss": 0.64,
+      "step": 18198
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.4791889332117827,
+      "learning_rate": 9.099387101305513e-07,
+      "loss": 0.627,
+      "step": 18199
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.1378182679735467,
+      "learning_rate": 9.098424952329284e-07,
+      "loss": 0.6699,
+      "step": 18200
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.1859077755375473,
+      "learning_rate": 9.097462811767533e-07,
+      "loss": 0.64,
+      "step": 18201
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.1424853905518177,
+      "learning_rate": 9.096500679629242e-07,
+      "loss": 0.5957,
+      "step": 18202
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 6.337498457470772,
+      "learning_rate": 9.095538555923392e-07,
+      "loss": 0.6185,
+      "step": 18203
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.225174939089237,
+      "learning_rate": 9.094576440658963e-07,
+      "loss": 0.6322,
+      "step": 18204
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.6184353598153765,
+      "learning_rate": 9.093614333844927e-07,
+      "loss": 0.6257,
+      "step": 18205
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.569148673328583,
+      "learning_rate": 9.092652235490271e-07,
+      "loss": 0.6406,
+      "step": 18206
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.415467863135101,
+      "learning_rate": 9.09169014560397e-07,
+      "loss": 0.6146,
+      "step": 18207
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.234231927589386,
+      "learning_rate": 9.090728064195008e-07,
+      "loss": 0.6445,
+      "step": 18208
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.321810834254782,
+      "learning_rate": 9.089765991272358e-07,
+      "loss": 0.6165,
+      "step": 18209
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.9990713325767775,
+      "learning_rate": 9.088803926845006e-07,
+      "loss": 0.6432,
+      "step": 18210
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.3240009919678175,
+      "learning_rate": 9.087841870921929e-07,
+      "loss": 0.6185,
+      "step": 18211
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.027374764297422,
+      "learning_rate": 9.086879823512099e-07,
+      "loss": 0.6816,
+      "step": 18212
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.302208783460543,
+      "learning_rate": 9.085917784624501e-07,
+      "loss": 0.6289,
+      "step": 18213
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.266054125638214,
+      "learning_rate": 9.084955754268116e-07,
+      "loss": 0.638,
+      "step": 18214
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.733800327276837,
+      "learning_rate": 9.083993732451918e-07,
+      "loss": 0.612,
+      "step": 18215
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.344563084735018,
+      "learning_rate": 9.08303171918489e-07,
+      "loss": 0.6589,
+      "step": 18216
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.613574430509282,
+      "learning_rate": 9.082069714476005e-07,
+      "loss": 0.6504,
+      "step": 18217
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.375606424129696,
+      "learning_rate": 9.081107718334249e-07,
+      "loss": 0.638,
+      "step": 18218
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.565006351632212,
+      "learning_rate": 9.080145730768594e-07,
+      "loss": 0.6673,
+      "step": 18219
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.571497743861842,
+      "learning_rate": 9.079183751788016e-07,
+      "loss": 0.6842,
+      "step": 18220
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.657974958944332,
+      "learning_rate": 9.078221781401503e-07,
+      "loss": 0.5996,
+      "step": 18221
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.00651032680078,
+      "learning_rate": 9.077259819618022e-07,
+      "loss": 0.6354,
+      "step": 18222
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.0661068449665345,
+      "learning_rate": 9.07629786644656e-07,
+      "loss": 0.6445,
+      "step": 18223
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.806421820101088,
+      "learning_rate": 9.075335921896095e-07,
+      "loss": 0.6439,
+      "step": 18224
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.0669652329281725,
+      "learning_rate": 9.074373985975598e-07,
+      "loss": 0.6602,
+      "step": 18225
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.709747493012315,
+      "learning_rate": 9.073412058694049e-07,
+      "loss": 0.6882,
+      "step": 18226
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.7147973048511895,
+      "learning_rate": 9.07245014006043e-07,
+      "loss": 0.6465,
+      "step": 18227
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.166982526552127,
+      "learning_rate": 9.071488230083714e-07,
+      "loss": 0.6081,
+      "step": 18228
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 6.536468816342005,
+      "learning_rate": 9.070526328772882e-07,
+      "loss": 0.6673,
+      "step": 18229
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.9503490236893914,
+      "learning_rate": 9.069564436136908e-07,
+      "loss": 0.6035,
+      "step": 18230
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.383141848223786,
+      "learning_rate": 9.068602552184775e-07,
+      "loss": 0.6686,
+      "step": 18231
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.3928224391650943,
+      "learning_rate": 9.067640676925455e-07,
+      "loss": 0.6458,
+      "step": 18232
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.962909427310737,
+      "learning_rate": 9.066678810367926e-07,
+      "loss": 0.6211,
+      "step": 18233
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.146912240721751,
+      "learning_rate": 9.065716952521166e-07,
+      "loss": 0.6432,
+      "step": 18234
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.6073605185959785,
+      "learning_rate": 9.064755103394152e-07,
+      "loss": 0.6419,
+      "step": 18235
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.649165239828051,
+      "learning_rate": 9.063793262995861e-07,
+      "loss": 0.6582,
+      "step": 18236
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.5984541770498772,
+      "learning_rate": 9.062831431335275e-07,
+      "loss": 0.6224,
+      "step": 18237
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.77731401218629,
+      "learning_rate": 9.061869608421363e-07,
+      "loss": 0.6673,
+      "step": 18238
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.170561744046858,
+      "learning_rate": 9.060907794263103e-07,
+      "loss": 0.6172,
+      "step": 18239
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.9383492485525102,
+      "learning_rate": 9.059945988869477e-07,
+      "loss": 0.6445,
+      "step": 18240
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.0324132265249277,
+      "learning_rate": 9.058984192249454e-07,
+      "loss": 0.6745,
+      "step": 18241
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.355677555794187,
+      "learning_rate": 9.058022404412018e-07,
+      "loss": 0.64,
+      "step": 18242
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.3797492617247573,
+      "learning_rate": 9.057060625366139e-07,
+      "loss": 0.6706,
+      "step": 18243
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.0737314031003073,
+      "learning_rate": 9.056098855120803e-07,
+      "loss": 0.6621,
+      "step": 18244
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 5.8506019913576655,
+      "learning_rate": 9.055137093684977e-07,
+      "loss": 0.6569,
+      "step": 18245
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.8372179401823674,
+      "learning_rate": 9.054175341067637e-07,
+      "loss": 0.6178,
+      "step": 18246
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.560056148561132,
+      "learning_rate": 9.053213597277762e-07,
+      "loss": 0.6113,
+      "step": 18247
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.5691756955292435,
+      "learning_rate": 9.052251862324332e-07,
+      "loss": 0.6374,
+      "step": 18248
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.9247914459582716,
+      "learning_rate": 9.051290136216315e-07,
+      "loss": 0.6237,
+      "step": 18249
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.08931870798089,
+      "learning_rate": 9.050328418962695e-07,
+      "loss": 0.6432,
+      "step": 18250
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.8199042946239565,
+      "learning_rate": 9.049366710572444e-07,
+      "loss": 0.6608,
+      "step": 18251
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.5109587764281818,
+      "learning_rate": 9.048405011054535e-07,
+      "loss": 0.6348,
+      "step": 18252
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.8834622721756746,
+      "learning_rate": 9.047443320417949e-07,
+      "loss": 0.6309,
+      "step": 18253
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.402150723466513,
+      "learning_rate": 9.046481638671654e-07,
+      "loss": 0.6335,
+      "step": 18254
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.5573506279055027,
+      "learning_rate": 9.045519965824635e-07,
+      "loss": 0.6302,
+      "step": 18255
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.8091121706845215,
+      "learning_rate": 9.044558301885858e-07,
+      "loss": 0.6582,
+      "step": 18256
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.344345546376157,
+      "learning_rate": 9.043596646864309e-07,
+      "loss": 0.6491,
+      "step": 18257
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 7.245976216688334,
+      "learning_rate": 9.042635000768953e-07,
+      "loss": 0.64,
+      "step": 18258
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.192767834778094,
+      "learning_rate": 9.04167336360877e-07,
+      "loss": 0.6374,
+      "step": 18259
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.444611525074938,
+      "learning_rate": 9.040711735392731e-07,
+      "loss": 0.6543,
+      "step": 18260
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.317623310956304,
+      "learning_rate": 9.03975011612982e-07,
+      "loss": 0.6471,
+      "step": 18261
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.114466597548727,
+      "learning_rate": 9.038788505829001e-07,
+      "loss": 0.6426,
+      "step": 18262
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.7349117847232645,
+      "learning_rate": 9.037826904499257e-07,
+      "loss": 0.5853,
+      "step": 18263
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.531006550600416,
+      "learning_rate": 9.036865312149562e-07,
+      "loss": 0.6237,
+      "step": 18264
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.939798796801863,
+      "learning_rate": 9.035903728788883e-07,
+      "loss": 0.6393,
+      "step": 18265
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.1851847881281374,
+      "learning_rate": 9.034942154426202e-07,
+      "loss": 0.6549,
+      "step": 18266
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.5954291423902442,
+      "learning_rate": 9.03398058907049e-07,
+      "loss": 0.6641,
+      "step": 18267
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.779779299860197,
+      "learning_rate": 9.033019032730722e-07,
+      "loss": 0.6829,
+      "step": 18268
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.079425380793196,
+      "learning_rate": 9.032057485415876e-07,
+      "loss": 0.6406,
+      "step": 18269
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 4.9276323744324975,
+      "learning_rate": 9.031095947134923e-07,
+      "loss": 0.6204,
+      "step": 18270
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.100011079752312,
+      "learning_rate": 9.030134417896835e-07,
+      "loss": 0.625,
+      "step": 18271
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.9304244266846875,
+      "learning_rate": 9.02917289771059e-07,
+      "loss": 0.6022,
+      "step": 18272
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.471118484985715,
+      "learning_rate": 9.028211386585157e-07,
+      "loss": 0.6169,
+      "step": 18273
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.7145839080249985,
+      "learning_rate": 9.027249884529517e-07,
+      "loss": 0.7168,
+      "step": 18274
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.713954233976205,
+      "learning_rate": 9.026288391552636e-07,
+      "loss": 0.6061,
+      "step": 18275
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.8361783635973787,
+      "learning_rate": 9.025326907663495e-07,
+      "loss": 0.5934,
+      "step": 18276
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.4805158342506153,
+      "learning_rate": 9.024365432871065e-07,
+      "loss": 0.6087,
+      "step": 18277
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.373653373642267,
+      "learning_rate": 9.023403967184316e-07,
+      "loss": 0.6055,
+      "step": 18278
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.245473197015915,
+      "learning_rate": 9.022442510612223e-07,
+      "loss": 0.596,
+      "step": 18279
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.212364415117663,
+      "learning_rate": 9.021481063163763e-07,
+      "loss": 0.6426,
+      "step": 18280
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.00520093837294,
+      "learning_rate": 9.020519624847904e-07,
+      "loss": 0.6211,
+      "step": 18281
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.633585047085869,
+      "learning_rate": 9.019558195673624e-07,
+      "loss": 0.6777,
+      "step": 18282
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.5339583759892776,
+      "learning_rate": 9.018596775649893e-07,
+      "loss": 0.6178,
+      "step": 18283
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.6081059504282535,
+      "learning_rate": 9.017635364785689e-07,
+      "loss": 0.6712,
+      "step": 18284
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.6974316850431315,
+      "learning_rate": 9.016673963089979e-07,
+      "loss": 0.668,
+      "step": 18285
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.4821231514968973,
+      "learning_rate": 9.015712570571736e-07,
+      "loss": 0.6478,
+      "step": 18286
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.322925927184762,
+      "learning_rate": 9.014751187239935e-07,
+      "loss": 0.5885,
+      "step": 18287
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.641770734849863,
+      "learning_rate": 9.013789813103548e-07,
+      "loss": 0.6374,
+      "step": 18288
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.5373466829724673,
+      "learning_rate": 9.012828448171548e-07,
+      "loss": 0.5957,
+      "step": 18289
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.2793686490578473,
+      "learning_rate": 9.01186709245291e-07,
+      "loss": 0.6276,
+      "step": 18290
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.272035222282402,
+      "learning_rate": 9.010905745956603e-07,
+      "loss": 0.6361,
+      "step": 18291
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.777644078013765,
+      "learning_rate": 9.009944408691598e-07,
+      "loss": 0.6608,
+      "step": 18292
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.4727318435254015,
+      "learning_rate": 9.008983080666871e-07,
+      "loss": 0.6654,
+      "step": 18293
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.4289742299333685,
+      "learning_rate": 9.008021761891391e-07,
+      "loss": 0.6673,
+      "step": 18294
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.993340659464121,
+      "learning_rate": 9.007060452374133e-07,
+      "loss": 0.6491,
+      "step": 18295
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.4621049365301286,
+      "learning_rate": 9.006099152124066e-07,
+      "loss": 0.6413,
+      "step": 18296
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.262285739585931,
+      "learning_rate": 9.005137861150167e-07,
+      "loss": 0.6543,
+      "step": 18297
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.579014670311217,
+      "learning_rate": 9.004176579461403e-07,
+      "loss": 0.679,
+      "step": 18298
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.2663453864725907,
+      "learning_rate": 9.003215307066744e-07,
+      "loss": 0.6387,
+      "step": 18299
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.893096549780903,
+      "learning_rate": 9.002254043975166e-07,
+      "loss": 0.6361,
+      "step": 18300
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.5988253776861496,
+      "learning_rate": 9.001292790195641e-07,
+      "loss": 0.6549,
+      "step": 18301
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.6140026359701243,
+      "learning_rate": 9.000331545737137e-07,
+      "loss": 0.6328,
+      "step": 18302
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.1879751033492885,
+      "learning_rate": 8.99937031060863e-07,
+      "loss": 0.6426,
+      "step": 18303
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.229040838679359,
+      "learning_rate": 8.998409084819087e-07,
+      "loss": 0.6595,
+      "step": 18304
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.132162634028491,
+      "learning_rate": 8.997447868377479e-07,
+      "loss": 0.6055,
+      "step": 18305
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.94857299648251,
+      "learning_rate": 8.99648666129278e-07,
+      "loss": 0.6693,
+      "step": 18306
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.2186271678689335,
+      "learning_rate": 8.995525463573958e-07,
+      "loss": 0.668,
+      "step": 18307
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.7408961024761997,
+      "learning_rate": 8.994564275229987e-07,
+      "loss": 0.6549,
+      "step": 18308
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.980323764778065,
+      "learning_rate": 8.993603096269835e-07,
+      "loss": 0.6051,
+      "step": 18309
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.0455746925399816,
+      "learning_rate": 8.992641926702479e-07,
+      "loss": 0.625,
+      "step": 18310
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.103277198651808,
+      "learning_rate": 8.991680766536883e-07,
+      "loss": 0.6426,
+      "step": 18311
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.6793503238865677,
+      "learning_rate": 8.990719615782017e-07,
+      "loss": 0.6673,
+      "step": 18312
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.048104260263636,
+      "learning_rate": 8.989758474446854e-07,
+      "loss": 0.5573,
+      "step": 18313
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.190087094799505,
+      "learning_rate": 8.988797342540367e-07,
+      "loss": 0.5684,
+      "step": 18314
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.330680909548514,
+      "learning_rate": 8.987836220071521e-07,
+      "loss": 0.6439,
+      "step": 18315
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.0759624238472707,
+      "learning_rate": 8.986875107049291e-07,
+      "loss": 0.6582,
+      "step": 18316
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 6.154466164293348,
+      "learning_rate": 8.985914003482649e-07,
+      "loss": 0.6309,
+      "step": 18317
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.4894649929296278,
+      "learning_rate": 8.984952909380556e-07,
+      "loss": 0.6348,
+      "step": 18318
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.8116778625582866,
+      "learning_rate": 8.983991824751989e-07,
+      "loss": 0.6536,
+      "step": 18319
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.6050346369517396,
+      "learning_rate": 8.983030749605914e-07,
+      "loss": 0.6361,
+      "step": 18320
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.2777133397674736,
+      "learning_rate": 8.982069683951305e-07,
+      "loss": 0.6432,
+      "step": 18321
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.357500808205302,
+      "learning_rate": 8.981108627797128e-07,
+      "loss": 0.6829,
+      "step": 18322
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.1593275315971057,
+      "learning_rate": 8.980147581152358e-07,
+      "loss": 0.5996,
+      "step": 18323
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.420034320980556,
+      "learning_rate": 8.979186544025958e-07,
+      "loss": 0.6051,
+      "step": 18324
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.169487310625603,
+      "learning_rate": 8.978225516426901e-07,
+      "loss": 0.6901,
+      "step": 18325
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.2177401927880607,
+      "learning_rate": 8.977264498364154e-07,
+      "loss": 0.624,
+      "step": 18326
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.725892455633093,
+      "learning_rate": 8.976303489846691e-07,
+      "loss": 0.6406,
+      "step": 18327
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.852582008416363,
+      "learning_rate": 8.975342490883474e-07,
+      "loss": 0.6543,
+      "step": 18328
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.07905801995058,
+      "learning_rate": 8.974381501483478e-07,
+      "loss": 0.6719,
+      "step": 18329
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.032940411388419,
+      "learning_rate": 8.973420521655675e-07,
+      "loss": 0.6628,
+      "step": 18330
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.085900047650812,
+      "learning_rate": 8.972459551409022e-07,
+      "loss": 0.666,
+      "step": 18331
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 4.399569325361978,
+      "learning_rate": 8.9714985907525e-07,
+      "loss": 0.6445,
+      "step": 18332
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.455195528684534,
+      "learning_rate": 8.970537639695069e-07,
+      "loss": 0.6452,
+      "step": 18333
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.946637993197398,
+      "learning_rate": 8.9695766982457e-07,
+      "loss": 0.6198,
+      "step": 18334
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.282500091597273,
+      "learning_rate": 8.968615766413365e-07,
+      "loss": 0.6439,
+      "step": 18335
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.9340864960526156,
+      "learning_rate": 8.96765484420703e-07,
+      "loss": 0.6165,
+      "step": 18336
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.437993685718336,
+      "learning_rate": 8.966693931635667e-07,
+      "loss": 0.6471,
+      "step": 18337
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 3.30635302216227,
+      "learning_rate": 8.965733028708239e-07,
+      "loss": 0.6042,
+      "step": 18338
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.947039192286493,
+      "learning_rate": 8.964772135433714e-07,
+      "loss": 0.6159,
+      "step": 18339
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.5708646540631745,
+      "learning_rate": 8.963811251821063e-07,
+      "loss": 0.64,
+      "step": 18340
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.1587144455410963,
+      "learning_rate": 8.962850377879252e-07,
+      "loss": 0.653,
+      "step": 18341
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.2409733794469044,
+      "learning_rate": 8.961889513617254e-07,
+      "loss": 0.638,
+      "step": 18342
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.3086003552791876,
+      "learning_rate": 8.960928659044033e-07,
+      "loss": 0.6191,
+      "step": 18343
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.0064934091326214,
+      "learning_rate": 8.959967814168552e-07,
+      "loss": 0.6074,
+      "step": 18344
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.750931041226108,
+      "learning_rate": 8.959006978999785e-07,
+      "loss": 0.6882,
+      "step": 18345
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.4804693665363517,
+      "learning_rate": 8.958046153546699e-07,
+      "loss": 0.6061,
+      "step": 18346
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.083039159243666,
+      "learning_rate": 8.957085337818259e-07,
+      "loss": 0.6439,
+      "step": 18347
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.89756170114702,
+      "learning_rate": 8.956124531823436e-07,
+      "loss": 0.6621,
+      "step": 18348
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.9361164400979285,
+      "learning_rate": 8.955163735571191e-07,
+      "loss": 0.6497,
+      "step": 18349
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.184556256738686,
+      "learning_rate": 8.954202949070502e-07,
+      "loss": 0.6217,
+      "step": 18350
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.768781323002286,
+      "learning_rate": 8.953242172330326e-07,
+      "loss": 0.6771,
+      "step": 18351
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.0750965952911264,
+      "learning_rate": 8.952281405359632e-07,
+      "loss": 0.6361,
+      "step": 18352
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.5535696855202428,
+      "learning_rate": 8.95132064816739e-07,
+      "loss": 0.6406,
+      "step": 18353
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.849575296640186,
+      "learning_rate": 8.950359900762563e-07,
+      "loss": 0.6198,
+      "step": 18354
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.0160582624217955,
+      "learning_rate": 8.949399163154118e-07,
+      "loss": 0.627,
+      "step": 18355
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.1526210370464187,
+      "learning_rate": 8.948438435351031e-07,
+      "loss": 0.6165,
+      "step": 18356
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 6.108876419107641,
+      "learning_rate": 8.947477717362257e-07,
+      "loss": 0.6178,
+      "step": 18357
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.612966571757245,
+      "learning_rate": 8.946517009196766e-07,
+      "loss": 0.6543,
+      "step": 18358
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.425421339371662,
+      "learning_rate": 8.945556310863524e-07,
+      "loss": 0.6426,
+      "step": 18359
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.265562422977711,
+      "learning_rate": 8.944595622371499e-07,
+      "loss": 0.5706,
+      "step": 18360
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.4561925671884017,
+      "learning_rate": 8.943634943729658e-07,
+      "loss": 0.6126,
+      "step": 18361
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.8262187676586255,
+      "learning_rate": 8.942674274946963e-07,
+      "loss": 0.6152,
+      "step": 18362
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.803348773623065,
+      "learning_rate": 8.941713616032387e-07,
+      "loss": 0.5967,
+      "step": 18363
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.760916329509893,
+      "learning_rate": 8.940752966994889e-07,
+      "loss": 0.5866,
+      "step": 18364
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.068411265565115,
+      "learning_rate": 8.939792327843434e-07,
+      "loss": 0.5898,
+      "step": 18365
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.3969958094214427,
+      "learning_rate": 8.938831698586993e-07,
+      "loss": 0.5986,
+      "step": 18366
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.4207816414734427,
+      "learning_rate": 8.937871079234531e-07,
+      "loss": 0.5794,
+      "step": 18367
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.560126499993377,
+      "learning_rate": 8.936910469795009e-07,
+      "loss": 0.5938,
+      "step": 18368
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.7711980299757295,
+      "learning_rate": 8.935949870277398e-07,
+      "loss": 0.6322,
+      "step": 18369
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.8434927168723085,
+      "learning_rate": 8.934989280690665e-07,
+      "loss": 0.6211,
+      "step": 18370
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.617937174265819,
+      "learning_rate": 8.934028701043765e-07,
+      "loss": 0.6328,
+      "step": 18371
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.1801483197462606,
+      "learning_rate": 8.933068131345671e-07,
+      "loss": 0.6966,
+      "step": 18372
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.730934670562446,
+      "learning_rate": 8.932107571605346e-07,
+      "loss": 0.6491,
+      "step": 18373
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.96987959426323,
+      "learning_rate": 8.931147021831758e-07,
+      "loss": 0.6764,
+      "step": 18374
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.6835608163136975,
+      "learning_rate": 8.930186482033865e-07,
+      "loss": 0.6289,
+      "step": 18375
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.9820744777316057,
+      "learning_rate": 8.929225952220643e-07,
+      "loss": 0.6745,
+      "step": 18376
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.4063092565106023,
+      "learning_rate": 8.928265432401046e-07,
+      "loss": 0.6237,
+      "step": 18377
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.264484926751685,
+      "learning_rate": 8.92730492258404e-07,
+      "loss": 0.6208,
+      "step": 18378
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.120029398964007,
+      "learning_rate": 8.926344422778594e-07,
+      "loss": 0.6396,
+      "step": 18379
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.8180924250883637,
+      "learning_rate": 8.925383932993671e-07,
+      "loss": 0.6693,
+      "step": 18380
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.739101037358439,
+      "learning_rate": 8.924423453238233e-07,
+      "loss": 0.6289,
+      "step": 18381
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.3676046999879476,
+      "learning_rate": 8.923462983521249e-07,
+      "loss": 0.6543,
+      "step": 18382
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.603989081806606,
+      "learning_rate": 8.922502523851682e-07,
+      "loss": 0.6302,
+      "step": 18383
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.6740166298796653,
+      "learning_rate": 8.921542074238488e-07,
+      "loss": 0.6745,
+      "step": 18384
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.686661527202374,
+      "learning_rate": 8.920581634690642e-07,
+      "loss": 0.6693,
+      "step": 18385
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.138111481163349,
+      "learning_rate": 8.9196212052171e-07,
+      "loss": 0.6523,
+      "step": 18386
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.073664816128324,
+      "learning_rate": 8.91866078582683e-07,
+      "loss": 0.6921,
+      "step": 18387
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.380419521162681,
+      "learning_rate": 8.917700376528794e-07,
+      "loss": 0.6038,
+      "step": 18388
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.288941759137504,
+      "learning_rate": 8.916739977331958e-07,
+      "loss": 0.6615,
+      "step": 18389
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.922004770257112,
+      "learning_rate": 8.915779588245282e-07,
+      "loss": 0.653,
+      "step": 18390
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 9.441626027904165,
+      "learning_rate": 8.914819209277732e-07,
+      "loss": 0.666,
+      "step": 18391
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.14772968736576,
+      "learning_rate": 8.913858840438267e-07,
+      "loss": 0.6523,
+      "step": 18392
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.8612987841627495,
+      "learning_rate": 8.912898481735858e-07,
+      "loss": 0.641,
+      "step": 18393
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.04252009349974,
+      "learning_rate": 8.911938133179459e-07,
+      "loss": 0.6868,
+      "step": 18394
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.3370730199613,
+      "learning_rate": 8.910977794778042e-07,
+      "loss": 0.6608,
+      "step": 18395
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.9971134484638218,
+      "learning_rate": 8.910017466540567e-07,
+      "loss": 0.6257,
+      "step": 18396
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.808052601707986,
+      "learning_rate": 8.90905714847599e-07,
+      "loss": 0.6263,
+      "step": 18397
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.0102469502706355,
+      "learning_rate": 8.908096840593282e-07,
+      "loss": 0.627,
+      "step": 18398
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.8912776821861987,
+      "learning_rate": 8.907136542901401e-07,
+      "loss": 0.6152,
+      "step": 18399
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.0307948343168625,
+      "learning_rate": 8.906176255409311e-07,
+      "loss": 0.6576,
+      "step": 18400
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.029211164515479,
+      "learning_rate": 8.905215978125978e-07,
+      "loss": 0.6452,
+      "step": 18401
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 4.129455668132014,
+      "learning_rate": 8.904255711060357e-07,
+      "loss": 0.6204,
+      "step": 18402
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.481688572478496,
+      "learning_rate": 8.903295454221418e-07,
+      "loss": 0.6178,
+      "step": 18403
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 3.407360990920802,
+      "learning_rate": 8.90233520761812e-07,
+      "loss": 0.6439,
+      "step": 18404
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 5.181572466575967,
+      "learning_rate": 8.901374971259421e-07,
+      "loss": 0.6237,
+      "step": 18405
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.2495795996084595,
+      "learning_rate": 8.900414745154289e-07,
+      "loss": 0.6777,
+      "step": 18406
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.1226855535574214,
+      "learning_rate": 8.899454529311681e-07,
+      "loss": 0.6146,
+      "step": 18407
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.8250020460018956,
+      "learning_rate": 8.898494323740564e-07,
+      "loss": 0.6309,
+      "step": 18408
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.0442715696966935,
+      "learning_rate": 8.897534128449898e-07,
+      "loss": 0.6081,
+      "step": 18409
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.462350903598123,
+      "learning_rate": 8.89657394344864e-07,
+      "loss": 0.6361,
+      "step": 18410
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.6761330866923325,
+      "learning_rate": 8.895613768745755e-07,
+      "loss": 0.6419,
+      "step": 18411
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.625652888423128,
+      "learning_rate": 8.894653604350206e-07,
+      "loss": 0.5964,
+      "step": 18412
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.3556165869507644,
+      "learning_rate": 8.893693450270951e-07,
+      "loss": 0.6484,
+      "step": 18413
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.761949244419,
+      "learning_rate": 8.892733306516954e-07,
+      "loss": 0.6549,
+      "step": 18414
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.380409170447536,
+      "learning_rate": 8.891773173097175e-07,
+      "loss": 0.6237,
+      "step": 18415
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.2601214885581546,
+      "learning_rate": 8.890813050020577e-07,
+      "loss": 0.6146,
+      "step": 18416
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 5.580150644653418,
+      "learning_rate": 8.889852937296117e-07,
+      "loss": 0.6471,
+      "step": 18417
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.5144652077673832,
+      "learning_rate": 8.888892834932756e-07,
+      "loss": 0.612,
+      "step": 18418
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.316346083959718,
+      "learning_rate": 8.887932742939459e-07,
+      "loss": 0.6387,
+      "step": 18419
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.660774190960675,
+      "learning_rate": 8.886972661325183e-07,
+      "loss": 0.6374,
+      "step": 18420
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.8866657708664127,
+      "learning_rate": 8.886012590098887e-07,
+      "loss": 0.6367,
+      "step": 18421
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.439921051396002,
+      "learning_rate": 8.88505252926954e-07,
+      "loss": 0.6354,
+      "step": 18422
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.569218042381998,
+      "learning_rate": 8.884092478846094e-07,
+      "loss": 0.6882,
+      "step": 18423
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.8698316577819694,
+      "learning_rate": 8.883132438837509e-07,
+      "loss": 0.584,
+      "step": 18424
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.68633962794471,
+      "learning_rate": 8.882172409252751e-07,
+      "loss": 0.6953,
+      "step": 18425
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.560931525316673,
+      "learning_rate": 8.881212390100773e-07,
+      "loss": 0.6126,
+      "step": 18426
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.6890104248012703,
+      "learning_rate": 8.880252381390541e-07,
+      "loss": 0.6556,
+      "step": 18427
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.313009282024038,
+      "learning_rate": 8.87929238313101e-07,
+      "loss": 0.5996,
+      "step": 18428
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.221311372891644,
+      "learning_rate": 8.878332395331149e-07,
+      "loss": 0.6276,
+      "step": 18429
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.208726150472772,
+      "learning_rate": 8.877372417999906e-07,
+      "loss": 0.6693,
+      "step": 18430
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.068925308494456,
+      "learning_rate": 8.876412451146246e-07,
+      "loss": 0.5859,
+      "step": 18431
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 6.810541687069363,
+      "learning_rate": 8.875452494779126e-07,
+      "loss": 0.6823,
+      "step": 18432
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.3870053075112545,
+      "learning_rate": 8.874492548907511e-07,
+      "loss": 0.6029,
+      "step": 18433
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.7298929765445976,
+      "learning_rate": 8.873532613540354e-07,
+      "loss": 0.6439,
+      "step": 18434
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.3529046340721753,
+      "learning_rate": 8.872572688686617e-07,
+      "loss": 0.6113,
+      "step": 18435
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 4.330681225778168,
+      "learning_rate": 8.871612774355262e-07,
+      "loss": 0.638,
+      "step": 18436
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 33525,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 1676,
+  "total_flos": 276711916732416.0,
+  "train_batch_size": 22,
+  "trial_name": null,
+  "trial_params": null
+}