{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 780,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 3.155007898233803,
      "learning_rate": 8.974358974358974e-08,
      "loss": 0.7002,
      "step": 1
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.1501311823834692,
      "learning_rate": 1.7948717948717948e-07,
      "loss": 0.5469,
      "step": 2
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.2709527976504122,
      "learning_rate": 2.692307692307692e-07,
      "loss": 0.6252,
      "step": 3
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.1490588816296174,
      "learning_rate": 3.5897435897435896e-07,
      "loss": 0.6215,
      "step": 4
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.9970198743522634,
      "learning_rate": 4.4871794871794865e-07,
      "loss": 0.5639,
      "step": 5
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.3295427761499448,
      "learning_rate": 5.384615384615384e-07,
      "loss": 0.6563,
      "step": 6
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.048225120103655,
      "learning_rate": 6.282051282051282e-07,
      "loss": 0.5892,
      "step": 7
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.8352144056173314,
      "learning_rate": 7.179487179487179e-07,
      "loss": 0.6691,
      "step": 8
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.037755430732799,
      "learning_rate": 8.076923076923077e-07,
      "loss": 0.6171,
      "step": 9
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.726311486575841,
      "learning_rate": 8.974358974358973e-07,
      "loss": 0.5419,
      "step": 10
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.8423737121143846,
      "learning_rate": 9.871794871794872e-07,
      "loss": 0.5734,
      "step": 11
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.8530931377361877,
      "learning_rate": 1.0769230769230769e-06,
      "loss": 0.6041,
      "step": 12
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.549681259857384,
      "learning_rate": 1.1666666666666666e-06,
      "loss": 0.5621,
      "step": 13
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.5648145355299543,
      "learning_rate": 1.2564102564102565e-06,
      "loss": 0.6351,
      "step": 14
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.6187703396840027,
      "learning_rate": 1.3461538461538462e-06,
      "loss": 0.5119,
      "step": 15
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.4643502254723932,
      "learning_rate": 1.4358974358974359e-06,
      "loss": 0.6282,
      "step": 16
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.0662967591437957,
      "learning_rate": 1.5256410256410255e-06,
      "loss": 0.5957,
      "step": 17
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.1946771330544164,
      "learning_rate": 1.6153846153846154e-06,
      "loss": 0.6085,
      "step": 18
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.0721275066847005,
      "learning_rate": 1.7051282051282051e-06,
      "loss": 0.5127,
      "step": 19
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.088536700329697,
      "learning_rate": 1.7948717948717946e-06,
      "loss": 0.512,
      "step": 20
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.0169376562070176,
      "learning_rate": 1.8846153846153845e-06,
      "loss": 0.5363,
      "step": 21
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.9535445729471628,
      "learning_rate": 1.9743589743589744e-06,
      "loss": 0.5398,
      "step": 22
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.8891775230026828,
      "learning_rate": 2.064102564102564e-06,
      "loss": 0.4532,
      "step": 23
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.033499357097621,
      "learning_rate": 2.1538461538461538e-06,
      "loss": 0.5398,
      "step": 24
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.0971684944020184,
      "learning_rate": 2.243589743589744e-06,
      "loss": 0.5513,
      "step": 25
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9400424507855618,
      "learning_rate": 2.333333333333333e-06,
      "loss": 0.4096,
      "step": 26
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.7906588387942997,
      "learning_rate": 2.423076923076923e-06,
      "loss": 0.5276,
      "step": 27
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.0798185942573713,
      "learning_rate": 2.512820512820513e-06,
      "loss": 0.5219,
      "step": 28
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8753577770843581,
      "learning_rate": 2.6025641025641026e-06,
      "loss": 0.5799,
      "step": 29
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.9186158642556645,
      "learning_rate": 2.6923076923076923e-06,
      "loss": 0.5151,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.855939003192239,
      "learning_rate": 2.782051282051282e-06,
      "loss": 0.496,
      "step": 31
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8486864087415182,
      "learning_rate": 2.8717948717948717e-06,
      "loss": 0.5204,
      "step": 32
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8021726042394228,
      "learning_rate": 2.9615384615384614e-06,
      "loss": 0.5043,
      "step": 33
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.0357378109965776,
      "learning_rate": 3.051282051282051e-06,
      "loss": 0.5788,
      "step": 34
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.7843582385646046,
      "learning_rate": 3.141025641025641e-06,
      "loss": 0.4452,
      "step": 35
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.0000957562889896,
      "learning_rate": 3.230769230769231e-06,
      "loss": 0.5486,
      "step": 36
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.861686036469186,
      "learning_rate": 3.32051282051282e-06,
      "loss": 0.468,
      "step": 37
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9887255288842596,
      "learning_rate": 3.4102564102564103e-06,
      "loss": 0.4729,
      "step": 38
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.0718650709658224,
      "learning_rate": 3.5e-06,
      "loss": 0.5457,
      "step": 39
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.9777425001308604,
      "learning_rate": 3.5897435897435892e-06,
      "loss": 0.5257,
      "step": 40
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.6671876468169715,
      "learning_rate": 3.6794871794871797e-06,
      "loss": 0.4937,
      "step": 41
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.7049607096852184,
      "learning_rate": 3.769230769230769e-06,
      "loss": 0.4769,
      "step": 42
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7345139622673087,
      "learning_rate": 3.858974358974359e-06,
      "loss": 0.3981,
      "step": 43
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.9180834947990224,
      "learning_rate": 3.948717948717949e-06,
      "loss": 0.536,
      "step": 44
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.8152016108847437,
      "learning_rate": 4.038461538461538e-06,
      "loss": 0.4885,
      "step": 45
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.9589341729000858,
      "learning_rate": 4.128205128205128e-06,
      "loss": 0.5547,
      "step": 46
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6766248131889971,
      "learning_rate": 4.217948717948718e-06,
      "loss": 0.4473,
      "step": 47
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.7792287546251313,
      "learning_rate": 4.3076923076923076e-06,
      "loss": 0.5166,
      "step": 48
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.6833930033458306,
      "learning_rate": 4.397435897435897e-06,
      "loss": 0.5004,
      "step": 49
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.766876180946489,
      "learning_rate": 4.487179487179488e-06,
      "loss": 0.5508,
      "step": 50
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5947570972439933,
      "learning_rate": 4.576923076923077e-06,
      "loss": 0.3858,
      "step": 51
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.7707975081953566,
      "learning_rate": 4.666666666666666e-06,
      "loss": 0.4734,
      "step": 52
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.496511191973904,
      "learning_rate": 4.756410256410257e-06,
      "loss": 0.4058,
      "step": 53
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5824823317523948,
      "learning_rate": 4.846153846153846e-06,
      "loss": 0.4332,
      "step": 54
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.7090461388411984,
      "learning_rate": 4.935897435897436e-06,
      "loss": 0.4714,
      "step": 55
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.7426447758933667,
      "learning_rate": 5.025641025641026e-06,
      "loss": 0.4239,
      "step": 56
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.6842573390959508,
      "learning_rate": 5.115384615384615e-06,
      "loss": 0.4872,
      "step": 57
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.5768381008703216,
      "learning_rate": 5.205128205128205e-06,
      "loss": 0.4409,
      "step": 58
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.9230070457298252,
      "learning_rate": 5.294871794871795e-06,
      "loss": 0.404,
      "step": 59
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.676341438975508,
      "learning_rate": 5.384615384615385e-06,
      "loss": 0.5008,
      "step": 60
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.6303835534795739,
      "learning_rate": 5.474358974358974e-06,
      "loss": 0.4683,
      "step": 61
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9373304143465806,
      "learning_rate": 5.564102564102564e-06,
      "loss": 0.5097,
      "step": 62
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8345122891614065,
      "learning_rate": 5.653846153846154e-06,
      "loss": 0.5298,
      "step": 63
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.7382245856352698,
      "learning_rate": 5.743589743589743e-06,
      "loss": 0.6012,
      "step": 64
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6453526661220774,
      "learning_rate": 5.833333333333333e-06,
      "loss": 0.4989,
      "step": 65
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.3484183912221845,
      "learning_rate": 5.923076923076923e-06,
      "loss": 0.4826,
      "step": 66
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.8908943045409214,
      "learning_rate": 6.0128205128205125e-06,
      "loss": 0.4744,
      "step": 67
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8316643602939897,
      "learning_rate": 6.102564102564102e-06,
      "loss": 0.446,
      "step": 68
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.8795168400920537,
      "learning_rate": 6.192307692307692e-06,
      "loss": 0.5211,
      "step": 69
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7156763474826677,
      "learning_rate": 6.282051282051282e-06,
      "loss": 0.4573,
      "step": 70
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.697549263992999,
      "learning_rate": 6.371794871794871e-06,
      "loss": 0.5127,
      "step": 71
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.8628486662084527,
      "learning_rate": 6.461538461538462e-06,
      "loss": 0.368,
      "step": 72
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7585623399087367,
      "learning_rate": 6.5512820512820515e-06,
      "loss": 0.4697,
      "step": 73
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6350247762941712,
      "learning_rate": 6.64102564102564e-06,
      "loss": 0.4568,
      "step": 74
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.625191280670622,
      "learning_rate": 6.730769230769231e-06,
      "loss": 0.4232,
      "step": 75
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.8473726342325616,
      "learning_rate": 6.8205128205128205e-06,
      "loss": 0.5019,
      "step": 76
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.743248260743822,
      "learning_rate": 6.91025641025641e-06,
      "loss": 0.475,
      "step": 77
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.62305010114068,
      "learning_rate": 7e-06,
      "loss": 0.4467,
      "step": 78
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.7466573021923082,
      "learning_rate": 6.9999649520318915e-06,
      "loss": 0.4301,
      "step": 79
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6130437414343075,
      "learning_rate": 6.999859808829483e-06,
      "loss": 0.4234,
      "step": 80
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.714296938271923,
      "learning_rate": 6.999684572498523e-06,
      "loss": 0.4479,
      "step": 81
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6478015443180727,
      "learning_rate": 6.999439246548541e-06,
      "loss": 0.4092,
      "step": 82
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.772703416513747,
      "learning_rate": 6.999123835892781e-06,
      "loss": 0.504,
      "step": 83
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8216260385526901,
      "learning_rate": 6.998738346848099e-06,
      "loss": 0.5069,
      "step": 84
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6744716503404546,
      "learning_rate": 6.998282787134845e-06,
      "loss": 0.4726,
      "step": 85
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.772184185014806,
      "learning_rate": 6.997757165876698e-06,
      "loss": 0.5105,
      "step": 86
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7107134524844578,
      "learning_rate": 6.9971614936004935e-06,
      "loss": 0.4361,
      "step": 87
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7373273309454584,
      "learning_rate": 6.996495782236003e-06,
      "loss": 0.4909,
      "step": 88
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.73191097304193,
      "learning_rate": 6.9957600451157e-06,
      "loss": 0.456,
      "step": 89
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.543591250610673,
      "learning_rate": 6.9949542969744955e-06,
      "loss": 0.4159,
      "step": 90
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.9119020098570365,
      "learning_rate": 6.9940785539494385e-06,
      "loss": 0.5059,
      "step": 91
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7231017193785318,
      "learning_rate": 6.9931328335793926e-06,
      "loss": 0.4661,
      "step": 92
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6314842551489181,
      "learning_rate": 6.992117154804688e-06,
      "loss": 0.4521,
      "step": 93
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7528843946113917,
      "learning_rate": 6.991031537966741e-06,
      "loss": 0.4864,
      "step": 94
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.773871809828818,
      "learning_rate": 6.989876004807644e-06,
      "loss": 0.5162,
      "step": 95
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5399343944436803,
      "learning_rate": 6.9886505784697354e-06,
      "loss": 0.4338,
      "step": 96
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7514912857415266,
      "learning_rate": 6.98735528349513e-06,
      "loss": 0.4747,
      "step": 97
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.555552457796697,
      "learning_rate": 6.985990145825233e-06,
      "loss": 0.4169,
      "step": 98
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.8134922361539518,
      "learning_rate": 6.984555192800216e-06,
      "loss": 0.5132,
      "step": 99
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.897574482765648,
      "learning_rate": 6.983050453158471e-06,
      "loss": 0.4651,
      "step": 100
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7158733456321913,
      "learning_rate": 6.981475957036039e-06,
      "loss": 0.4445,
      "step": 101
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.4473946002542672,
      "learning_rate": 6.979831735965997e-06,
      "loss": 0.4479,
      "step": 102
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7162901512740472,
      "learning_rate": 6.9781178228778385e-06,
      "loss": 0.4596,
      "step": 103
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6337084396881052,
      "learning_rate": 6.9763342520968e-06,
      "loss": 0.3578,
      "step": 104
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6651285180833972,
      "learning_rate": 6.974481059343188e-06,
      "loss": 0.3986,
      "step": 105
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5096225766725835,
      "learning_rate": 6.972558281731655e-06,
      "loss": 0.384,
      "step": 106
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.511217953681732,
      "learning_rate": 6.970565957770456e-06,
      "loss": 0.4286,
      "step": 107
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7549162394009117,
      "learning_rate": 6.96850412736068e-06,
      "loss": 0.4961,
      "step": 108
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6794558501418015,
      "learning_rate": 6.9663728317954505e-06,
      "loss": 0.488,
      "step": 109
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.8013458849210429,
      "learning_rate": 6.9641721137591e-06,
      "loss": 0.5055,
      "step": 110
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5223342051762156,
      "learning_rate": 6.961902017326311e-06,
      "loss": 0.4384,
      "step": 111
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.75400017947144,
      "learning_rate": 6.959562587961235e-06,
      "loss": 0.4275,
      "step": 112
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.4432884930631757,
      "learning_rate": 6.9571538725165855e-06,
      "loss": 0.3806,
      "step": 113
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.3728512834355604,
      "learning_rate": 6.9546759192326944e-06,
      "loss": 0.3347,
      "step": 114
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7014083625164993,
      "learning_rate": 6.95212877773655e-06,
      "loss": 0.4893,
      "step": 115
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7435972771230723,
      "learning_rate": 6.949512499040799e-06,
      "loss": 0.4518,
      "step": 116
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6443871246884252,
      "learning_rate": 6.946827135542729e-06,
      "loss": 0.4675,
      "step": 117
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.4381128278032245,
      "learning_rate": 6.944072741023215e-06,
      "loss": 0.3713,
      "step": 118
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6569315631240311,
      "learning_rate": 6.941249370645649e-06,
      "loss": 0.4382,
      "step": 119
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6251022822204857,
      "learning_rate": 6.938357080954826e-06,
      "loss": 0.4447,
      "step": 120
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7598566755719935,
      "learning_rate": 6.935395929875821e-06,
      "loss": 0.4624,
      "step": 121
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5423964387806426,
      "learning_rate": 6.93236597671282e-06,
      "loss": 0.3977,
      "step": 122
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7628631989109886,
      "learning_rate": 6.929267282147936e-06,
      "loss": 0.446,
      "step": 123
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6468615154504636,
      "learning_rate": 6.9260999082400014e-06,
      "loss": 0.474,
      "step": 124
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.69755130421299,
      "learning_rate": 6.922863918423311e-06,
      "loss": 0.4332,
      "step": 125
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8083022682191605,
      "learning_rate": 6.91955937750636e-06,
      "loss": 0.5805,
      "step": 126
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6518789318820335,
      "learning_rate": 6.916186351670546e-06,
      "loss": 0.4257,
      "step": 127
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.488823939533215,
      "learning_rate": 6.912744908468841e-06,
      "loss": 0.3751,
      "step": 128
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7757966013804762,
      "learning_rate": 6.909235116824441e-06,
      "loss": 0.469,
      "step": 129
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.592505278493427,
      "learning_rate": 6.905657047029383e-06,
      "loss": 0.4143,
      "step": 130
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7305116612692355,
      "learning_rate": 6.90201077074314e-06,
      "loss": 0.4315,
      "step": 131
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7311239078049492,
      "learning_rate": 6.898296360991182e-06,
      "loss": 0.4603,
      "step": 132
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7269770749278417,
      "learning_rate": 6.894513892163519e-06,
      "loss": 0.4569,
      "step": 133
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7096122263349194,
      "learning_rate": 6.890663440013204e-06,
      "loss": 0.4859,
      "step": 134
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8247290588131861,
      "learning_rate": 6.886745081654823e-06,
      "loss": 0.4798,
      "step": 135
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6368752515489966,
      "learning_rate": 6.882758895562948e-06,
      "loss": 0.4624,
      "step": 136
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7503694358624682,
      "learning_rate": 6.8787049615705635e-06,
      "loss": 0.4587,
      "step": 137
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6383457284082776,
      "learning_rate": 6.8745833608674685e-06,
      "loss": 0.4008,
      "step": 138
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5622812578004324,
      "learning_rate": 6.870394175998651e-06,
      "loss": 0.4348,
      "step": 139
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.5762025328026612,
      "learning_rate": 6.866137490862636e-06,
      "loss": 0.4095,
      "step": 140
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.673415899230875,
      "learning_rate": 6.861813390709803e-06,
      "loss": 0.4659,
      "step": 141
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.5962555180641838,
      "learning_rate": 6.857421962140681e-06,
      "loss": 0.4213,
      "step": 142
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.7562554854422003,
      "learning_rate": 6.852963293104211e-06,
      "loss": 0.5064,
      "step": 143
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.6633585065572871,
      "learning_rate": 6.848437472895989e-06,
      "loss": 0.4975,
      "step": 144
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.7373610441369685,
      "learning_rate": 6.84384459215647e-06,
      "loss": 0.4097,
      "step": 145
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.8113935248295507,
      "learning_rate": 6.839184742869166e-06,
      "loss": 0.539,
      "step": 146
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.6602786354588224,
      "learning_rate": 6.8344580183587866e-06,
      "loss": 0.4478,
      "step": 147
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.738271233998496,
      "learning_rate": 6.829664513289387e-06,
      "loss": 0.424,
      "step": 148
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.814406857491063,
      "learning_rate": 6.824804323662456e-06,
      "loss": 0.5051,
      "step": 149
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.678131337111829,
      "learning_rate": 6.8198775468150085e-06,
      "loss": 0.5028,
      "step": 150
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.6059802835602406,
      "learning_rate": 6.814884281417627e-06,
      "loss": 0.4861,
      "step": 151
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.5551551029291752,
      "learning_rate": 6.8098246274724835e-06,
      "loss": 0.4236,
      "step": 152
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.6508065774889,
      "learning_rate": 6.8046986863113455e-06,
      "loss": 0.5092,
      "step": 153
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.6796345029479292,
      "learning_rate": 6.7995065605935405e-06,
      "loss": 0.4368,
      "step": 154
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.6592107174838626,
      "learning_rate": 6.7942483543039e-06,
      "loss": 0.3755,
      "step": 155
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.4850945503983306,
      "learning_rate": 6.788924172750679e-06,
      "loss": 0.419,
      "step": 156
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.636480713605989,
      "learning_rate": 6.783534122563447e-06,
      "loss": 0.4985,
      "step": 157
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.4905731963500435,
      "learning_rate": 6.7780783116909495e-06,
      "loss": 0.3661,
      "step": 158
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.656763138900859,
      "learning_rate": 6.772556849398952e-06,
      "loss": 0.4249,
      "step": 159
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.6619926692208726,
      "learning_rate": 6.7669698462680434e-06,
      "loss": 0.4119,
      "step": 160
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.5244839973693285,
      "learning_rate": 6.761317414191428e-06,
      "loss": 0.3802,
      "step": 161
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.705620627532888,
      "learning_rate": 6.755599666372685e-06,
      "loss": 0.439,
      "step": 162
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.6267790742105397,
      "learning_rate": 6.749816717323493e-06,
      "loss": 0.3791,
      "step": 163
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.6069385218303462,
      "learning_rate": 6.743968682861346e-06,
      "loss": 0.4668,
      "step": 164
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.656574398737437,
      "learning_rate": 6.738055680107233e-06,
      "loss": 0.404,
      "step": 165
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.4902964557198772,
      "learning_rate": 6.7320778274832836e-06,
      "loss": 0.3513,
      "step": 166
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.5935267332780712,
      "learning_rate": 6.726035244710406e-06,
      "loss": 0.3731,
      "step": 167
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.6490590043388837,
      "learning_rate": 6.7199280528058844e-06,
      "loss": 0.3951,
      "step": 168
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.5982484017746137,
      "learning_rate": 6.713756374080959e-06,
      "loss": 0.3872,
      "step": 169
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.7217766197864206,
      "learning_rate": 6.70752033213837e-06,
      "loss": 0.423,
      "step": 170
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.616116261962514,
      "learning_rate": 6.7012200518698904e-06,
      "loss": 0.3299,
      "step": 171
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.7860181512527011,
      "learning_rate": 6.6948556594538185e-06,
      "loss": 0.4138,
      "step": 172
    },
    {
      "epoch": 1.11,
      "grad_norm": 1.5952278874239347,
      "learning_rate": 6.688427282352449e-06,
      "loss": 0.4096,
      "step": 173
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.7136967077306529,
      "learning_rate": 6.681935049309533e-06,
      "loss": 0.4069,
      "step": 174
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.5749205024444042,
      "learning_rate": 6.6753790903476814e-06,
      "loss": 0.3344,
      "step": 175
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.5363523923720421,
      "learning_rate": 6.668759536765778e-06,
      "loss": 0.3373,
      "step": 176
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.6413945984220908,
      "learning_rate": 6.6620765211363376e-06,
      "loss": 0.3469,
      "step": 177
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.6144791165228076,
      "learning_rate": 6.655330177302857e-06,
      "loss": 0.352,
      "step": 178
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.6043475836227805,
      "learning_rate": 6.64852064037713e-06,
      "loss": 0.2867,
      "step": 179
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.6935976486539148,
      "learning_rate": 6.6416480467365494e-06,
      "loss": 0.3408,
      "step": 180
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.7124502452903085,
      "learning_rate": 6.634712534021367e-06,
      "loss": 0.353,
      "step": 181
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.4992513277039652,
      "learning_rate": 6.627714241131943e-06,
      "loss": 0.2486,
      "step": 182
    },
    {
      "epoch": 1.17,
      "grad_norm": 1.593568781257124,
      "learning_rate": 6.62065330822596e-06,
      "loss": 0.3488,
      "step": 183
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.6009882930842492,
      "learning_rate": 6.613529876715619e-06,
      "loss": 0.3361,
      "step": 184
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.6418303297336283,
      "learning_rate": 6.606344089264805e-06,
      "loss": 0.3674,
      "step": 185
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.6506486655738626,
      "learning_rate": 6.599096089786234e-06,
      "loss": 0.3231,
      "step": 186
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.576597053257656,
      "learning_rate": 6.591786023438565e-06,
      "loss": 0.3197,
      "step": 187
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5882663115394593,
      "learning_rate": 6.5844140366234956e-06,
      "loss": 0.3217,
      "step": 188
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.5005363443170452,
      "learning_rate": 6.576980276982832e-06,
      "loss": 0.3101,
      "step": 189
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.7065502782313382,
      "learning_rate": 6.569484893395527e-06,
      "loss": 0.3607,
      "step": 190
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.4576343806529883,
      "learning_rate": 6.5619280359747045e-06,
      "loss": 0.273,
      "step": 191
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.6846657499731517,
      "learning_rate": 6.55430985606465e-06,
      "loss": 0.3221,
      "step": 192
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.5588135974495478,
      "learning_rate": 6.546630506237778e-06,
      "loss": 0.2656,
      "step": 193
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.669978644846293,
      "learning_rate": 6.538890140291578e-06,
      "loss": 0.2425,
      "step": 194
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.788539476603885,
      "learning_rate": 6.531088913245536e-06,
      "loss": 0.3166,
      "step": 195
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.7210858323037903,
      "learning_rate": 6.5232269813380254e-06,
      "loss": 0.3006,
      "step": 196
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.702169038700813,
      "learning_rate": 6.5153045020231855e-06,
      "loss": 0.2967,
      "step": 197
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.6563664353691507,
      "learning_rate": 6.507321633967758e-06,
      "loss": 0.2713,
      "step": 198
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.5309368809239376,
      "learning_rate": 6.499278537047919e-06,
      "loss": 0.2147,
      "step": 199
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.7209942943977308,
      "learning_rate": 6.49117537234607e-06,
      "loss": 0.3018,
      "step": 200
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.7129877502808601,
      "learning_rate": 6.483012302147617e-06,
      "loss": 0.2729,
      "step": 201
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.7496718789927703,
      "learning_rate": 6.474789489937715e-06,
      "loss": 0.2876,
      "step": 202
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.5933095070593675,
      "learning_rate": 6.4665071003979985e-06,
      "loss": 0.2438,
      "step": 203
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.7227892810373833,
      "learning_rate": 6.4581652994032816e-06,
      "loss": 0.2842,
      "step": 204
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.6267649305204939,
      "learning_rate": 6.449764254018236e-06,
      "loss": 0.2789,
      "step": 205
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.5850041512660988,
      "learning_rate": 6.441304132494045e-06,
      "loss": 0.297,
      "step": 206
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.4696703594193863,
      "learning_rate": 6.432785104265034e-06,
      "loss": 0.2101,
      "step": 207
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.656066428185347,
      "learning_rate": 6.424207339945278e-06,
      "loss": 0.2409,
      "step": 208
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.4167653617356255,
      "learning_rate": 6.415571011325181e-06,
      "loss": 0.2209,
      "step": 209
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.4983529166587497,
      "learning_rate": 6.406876291368041e-06,
      "loss": 0.2368,
      "step": 210
    },
    {
      "epoch": 1.35,
      "grad_norm": 1.5865913377222434,
      "learning_rate": 6.3981233542065824e-06,
      "loss": 0.2506,
      "step": 211
    },
    {
      "epoch": 1.36,
      "grad_norm": 1.5796504194067498,
      "learning_rate": 6.3893123751394695e-06,
      "loss": 0.2091,
      "step": 212
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.5981542013013652,
      "learning_rate": 6.380443530627797e-06,
      "loss": 0.2462,
      "step": 213
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.490322079361194,
      "learning_rate": 6.371516998291552e-06,
      "loss": 0.2259,
      "step": 214
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.596961848168018,
      "learning_rate": 6.3625329569060595e-06,
      "loss": 0.1738,
      "step": 215
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.5745062034098327,
      "learning_rate": 6.3534915863984045e-06,
      "loss": 0.2493,
      "step": 216
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.5450698814984771,
      "learning_rate": 6.344393067843825e-06,
      "loss": 0.2377,
      "step": 217
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.7029756458678342,
      "learning_rate": 6.335237583462083e-06,
      "loss": 0.2416,
      "step": 218
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.7441512485411754,
      "learning_rate": 6.326025316613824e-06,
      "loss": 0.254,
      "step": 219
    },
    {
      "epoch": 1.41,
      "grad_norm": 1.6605837222679327,
      "learning_rate": 6.3167564517968944e-06,
      "loss": 0.314,
      "step": 220
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.565133849108226,
      "learning_rate": 6.307431174642653e-06,
      "loss": 0.244,
      "step": 221
    },
    {
      "epoch": 1.42,
      "grad_norm": 1.5836143873013986,
      "learning_rate": 6.2980496719122544e-06,
      "loss": 0.2213,
      "step": 222
    },
    {
      "epoch": 1.43,
      "grad_norm": 1.528965218826486,
      "learning_rate": 6.288612131492901e-06,
      "loss": 0.2033,
      "step": 223
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.5075906007645412,
      "learning_rate": 6.279118742394089e-06,
      "loss": 0.1995,
      "step": 224
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.6753818522307469,
      "learning_rate": 6.2695696947438165e-06,
      "loss": 0.2259,
      "step": 225
    },
    {
      "epoch": 1.45,
      "grad_norm": 1.5963342713697721,
      "learning_rate": 6.25996517978478e-06,
      "loss": 0.2074,
      "step": 226
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.619015544516661,
      "learning_rate": 6.2503053898705416e-06,
      "loss": 0.2407,
      "step": 227
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.5457927237812485,
      "learning_rate": 6.2405905184616776e-06,
      "loss": 0.1524,
      "step": 228
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.5882620229865172,
      "learning_rate": 6.230820760121904e-06,
      "loss": 0.1884,
      "step": 229
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.561396094087553,
      "learning_rate": 6.220996310514181e-06,
      "loss": 0.1982,
      "step": 230
    },
    {
      "epoch": 1.48,
      "grad_norm": 1.5817943374237846,
      "learning_rate": 6.21111736639679e-06,
      "loss": 0.1935,
      "step": 231
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.6950787867836972,
      "learning_rate": 6.201184125619403e-06,
      "loss": 0.2077,
      "step": 232
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.6754831923835654,
      "learning_rate": 6.191196787119104e-06,
      "loss": 0.213,
      "step": 233
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.6029146218294437,
      "learning_rate": 6.181155550916423e-06,
      "loss": 0.2059,
      "step": 234
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.5323755489476472,
      "learning_rate": 6.171060618111317e-06,
      "loss": 0.174,
      "step": 235
    },
    {
      "epoch": 1.51,
      "grad_norm": 1.557257858564408,
      "learning_rate": 6.160912190879146e-06,
      "loss": 0.1968,
      "step": 236
    },
    {
      "epoch": 1.52,
      "grad_norm": 1.6171718420036003,
      "learning_rate": 6.15071047246663e-06,
      "loss": 0.18,
      "step": 237
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.480227657890647,
      "learning_rate": 6.140455667187765e-06,
      "loss": 0.1717,
      "step": 238
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.551408341338327,
      "learning_rate": 6.13014798041975e-06,
      "loss": 0.2161,
      "step": 239
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.6207949878686716,
      "learning_rate": 6.119787618598854e-06,
      "loss": 0.2148,
      "step": 240
    },
    {
      "epoch": 1.54,
      "grad_norm": 1.5643103300094428,
      "learning_rate": 6.109374789216296e-06,
      "loss": 0.2138,
      "step": 241
    },
    {
      "epoch": 1.55,
      "grad_norm": 1.6967073475220011,
      "learning_rate": 6.098909700814082e-06,
      "loss": 0.2429,
      "step": 242
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5728184775316987,
      "learning_rate": 6.08839256298083e-06,
      "loss": 0.1942,
      "step": 243
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5637927560520009,
      "learning_rate": 6.077823586347579e-06,
      "loss": 0.2019,
      "step": 244
    },
    {
      "epoch": 1.57,
      "grad_norm": 1.568288688763992,
      "learning_rate": 6.06720298258356e-06,
      "loss": 0.1926,
      "step": 245
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.4419153948220185,
      "learning_rate": 6.056530964391961e-06,
      "loss": 0.1975,
      "step": 246
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.686492281338865,
      "learning_rate": 6.0458077455056704e-06,
      "loss": 0.219,
      "step": 247
    },
    {
      "epoch": 1.59,
      "grad_norm": 1.5332314891866488,
      "learning_rate": 6.035033540682993e-06,
      "loss": 0.2005,
      "step": 248
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5577465097613148,
      "learning_rate": 6.024208565703351e-06,
      "loss": 0.2131,
      "step": 249
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.5794506409096423,
      "learning_rate": 6.013333037362959e-06,
      "loss": 0.2096,
      "step": 250
    },
    {
      "epoch": 1.61,
      "grad_norm": 1.6458078130518516,
      "learning_rate": 6.002407173470486e-06,
      "loss": 0.2421,
      "step": 251
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.5012439233130601,
      "learning_rate": 5.991431192842692e-06,
      "loss": 0.2083,
      "step": 252
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.5760275134953128,
      "learning_rate": 5.980405315300045e-06,
      "loss": 0.2047,
      "step": 253
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.4555813197229432,
      "learning_rate": 5.969329761662319e-06,
      "loss": 0.1822,
      "step": 254
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.5857225468346239,
      "learning_rate": 5.9582047537441716e-06,
      "loss": 0.219,
      "step": 255
    },
    {
      "epoch": 1.64,
      "grad_norm": 1.745530982159658,
      "learning_rate": 5.9470305143507e-06,
      "loss": 0.1774,
      "step": 256
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.6320256493079581,
      "learning_rate": 5.9358072672729845e-06,
      "loss": 0.1942,
      "step": 257
    },
    {
      "epoch": 1.65,
      "grad_norm": 1.6361095890331265,
      "learning_rate": 5.924535237283598e-06,
      "loss": 0.2505,
      "step": 258
    },
    {
      "epoch": 1.66,
      "grad_norm": 1.5993814196455234,
      "learning_rate": 5.913214650132112e-06,
      "loss": 0.1935,
      "step": 259
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.3895790850912417,
      "learning_rate": 5.901845732540568e-06,
      "loss": 0.1309,
      "step": 260
    },
    {
      "epoch": 1.67,
      "grad_norm": 1.4752869859546751,
      "learning_rate": 5.8904287121989455e-06,
      "loss": 0.1537,
      "step": 261
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.5193526100418906,
      "learning_rate": 5.878963817760597e-06,
      "loss": 0.1722,
      "step": 262
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.4815632509360366,
      "learning_rate": 5.867451278837666e-06,
      "loss": 0.1962,
      "step": 263
    },
    {
      "epoch": 1.69,
      "grad_norm": 1.6845916508822236,
      "learning_rate": 5.855891325996495e-06,
      "loss": 0.2095,
      "step": 264
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.8207264207099068,
      "learning_rate": 5.8442841907530035e-06,
      "loss": 0.2485,
      "step": 265
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.5523354061842987,
      "learning_rate": 5.83263010556805e-06,
      "loss": 0.2203,
      "step": 266
    },
    {
      "epoch": 1.71,
      "grad_norm": 1.488592846305629,
      "learning_rate": 5.820929303842783e-06,
      "loss": 0.2163,
      "step": 267
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.5016962959606086,
      "learning_rate": 5.809182019913959e-06,
      "loss": 0.1769,
      "step": 268
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.4132734874193642,
      "learning_rate": 5.797388489049253e-06,
      "loss": 0.1891,
      "step": 269
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.3907078826471153,
      "learning_rate": 5.785548947442547e-06,
      "loss": 0.1639,
      "step": 270
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5952473846369721,
      "learning_rate": 5.7736636322092016e-06,
      "loss": 0.2225,
      "step": 271
    },
    {
      "epoch": 1.74,
      "grad_norm": 1.5983856197422255,
      "learning_rate": 5.7617327813813e-06,
      "loss": 0.1967,
      "step": 272
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.6251445895164713,
      "learning_rate": 5.749756633902887e-06,
      "loss": 0.2141,
      "step": 273
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.5239667480122763,
      "learning_rate": 5.7377354296251855e-06,
      "loss": 0.1789,
      "step": 274
    },
    {
      "epoch": 1.76,
      "grad_norm": 1.5733312077254948,
      "learning_rate": 5.725669409301782e-06,
      "loss": 0.1891,
      "step": 275
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.579084689666494,
      "learning_rate": 5.71355881458382e-06,
      "loss": 0.2104,
      "step": 276
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.6098678412924634,
      "learning_rate": 5.701403888015149e-06,
      "loss": 0.1883,
      "step": 277
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.5520664946596863,
      "learning_rate": 5.689204873027471e-06,
      "loss": 0.1873,
      "step": 278
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.643706186330864,
      "learning_rate": 5.676962013935464e-06,
      "loss": 0.1833,
      "step": 279
    },
    {
      "epoch": 1.79,
      "grad_norm": 1.574632884356156,
      "learning_rate": 5.664675555931892e-06,
      "loss": 0.2086,
      "step": 280
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.7040654897451826,
      "learning_rate": 5.652345745082691e-06,
      "loss": 0.1947,
      "step": 281
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.7990396739414352,
      "learning_rate": 5.639972828322043e-06,
      "loss": 0.2834,
      "step": 282
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.545632160225994,
      "learning_rate": 5.627557053447427e-06,
      "loss": 0.1823,
      "step": 283
    },
    {
      "epoch": 1.82,
      "grad_norm": 1.4802183704681011,
      "learning_rate": 5.615098669114664e-06,
      "loss": 0.1591,
      "step": 284
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.6097297868594553,
      "learning_rate": 5.6025979248329265e-06,
      "loss": 0.1851,
      "step": 285
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.5317448245457972,
      "learning_rate": 5.590055070959752e-06,
      "loss": 0.1933,
      "step": 286
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.6207216281170533,
      "learning_rate": 5.577470358696021e-06,
      "loss": 0.187,
      "step": 287
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.6156124581387827,
      "learning_rate": 5.564844040080931e-06,
      "loss": 0.2033,
      "step": 288
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.6248007053651101,
      "learning_rate": 5.5521763679869445e-06,
      "loss": 0.2027,
      "step": 289
    },
    {
      "epoch": 1.86,
      "grad_norm": 1.703201832873631,
      "learning_rate": 5.53946759611473e-06,
      "loss": 0.2355,
      "step": 290
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.6347182792762331,
      "learning_rate": 5.526717978988076e-06,
      "loss": 0.196,
      "step": 291
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.6347383876881978,
      "learning_rate": 5.513927771948798e-06,
      "loss": 0.2136,
      "step": 292
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.6210512023086405,
      "learning_rate": 5.5010972311516184e-06,
      "loss": 0.2023,
      "step": 293
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.5825507774419676,
      "learning_rate": 5.488226613559045e-06,
      "loss": 0.174,
      "step": 294
    },
    {
      "epoch": 1.89,
      "grad_norm": 1.6023419187659833,
      "learning_rate": 5.475316176936217e-06,
      "loss": 0.1988,
      "step": 295
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.5025326070404128,
      "learning_rate": 5.462366179845746e-06,
      "loss": 0.182,
      "step": 296
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.599375014112513,
      "learning_rate": 5.449376881642537e-06,
      "loss": 0.2045,
      "step": 297
    },
    {
      "epoch": 1.91,
      "grad_norm": 1.5465339822704995,
      "learning_rate": 5.436348542468598e-06,
      "loss": 0.2029,
      "step": 298
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.6720010038710715,
      "learning_rate": 5.423281423247821e-06,
      "loss": 0.2109,
      "step": 299
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.6400451846120492,
      "learning_rate": 5.4101757856807655e-06,
      "loss": 0.2411,
      "step": 300
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.6036906269837266,
      "learning_rate": 5.397031892239414e-06,
      "loss": 0.1787,
      "step": 301
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.726874638894533,
      "learning_rate": 5.383850006161913e-06,
      "loss": 0.2479,
      "step": 302
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.5387638769935164,
      "learning_rate": 5.370630391447303e-06,
      "loss": 0.1976,
      "step": 303
    },
    {
      "epoch": 1.95,
      "grad_norm": 1.6078608967191599,
      "learning_rate": 5.357373312850236e-06,
      "loss": 0.1819,
      "step": 304
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.6795308057760014,
      "learning_rate": 5.3440790358756615e-06,
      "loss": 0.2084,
      "step": 305
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.6968139505178514,
      "learning_rate": 5.330747826773522e-06,
      "loss": 0.2359,
      "step": 306
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.6327022776957227,
      "learning_rate": 5.317379952533411e-06,
      "loss": 0.2256,
      "step": 307
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.5535212129206661,
      "learning_rate": 5.303975680879232e-06,
      "loss": 0.204,
      "step": 308
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.6859303635271699,
      "learning_rate": 5.290535280263835e-06,
      "loss": 0.2601,
      "step": 309
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.6081244411677793,
      "learning_rate": 5.277059019863637e-06,
      "loss": 0.1961,
      "step": 310
    },
    {
      "epoch": 1.99,
      "grad_norm": 1.4376728553681852,
      "learning_rate": 5.263547169573235e-06,
      "loss": 0.1625,
      "step": 311
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4754384583406648,
      "learning_rate": 5.25e-06,
      "loss": 0.2084,
      "step": 312
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.629012127398148,
      "learning_rate": 5.236417782458656e-06,
      "loss": 0.2223,
      "step": 313
    },
    {
      "epoch": 2.01,
      "grad_norm": 1.4053184464424964,
      "learning_rate": 5.222800788965847e-06,
      "loss": 0.1541,
      "step": 314
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.5168677999761422,
      "learning_rate": 5.2091492922346894e-06,
      "loss": 0.1835,
      "step": 315
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.5300778099879906,
      "learning_rate": 5.195463565669309e-06,
      "loss": 0.1801,
      "step": 316
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.5081736458359112,
      "learning_rate": 5.18174388335937e-06,
      "loss": 0.1612,
      "step": 317
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.5873664179989553,
      "learning_rate": 5.167990520074577e-06,
      "loss": 0.1686,
      "step": 318
    },
    {
      "epoch": 2.04,
      "grad_norm": 1.586158856785377,
      "learning_rate": 5.154203751259183e-06,
      "loss": 0.1511,
      "step": 319
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.6403439753706173,
      "learning_rate": 5.140383853026463e-06,
      "loss": 0.2097,
      "step": 320
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.5021116951267204,
      "learning_rate": 5.12653110215319e-06,
      "loss": 0.1494,
      "step": 321
    },
    {
      "epoch": 2.06,
      "grad_norm": 1.5031304131151833,
      "learning_rate": 5.11264577607409e-06,
      "loss": 0.155,
      "step": 322
    },
    {
      "epoch": 2.07,
      "grad_norm": 1.4979853778197805,
      "learning_rate": 5.098728152876287e-06,
      "loss": 0.1474,
      "step": 323
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.583104755788855,
      "learning_rate": 5.084778511293731e-06,
      "loss": 0.1551,
      "step": 324
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.5052221933069465,
      "learning_rate": 5.070797130701618e-06,
      "loss": 0.1641,
      "step": 325
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.5841092318010859,
      "learning_rate": 5.056784291110794e-06,
      "loss": 0.1523,
      "step": 326
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.56655325522275,
      "learning_rate": 5.04274027316215e-06,
      "loss": 0.107,
      "step": 327
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.546032887920805,
      "learning_rate": 5.028665358120995e-06,
      "loss": 0.1437,
      "step": 328
    },
    {
      "epoch": 2.11,
      "grad_norm": 1.6885562930270839,
      "learning_rate": 5.014559827871426e-06,
      "loss": 0.1569,
      "step": 329
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.4483394611955436,
      "learning_rate": 5.00042396491069e-06,
      "loss": 0.137,
      "step": 330
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.4448536208556515,
      "learning_rate": 4.9862580523435116e-06,
      "loss": 0.1276,
      "step": 331
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.4797511760049469,
      "learning_rate": 4.972062373876435e-06,
      "loss": 0.1249,
      "step": 332
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.5847877833278934,
      "learning_rate": 4.95783721381214e-06,
      "loss": 0.124,
      "step": 333
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.5568154599963886,
      "learning_rate": 4.943582857043742e-06,
      "loss": 0.1214,
      "step": 334
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.4041826443755792,
      "learning_rate": 4.9292995890490945e-06,
      "loss": 0.0838,
      "step": 335
    },
    {
      "epoch": 2.15,
      "grad_norm": 1.6919217726047346,
      "learning_rate": 4.914987695885067e-06,
      "loss": 0.1051,
      "step": 336
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.7270792065913758,
      "learning_rate": 4.900647464181817e-06,
      "loss": 0.1149,
      "step": 337
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.4384806218796988,
      "learning_rate": 4.886279181137049e-06,
      "loss": 0.0733,
      "step": 338
    },
    {
      "epoch": 2.17,
      "grad_norm": 1.5148604456334718,
      "learning_rate": 4.871883134510263e-06,
      "loss": 0.1204,
      "step": 339
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.5573623515183024,
      "learning_rate": 4.8574596126169925e-06,
      "loss": 0.1158,
      "step": 340
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.667599893884288,
      "learning_rate": 4.843008904323029e-06,
      "loss": 0.1218,
      "step": 341
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.6336415171179288,
      "learning_rate": 4.828531299038638e-06,
      "loss": 0.1064,
      "step": 342
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.7857223150803634,
      "learning_rate": 4.81402708671276e-06,
      "loss": 0.1101,
      "step": 343
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.690625690003038,
      "learning_rate": 4.799496557827208e-06,
      "loss": 0.1141,
      "step": 344
    },
    {
      "epoch": 2.21,
      "grad_norm": 1.4918432091961047,
      "learning_rate": 4.7849400033908465e-06,
      "loss": 0.113,
      "step": 345
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.616344221375453,
      "learning_rate": 4.770357714933765e-06,
      "loss": 0.1263,
      "step": 346
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.493204431026351,
      "learning_rate": 4.755749984501437e-06,
      "loss": 0.0964,
      "step": 347
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.5095159402266893,
      "learning_rate": 4.741117104648874e-06,
      "loss": 0.1008,
      "step": 348
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.2650625741158785,
      "learning_rate": 4.726459368434768e-06,
      "loss": 0.0764,
      "step": 349
    },
    {
      "epoch": 2.24,
      "grad_norm": 1.4886734247960312,
      "learning_rate": 4.711777069415615e-06,
      "loss": 0.0717,
      "step": 350
    },
    {
      "epoch": 2.25,
      "grad_norm": 1.5028429790381546,
      "learning_rate": 4.697070501639841e-06,
      "loss": 0.0978,
      "step": 351
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.5105459796954572,
      "learning_rate": 4.682339959641915e-06,
      "loss": 0.0997,
      "step": 352
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.4581717109467751,
      "learning_rate": 4.667585738436448e-06,
      "loss": 0.0935,
      "step": 353
    },
    {
      "epoch": 2.27,
      "grad_norm": 1.3990942636732424,
      "learning_rate": 4.652808133512279e-06,
      "loss": 0.0848,
      "step": 354
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.3125915080099024,
      "learning_rate": 4.638007440826568e-06,
      "loss": 0.06,
      "step": 355
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.5922559613481908,
      "learning_rate": 4.62318395679886e-06,
      "loss": 0.09,
      "step": 356
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.6473855344758028,
      "learning_rate": 4.6083379783051545e-06,
      "loss": 0.0883,
      "step": 357
    },
    {
      "epoch": 2.29,
      "grad_norm": 1.4863020248858034,
      "learning_rate": 4.593469802671951e-06,
      "loss": 0.0729,
      "step": 358
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.559787426657181,
      "learning_rate": 4.5785797276703075e-06,
      "loss": 0.0678,
      "step": 359
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.7039467911066226,
      "learning_rate": 4.563668051509864e-06,
      "loss": 0.0994,
      "step": 360
    },
    {
      "epoch": 2.31,
      "grad_norm": 1.7731905609375875,
      "learning_rate": 4.548735072832879e-06,
      "loss": 0.1051,
      "step": 361
    },
    {
      "epoch": 2.32,
      "grad_norm": 1.5882807257594604,
      "learning_rate": 4.533781090708244e-06,
      "loss": 0.0919,
      "step": 362
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.4184027450810133,
      "learning_rate": 4.518806404625495e-06,
      "loss": 0.0648,
      "step": 363
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.5279149470972586,
      "learning_rate": 4.503811314488816e-06,
      "loss": 0.0683,
      "step": 364
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.4225043393982648,
      "learning_rate": 4.48879612061103e-06,
      "loss": 0.0767,
      "step": 365
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.459950562847426,
      "learning_rate": 4.473761123707584e-06,
      "loss": 0.0894,
      "step": 366
    },
    {
      "epoch": 2.35,
      "grad_norm": 1.4577868793787234,
      "learning_rate": 4.458706624890534e-06,
      "loss": 0.0863,
      "step": 367
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.4796537233042946,
      "learning_rate": 4.443632925662504e-06,
      "loss": 0.0741,
      "step": 368
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.3954928355766592,
      "learning_rate": 4.428540327910652e-06,
      "loss": 0.0744,
      "step": 369
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.3599680592870689,
      "learning_rate": 4.41342913390063e-06,
      "loss": 0.0769,
      "step": 370
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.3432745418057788,
      "learning_rate": 4.398299646270518e-06,
      "loss": 0.0469,
      "step": 371
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.3618470064275703,
      "learning_rate": 4.3831521680247765e-06,
      "loss": 0.0801,
      "step": 372
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.335796702130415,
      "learning_rate": 4.3679870025281644e-06,
      "loss": 0.0733,
      "step": 373
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.4972832230856088,
      "learning_rate": 4.352804453499677e-06,
      "loss": 0.0701,
      "step": 374
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.566065032089918,
      "learning_rate": 4.3376048250064525e-06,
      "loss": 0.0878,
      "step": 375
    },
    {
      "epoch": 2.41,
      "grad_norm": 1.9440213727928068,
      "learning_rate": 4.322388421457687e-06,
      "loss": 0.1117,
      "step": 376
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.6075238415353954,
      "learning_rate": 4.30715554759854e-06,
      "loss": 0.0834,
      "step": 377
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.4916088511503431,
      "learning_rate": 4.2919065085040285e-06,
      "loss": 0.0674,
      "step": 378
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.5864994661720633,
      "learning_rate": 4.276641609572911e-06,
      "loss": 0.0748,
      "step": 379
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.5877220777630603,
      "learning_rate": 4.261361156521586e-06,
      "loss": 0.0673,
      "step": 380
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.5952419379371832,
      "learning_rate": 4.246065455377956e-06,
      "loss": 0.0723,
      "step": 381
    },
    {
      "epoch": 2.45,
      "grad_norm": 1.3659196106216351,
      "learning_rate": 4.230754812475306e-06,
      "loss": 0.0678,
      "step": 382
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.5150129082800552,
      "learning_rate": 4.215429534446161e-06,
      "loss": 0.0873,
      "step": 383
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.328231293830379,
      "learning_rate": 4.200089928216156e-06,
      "loss": 0.0469,
      "step": 384
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.3644208951729628,
      "learning_rate": 4.1847363009978776e-06,
      "loss": 0.0558,
      "step": 385
    },
    {
      "epoch": 2.47,
      "grad_norm": 1.3631190033041523,
      "learning_rate": 4.169368960284718e-06,
      "loss": 0.0642,
      "step": 386
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.3380963083993747,
      "learning_rate": 4.153988213844717e-06,
      "loss": 0.0651,
      "step": 387
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.6232655969183094,
      "learning_rate": 4.138594369714394e-06,
      "loss": 0.0768,
      "step": 388
    },
    {
      "epoch": 2.49,
      "grad_norm": 1.738615411907169,
      "learning_rate": 4.123187736192583e-06,
      "loss": 0.0631,
      "step": 389
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.3541880062004612,
      "learning_rate": 4.107768621834257e-06,
      "loss": 0.0646,
      "step": 390
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.5937219608859803,
      "learning_rate": 4.092337335444343e-06,
      "loss": 0.0543,
      "step": 391
    },
    {
      "epoch": 2.51,
      "grad_norm": 1.2652905019689844,
      "learning_rate": 4.076894186071548e-06,
      "loss": 0.0624,
      "step": 392
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.585692487433423,
      "learning_rate": 4.061439483002161e-06,
      "loss": 0.056,
      "step": 393
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.4234670685396151,
      "learning_rate": 4.045973535753863e-06,
      "loss": 0.0563,
      "step": 394
    },
    {
      "epoch": 2.53,
      "grad_norm": 1.3660930904759059,
      "learning_rate": 4.030496654069524e-06,
      "loss": 0.0683,
      "step": 395
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.5737973741759328,
      "learning_rate": 4.015009147911007e-06,
      "loss": 0.072,
      "step": 396
    },
    {
      "epoch": 2.54,
      "grad_norm": 1.4929980457042287,
      "learning_rate": 3.9995113274529506e-06,
      "loss": 0.0735,
      "step": 397
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.4565626149680446,
      "learning_rate": 3.984003503076566e-06,
      "loss": 0.0871,
      "step": 398
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.3794177772470047,
      "learning_rate": 3.968485985363416e-06,
      "loss": 0.0614,
      "step": 399
    },
    {
      "epoch": 2.56,
      "grad_norm": 1.44393677049457,
      "learning_rate": 3.952959085089193e-06,
      "loss": 0.0632,
      "step": 400
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.3495277752899084,
      "learning_rate": 3.937423113217505e-06,
      "loss": 0.0619,
      "step": 401
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.2902982047483085,
      "learning_rate": 3.92187838089363e-06,
      "loss": 0.0624,
      "step": 402
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.4252355923208517,
      "learning_rate": 3.9063251994383055e-06,
      "loss": 0.0619,
      "step": 403
    },
    {
      "epoch": 2.59,
      "grad_norm": 1.3702111372981403,
      "learning_rate": 3.8907638803414774e-06,
      "loss": 0.0659,
      "step": 404
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.358396927631374,
      "learning_rate": 3.875194735256067e-06,
      "loss": 0.0713,
      "step": 405
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.4249331438647441,
      "learning_rate": 3.859618075991735e-06,
      "loss": 0.0658,
      "step": 406
    },
    {
      "epoch": 2.61,
      "grad_norm": 1.3059396981665108,
      "learning_rate": 3.844034214508625e-06,
      "loss": 0.0751,
      "step": 407
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.3155398633895121,
      "learning_rate": 3.828443462911128e-06,
      "loss": 0.0679,
      "step": 408
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4663166371568865,
      "learning_rate": 3.8128461334416223e-06,
      "loss": 0.0669,
      "step": 409
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.4387435039719887,
      "learning_rate": 3.7972425384742264e-06,
      "loss": 0.0578,
      "step": 410
    },
    {
      "epoch": 2.63,
      "grad_norm": 1.5441681248495072,
      "learning_rate": 3.781632990508541e-06,
      "loss": 0.0668,
      "step": 411
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.509803326861479,
      "learning_rate": 3.766017802163386e-06,
      "loss": 0.0509,
      "step": 412
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.4135151374037613,
      "learning_rate": 3.7503972861705478e-06,
      "loss": 0.0574,
      "step": 413
    },
    {
      "epoch": 2.65,
      "grad_norm": 1.5520725445811439,
      "learning_rate": 3.7347717553685084e-06,
      "loss": 0.0861,
      "step": 414
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.7837110913227487,
      "learning_rate": 3.7191415226961867e-06,
      "loss": 0.0617,
      "step": 415
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.384690904784385,
      "learning_rate": 3.703506901186665e-06,
      "loss": 0.0421,
      "step": 416
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.58160512232892,
      "learning_rate": 3.6878682039609253e-06,
      "loss": 0.0465,
      "step": 417
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.4583873146929784,
      "learning_rate": 3.6722257442215736e-06,
      "loss": 0.054,
      "step": 418
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.392182390968472,
      "learning_rate": 3.6565798352465697e-06,
      "loss": 0.0648,
      "step": 419
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.410291768179273,
      "learning_rate": 3.640930790382953e-06,
      "loss": 0.0539,
      "step": 420
    },
    {
      "epoch": 2.7,
      "grad_norm": 2.2129885462771672,
      "learning_rate": 3.625278923040567e-06,
      "loss": 0.119,
      "step": 421
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.3548595450124818,
      "learning_rate": 3.6096245466857808e-06,
      "loss": 0.0741,
      "step": 422
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.3440278963530723,
      "learning_rate": 3.5939679748352146e-06,
      "loss": 0.0773,
      "step": 423
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.306530229127965,
      "learning_rate": 3.578309521049456e-06,
      "loss": 0.0562,
      "step": 424
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.4782640258474615,
      "learning_rate": 3.562649498926785e-06,
      "loss": 0.0696,
      "step": 425
    },
    {
      "epoch": 2.73,
      "grad_norm": 1.480966118357646,
      "learning_rate": 3.546988222096891e-06,
      "loss": 0.0611,
      "step": 426
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.465534512953903,
      "learning_rate": 3.531326004214592e-06,
      "loss": 0.07,
      "step": 427
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.3384172211856384,
      "learning_rate": 3.515663158953552e-06,
      "loss": 0.0623,
      "step": 428
    },
    {
      "epoch": 2.75,
      "grad_norm": 1.325181333356099,
      "learning_rate": 3.5e-06,
      "loss": 0.0679,
      "step": 429
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.1268047140927235,
      "learning_rate": 3.484336841046448e-06,
      "loss": 0.0549,
      "step": 430
    },
    {
      "epoch": 2.76,
      "grad_norm": 1.287545657132799,
      "learning_rate": 3.468673995785409e-06,
      "loss": 0.0554,
      "step": 431
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.480635441568218,
      "learning_rate": 3.4530117779031096e-06,
      "loss": 0.0721,
      "step": 432
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.251760291533078,
      "learning_rate": 3.4373505010732152e-06,
      "loss": 0.0488,
      "step": 433
    },
    {
      "epoch": 2.78,
      "grad_norm": 1.3619293632526979,
      "learning_rate": 3.4216904789505444e-06,
      "loss": 0.0567,
      "step": 434
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.3634889684832445,
      "learning_rate": 3.4060320251647866e-06,
      "loss": 0.0505,
      "step": 435
    },
    {
      "epoch": 2.79,
      "grad_norm": 1.43374246584491,
      "learning_rate": 3.3903754533142195e-06,
      "loss": 0.0658,
      "step": 436
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.553124280424844,
      "learning_rate": 3.374721076959433e-06,
      "loss": 0.0609,
      "step": 437
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.7225796812921945,
      "learning_rate": 3.359069209617048e-06,
      "loss": 0.1126,
      "step": 438
    },
    {
      "epoch": 2.81,
      "grad_norm": 1.374779091521454,
      "learning_rate": 3.3434201647534306e-06,
      "loss": 0.0529,
      "step": 439
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.4196983369781266,
      "learning_rate": 3.3277742557784263e-06,
      "loss": 0.0468,
      "step": 440
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.475726843734466,
      "learning_rate": 3.312131796039074e-06,
      "loss": 0.0546,
      "step": 441
    },
    {
      "epoch": 2.83,
      "grad_norm": 1.303979627867413,
      "learning_rate": 3.296493098813335e-06,
      "loss": 0.0628,
      "step": 442
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.3984315031120913,
      "learning_rate": 3.280858477303813e-06,
      "loss": 0.0573,
      "step": 443
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.5264883558014777,
      "learning_rate": 3.265228244631491e-06,
      "loss": 0.0669,
      "step": 444
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.3900299809827257,
      "learning_rate": 3.2496027138294534e-06,
      "loss": 0.06,
      "step": 445
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.621781645789055,
      "learning_rate": 3.2339821978366144e-06,
      "loss": 0.087,
      "step": 446
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.406176966599867,
      "learning_rate": 3.2183670094914596e-06,
      "loss": 0.0549,
      "step": 447
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.4185651707707967,
      "learning_rate": 3.2027574615257726e-06,
      "loss": 0.0672,
      "step": 448
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.4517014206041816,
      "learning_rate": 3.1871538665583784e-06,
      "loss": 0.0693,
      "step": 449
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.3197298201511651,
      "learning_rate": 3.171556537088873e-06,
      "loss": 0.051,
      "step": 450
    },
    {
      "epoch": 2.89,
      "grad_norm": 1.4027977888678176,
      "learning_rate": 3.155965785491375e-06,
      "loss": 0.0628,
      "step": 451
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.3210542029533012,
      "learning_rate": 3.140381924008266e-06,
      "loss": 0.0537,
      "step": 452
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.3983814988636492,
      "learning_rate": 3.1248052647439327e-06,
      "loss": 0.0607,
      "step": 453
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.3381469983224843,
      "learning_rate": 3.109236119658523e-06,
      "loss": 0.0629,
      "step": 454
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.363641990737864,
      "learning_rate": 3.0936748005616936e-06,
      "loss": 0.0602,
      "step": 455
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.337994198492209,
      "learning_rate": 3.0781216191063695e-06,
      "loss": 0.0799,
      "step": 456
    },
    {
      "epoch": 2.93,
      "grad_norm": 1.3335030606735412,
      "learning_rate": 3.0625768867824957e-06,
      "loss": 0.0545,
      "step": 457
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.4769773993333548,
      "learning_rate": 3.047040914910806e-06,
      "loss": 0.0783,
      "step": 458
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.3197947799572127,
      "learning_rate": 3.0315140146365854e-06,
      "loss": 0.0584,
      "step": 459
    },
    {
      "epoch": 2.95,
      "grad_norm": 1.4192189944677764,
      "learning_rate": 3.015996496923435e-06,
      "loss": 0.0531,
      "step": 460
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.419406225194028,
      "learning_rate": 3.00048867254705e-06,
      "loss": 0.0583,
      "step": 461
    },
    {
      "epoch": 2.96,
      "grad_norm": 1.4550693486915325,
      "learning_rate": 2.9849908520889936e-06,
      "loss": 0.0714,
      "step": 462
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.34618411271052,
      "learning_rate": 2.9695033459304766e-06,
      "loss": 0.0743,
      "step": 463
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.4148143017886536,
      "learning_rate": 2.954026464246138e-06,
      "loss": 0.0687,
      "step": 464
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.494976547180411,
      "learning_rate": 2.9385605169978387e-06,
      "loss": 0.0905,
      "step": 465
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.4492598962322225,
      "learning_rate": 2.923105813928453e-06,
      "loss": 0.0686,
      "step": 466
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.254765150326393,
      "learning_rate": 2.907662664555658e-06,
      "loss": 0.0492,
      "step": 467
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.3978603716574316,
      "learning_rate": 2.8922313781657437e-06,
      "loss": 0.0733,
      "step": 468
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.3568449348317895,
      "learning_rate": 2.876812263807417e-06,
      "loss": 0.0692,
      "step": 469
    },
    {
      "epoch": 3.01,
      "grad_norm": 1.5106491687723738,
      "learning_rate": 2.861405630285606e-06,
      "loss": 0.0533,
      "step": 470
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.5179556314161937,
      "learning_rate": 2.8460117861552833e-06,
      "loss": 0.055,
      "step": 471
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.5694567578058851,
      "learning_rate": 2.8306310397152817e-06,
      "loss": 0.0598,
      "step": 472
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.2954069120683234,
      "learning_rate": 2.815263699002124e-06,
      "loss": 0.0538,
      "step": 473
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.3105848212645954,
      "learning_rate": 2.799910071783845e-06,
      "loss": 0.0462,
      "step": 474
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.3098901122342839,
      "learning_rate": 2.7845704655538383e-06,
      "loss": 0.0471,
      "step": 475
    },
    {
      "epoch": 3.05,
      "grad_norm": 1.3134707970328,
      "learning_rate": 2.7692451875246956e-06,
      "loss": 0.0683,
      "step": 476
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.1406228014597528,
      "learning_rate": 2.7539345446220444e-06,
      "loss": 0.0477,
      "step": 477
    },
    {
      "epoch": 3.06,
      "grad_norm": 1.0924596106691973,
      "learning_rate": 2.7386388434784143e-06,
      "loss": 0.049,
      "step": 478
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.257749327487133,
      "learning_rate": 2.723358390427089e-06,
      "loss": 0.0481,
      "step": 479
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.2601405115815025,
      "learning_rate": 2.708093491495973e-06,
      "loss": 0.0464,
      "step": 480
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.2396907483275874,
      "learning_rate": 2.6928444524014595e-06,
      "loss": 0.0526,
      "step": 481
    },
    {
      "epoch": 3.09,
      "grad_norm": 1.220905178010921,
      "learning_rate": 2.6776115785423123e-06,
      "loss": 0.0469,
      "step": 482
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.134133024109752,
      "learning_rate": 2.6623951749935487e-06,
      "loss": 0.0334,
      "step": 483
    },
    {
      "epoch": 3.1,
      "grad_norm": 1.2225938422846696,
      "learning_rate": 2.6471955465003237e-06,
      "loss": 0.0405,
      "step": 484
    },
    {
      "epoch": 3.11,
      "grad_norm": 1.192978092896825,
      "learning_rate": 2.6320129974718355e-06,
      "loss": 0.0511,
      "step": 485
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.2144105464154307,
      "learning_rate": 2.616847831975224e-06,
      "loss": 0.043,
      "step": 486
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.3270694940879673,
      "learning_rate": 2.601700353729481e-06,
      "loss": 0.044,
      "step": 487
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.2909647509362592,
      "learning_rate": 2.58657086609937e-06,
      "loss": 0.0382,
      "step": 488
    },
    {
      "epoch": 3.13,
      "grad_norm": 1.557891789297615,
      "learning_rate": 2.5714596720893473e-06,
      "loss": 0.0484,
      "step": 489
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.3632837500193322,
      "learning_rate": 2.5563670743374973e-06,
      "loss": 0.0406,
      "step": 490
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.128374695442261,
      "learning_rate": 2.5412933751094662e-06,
      "loss": 0.0273,
      "step": 491
    },
    {
      "epoch": 3.15,
      "grad_norm": 1.2324598795816455,
      "learning_rate": 2.5262388762924157e-06,
      "loss": 0.0341,
      "step": 492
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.343419423571675,
      "learning_rate": 2.5112038793889706e-06,
      "loss": 0.0373,
      "step": 493
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.1526235554463253,
      "learning_rate": 2.496188685511185e-06,
      "loss": 0.0276,
      "step": 494
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.201453054797899,
      "learning_rate": 2.481193595374505e-06,
      "loss": 0.0439,
      "step": 495
    },
    {
      "epoch": 3.18,
      "grad_norm": 1.4351161008323083,
      "learning_rate": 2.4662189092917563e-06,
      "loss": 0.0429,
      "step": 496
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.5663274417487951,
      "learning_rate": 2.4512649271671214e-06,
      "loss": 0.0554,
      "step": 497
    },
    {
      "epoch": 3.19,
      "grad_norm": 1.3595537176740793,
      "learning_rate": 2.436331948490136e-06,
      "loss": 0.0393,
      "step": 498
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.4896573920852347,
      "learning_rate": 2.4214202723296924e-06,
      "loss": 0.0405,
      "step": 499
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.2767253714566218,
      "learning_rate": 2.4065301973280486e-06,
      "loss": 0.0399,
      "step": 500
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.2378296334262433,
      "learning_rate": 2.391662021694847e-06,
      "loss": 0.0405,
      "step": 501
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.3468312150337725,
      "learning_rate": 2.3768160432011395e-06,
      "loss": 0.0404,
      "step": 502
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.256283638660665,
      "learning_rate": 2.3619925591734323e-06,
      "loss": 0.0333,
      "step": 503
    },
    {
      "epoch": 3.23,
      "grad_norm": 1.1908716392621157,
      "learning_rate": 2.3471918664877217e-06,
      "loss": 0.0338,
      "step": 504
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.0561449925755435,
      "learning_rate": 2.332414261563553e-06,
      "loss": 0.0252,
      "step": 505
    },
    {
      "epoch": 3.24,
      "grad_norm": 1.0161119154501226,
      "learning_rate": 2.317660040358085e-06,
      "loss": 0.0213,
      "step": 506
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.3396465580980972,
      "learning_rate": 2.3029294983601598e-06,
      "loss": 0.033,
      "step": 507
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.0469883632933366,
      "learning_rate": 2.2882229305843866e-06,
      "loss": 0.0275,
      "step": 508
    },
    {
      "epoch": 3.26,
      "grad_norm": 1.1713965135820492,
      "learning_rate": 2.2735406315652323e-06,
      "loss": 0.0334,
      "step": 509
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.0156278207651381,
      "learning_rate": 2.258882895351125e-06,
      "loss": 0.0316,
      "step": 510
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.9597778807188877,
      "learning_rate": 2.2442500154985643e-06,
      "loss": 0.0211,
      "step": 511
    },
    {
      "epoch": 3.28,
      "grad_norm": 1.056684870602395,
      "learning_rate": 2.229642285066236e-06,
      "loss": 0.031,
      "step": 512
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.2111529373318868,
      "learning_rate": 2.215059996609154e-06,
      "loss": 0.034,
      "step": 513
    },
    {
      "epoch": 3.29,
      "grad_norm": 1.0403499427426597,
      "learning_rate": 2.200503442172792e-06,
      "loss": 0.0266,
      "step": 514
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.023896201002211,
      "learning_rate": 2.185972913287241e-06,
      "loss": 0.0245,
      "step": 515
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.09541799539132,
      "learning_rate": 2.1714687009613628e-06,
      "loss": 0.0325,
      "step": 516
    },
    {
      "epoch": 3.31,
      "grad_norm": 1.2592978524239982,
      "learning_rate": 2.156991095676971e-06,
      "loss": 0.038,
      "step": 517
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.0543296286017256,
      "learning_rate": 2.1425403873830083e-06,
      "loss": 0.0346,
      "step": 518
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.0001186604444399,
      "learning_rate": 2.1281168654897376e-06,
      "loss": 0.0208,
      "step": 519
    },
    {
      "epoch": 3.33,
      "grad_norm": 1.10411761212027,
      "learning_rate": 2.113720818862951e-06,
      "loss": 0.026,
      "step": 520
    },
    {
      "epoch": 3.34,
      "grad_norm": 1.076138981995205,
      "learning_rate": 2.099352535818182e-06,
      "loss": 0.0244,
      "step": 521
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.0196811903638796,
      "learning_rate": 2.085012304114933e-06,
      "loss": 0.0296,
      "step": 522
    },
    {
      "epoch": 3.35,
      "grad_norm": 1.0314845239229538,
      "learning_rate": 2.070700410950906e-06,
      "loss": 0.0324,
      "step": 523
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.5222487359020664,
      "learning_rate": 2.0564171429562587e-06,
      "loss": 0.029,
      "step": 524
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.1000193388139592,
      "learning_rate": 2.042162786187862e-06,
      "loss": 0.0252,
      "step": 525
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.0947526780154235,
      "learning_rate": 2.027937626123565e-06,
      "loss": 0.027,
      "step": 526
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.1771718630654349,
      "learning_rate": 2.0137419476564896e-06,
      "loss": 0.0157,
      "step": 527
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.9908516635250878,
      "learning_rate": 1.9995760350893098e-06,
      "loss": 0.0286,
      "step": 528
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.9791274664173555,
      "learning_rate": 1.985440172128573e-06,
      "loss": 0.0243,
      "step": 529
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.3203297232694446,
      "learning_rate": 1.9713346418790058e-06,
      "loss": 0.0261,
      "step": 530
    },
    {
      "epoch": 3.4,
      "grad_norm": 1.13911901666537,
      "learning_rate": 1.957259726837849e-06,
      "loss": 0.029,
      "step": 531
    },
    {
      "epoch": 3.41,
      "grad_norm": 1.048570931624032,
      "learning_rate": 1.9432157088892064e-06,
      "loss": 0.0374,
      "step": 532
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.2103418865964828,
      "learning_rate": 1.9292028692983824e-06,
      "loss": 0.0306,
      "step": 533
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.029334123069811,
      "learning_rate": 1.91522148870627e-06,
      "loss": 0.0217,
      "step": 534
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.0198377079922467,
      "learning_rate": 1.9012718471237144e-06,
      "loss": 0.0251,
      "step": 535
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.166818734002997,
      "learning_rate": 1.887354223925911e-06,
      "loss": 0.0215,
      "step": 536
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.085856728273484,
      "learning_rate": 1.87346889784681e-06,
      "loss": 0.0241,
      "step": 537
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.9340299089299968,
      "learning_rate": 1.8596161469735374e-06,
      "loss": 0.0222,
      "step": 538
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.9582885459137026,
      "learning_rate": 1.8457962487408175e-06,
      "loss": 0.0306,
      "step": 539
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.7818765763548071,
      "learning_rate": 1.8320094799254222e-06,
      "loss": 0.0108,
      "step": 540
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.8674372954189106,
      "learning_rate": 1.8182561166406308e-06,
      "loss": 0.0173,
      "step": 541
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.9225859827659534,
      "learning_rate": 1.8045364343306915e-06,
      "loss": 0.0189,
      "step": 542
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.9050743041899307,
      "learning_rate": 1.7908507077653124e-06,
      "loss": 0.0185,
      "step": 543
    },
    {
      "epoch": 3.49,
      "grad_norm": 0.9749205541180587,
      "learning_rate": 1.7771992110341533e-06,
      "loss": 0.0219,
      "step": 544
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.0069281642083787,
      "learning_rate": 1.7635822175413446e-06,
      "loss": 0.0231,
      "step": 545
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.8029510509892319,
      "learning_rate": 1.7500000000000008e-06,
      "loss": 0.02,
      "step": 546
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.0472669216241044,
      "learning_rate": 1.7364528304267646e-06,
      "loss": 0.0169,
      "step": 547
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.9119333217497559,
      "learning_rate": 1.7229409801363635e-06,
      "loss": 0.0194,
      "step": 548
    },
    {
      "epoch": 3.52,
      "grad_norm": 1.0558846501029286,
      "learning_rate": 1.7094647197361656e-06,
      "loss": 0.018,
      "step": 549
    },
    {
      "epoch": 3.53,
      "grad_norm": 1.0384510779811111,
      "learning_rate": 1.6960243191207686e-06,
      "loss": 0.0182,
      "step": 550
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.8341867375991961,
      "learning_rate": 1.6826200474665891e-06,
      "loss": 0.0204,
      "step": 551
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.1531429797253752,
      "learning_rate": 1.669252173226479e-06,
      "loss": 0.0221,
      "step": 552
    },
    {
      "epoch": 3.54,
      "grad_norm": 1.046052155783273,
      "learning_rate": 1.6559209641243388e-06,
      "loss": 0.0251,
      "step": 553
    },
    {
      "epoch": 3.55,
      "grad_norm": 1.0056558783597644,
      "learning_rate": 1.642626687149765e-06,
      "loss": 0.0297,
      "step": 554
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.0388548400576412,
      "learning_rate": 1.629369608552696e-06,
      "loss": 0.0207,
      "step": 555
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.9676580813043666,
      "learning_rate": 1.6161499938380873e-06,
      "loss": 0.0207,
      "step": 556
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.910615049954989,
      "learning_rate": 1.6029681077605864e-06,
      "loss": 0.0194,
      "step": 557
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.8154656860131219,
      "learning_rate": 1.5898242143192336e-06,
      "loss": 0.0179,
      "step": 558
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.8639595706943999,
      "learning_rate": 1.576718576752179e-06,
      "loss": 0.0156,
      "step": 559
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.9189700127422249,
      "learning_rate": 1.5636514575314024e-06,
      "loss": 0.0192,
      "step": 560
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.9330229070746087,
      "learning_rate": 1.550623118357463e-06,
      "loss": 0.0224,
      "step": 561
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.9035600157428585,
      "learning_rate": 1.5376338201542538e-06,
      "loss": 0.0198,
      "step": 562
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.8321348110175191,
      "learning_rate": 1.5246838230637831e-06,
      "loss": 0.0222,
      "step": 563
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.8596805606431125,
      "learning_rate": 1.511773386440955e-06,
      "loss": 0.0195,
      "step": 564
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.9537922409494741,
      "learning_rate": 1.4989027688483808e-06,
      "loss": 0.0205,
      "step": 565
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.7806033695969874,
      "learning_rate": 1.4860722280512022e-06,
      "loss": 0.0168,
      "step": 566
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.8391056339361249,
      "learning_rate": 1.473282021011924e-06,
      "loss": 0.0192,
      "step": 567
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.9361363433592843,
      "learning_rate": 1.4605324038852707e-06,
      "loss": 0.0137,
      "step": 568
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.9165200492322241,
      "learning_rate": 1.4478236320130554e-06,
      "loss": 0.0183,
      "step": 569
    },
    {
      "epoch": 3.65,
      "grad_norm": 1.0311350391531426,
      "learning_rate": 1.4351559599190708e-06,
      "loss": 0.0278,
      "step": 570
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.002796327201144,
      "learning_rate": 1.4225296413039794e-06,
      "loss": 0.0207,
      "step": 571
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.8663434564935976,
      "learning_rate": 1.4099449290402492e-06,
      "loss": 0.0145,
      "step": 572
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.7829556851775933,
      "learning_rate": 1.3974020751670734e-06,
      "loss": 0.0119,
      "step": 573
    },
    {
      "epoch": 3.68,
      "grad_norm": 0.7838935562572859,
      "learning_rate": 1.3849013308853369e-06,
      "loss": 0.014,
      "step": 574
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.8089531524154776,
      "learning_rate": 1.3724429465525733e-06,
      "loss": 0.0172,
      "step": 575
    },
    {
      "epoch": 3.69,
      "grad_norm": 1.0476096426575607,
      "learning_rate": 1.360027171677957e-06,
      "loss": 0.0159,
      "step": 576
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.1492757887504752,
      "learning_rate": 1.3476542549173097e-06,
      "loss": 0.0509,
      "step": 577
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.9034375118433278,
      "learning_rate": 1.335324444068108e-06,
      "loss": 0.0221,
      "step": 578
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.8650183804828405,
      "learning_rate": 1.3230379860645363e-06,
      "loss": 0.0213,
      "step": 579
    },
    {
      "epoch": 3.72,
      "grad_norm": 1.0015608184240266,
      "learning_rate": 1.3107951269725286e-06,
      "loss": 0.018,
      "step": 580
    },
    {
      "epoch": 3.72,
      "grad_norm": 0.9368052136067349,
      "learning_rate": 1.2985961119848508e-06,
      "loss": 0.0197,
      "step": 581
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.959349340544164,
      "learning_rate": 1.28644118541618e-06,
      "loss": 0.0201,
      "step": 582
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.8723031906642617,
      "learning_rate": 1.2743305906982184e-06,
      "loss": 0.0186,
      "step": 583
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.9478403521550213,
      "learning_rate": 1.2622645703748163e-06,
      "loss": 0.0194,
      "step": 584
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.8494970290194015,
      "learning_rate": 1.2502433660971122e-06,
      "loss": 0.0186,
      "step": 585
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.8551225612629588,
      "learning_rate": 1.2382672186187003e-06,
      "loss": 0.017,
      "step": 586
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.8488361613538855,
      "learning_rate": 1.2263363677907975e-06,
      "loss": 0.0169,
      "step": 587
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.9808510003867794,
      "learning_rate": 1.214451052557453e-06,
      "loss": 0.0221,
      "step": 588
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.8108841674151576,
      "learning_rate": 1.202611510950747e-06,
      "loss": 0.0151,
      "step": 589
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.8886038022214681,
      "learning_rate": 1.1908179800860415e-06,
      "loss": 0.0182,
      "step": 590
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.9119958116007242,
      "learning_rate": 1.1790706961572176e-06,
      "loss": 0.0138,
      "step": 591
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.7574159445793109,
      "learning_rate": 1.167369894431949e-06,
      "loss": 0.0162,
      "step": 592
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.8830721642283024,
      "learning_rate": 1.1557158092469968e-06,
      "loss": 0.017,
      "step": 593
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.8752065538129817,
      "learning_rate": 1.1441086740035036e-06,
      "loss": 0.0312,
      "step": 594
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.7802871014880685,
      "learning_rate": 1.1325487211623343e-06,
      "loss": 0.014,
      "step": 595
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.822063110348498,
      "learning_rate": 1.121036182239403e-06,
      "loss": 0.0109,
      "step": 596
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.7316479690003538,
      "learning_rate": 1.1095712878010542e-06,
      "loss": 0.0104,
      "step": 597
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.7093603836144392,
      "learning_rate": 1.0981542674594327e-06,
      "loss": 0.0184,
      "step": 598
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.7794338741880872,
      "learning_rate": 1.08678534986789e-06,
      "loss": 0.0168,
      "step": 599
    },
    {
      "epoch": 3.85,
      "grad_norm": 1.0811588463022737,
      "learning_rate": 1.0754647627164022e-06,
      "loss": 0.0191,
      "step": 600
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.9111404871044384,
      "learning_rate": 1.064192732727016e-06,
      "loss": 0.0189,
      "step": 601
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.8869358514707676,
      "learning_rate": 1.0529694856493002e-06,
      "loss": 0.0265,
      "step": 602
    },
    {
      "epoch": 3.87,
      "grad_norm": 1.0300105674738138,
      "learning_rate": 1.0417952462558286e-06,
      "loss": 0.0174,
      "step": 603
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.9154400665993732,
      "learning_rate": 1.0306702383376813e-06,
      "loss": 0.0216,
      "step": 604
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.7790200245067616,
      "learning_rate": 1.0195946846999551e-06,
      "loss": 0.0207,
      "step": 605
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.9245289976289252,
      "learning_rate": 1.0085688071573086e-06,
      "loss": 0.0157,
      "step": 606
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.845440385987704,
      "learning_rate": 9.97592826529514e-07,
      "loss": 0.0168,
      "step": 607
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.6828942627029853,
      "learning_rate": 9.866669626370412e-07,
      "loss": 0.0136,
      "step": 608
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.7507599698976973,
      "learning_rate": 9.757914342966495e-07,
      "loss": 0.0159,
      "step": 609
    },
    {
      "epoch": 3.91,
      "grad_norm": 0.7608568860765801,
      "learning_rate": 9.649664593170062e-07,
      "loss": 0.0182,
      "step": 610
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.6964025046289981,
      "learning_rate": 9.541922544943295e-07,
      "loss": 0.0143,
      "step": 611
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.8040858287775848,
      "learning_rate": 9.434690356080394e-07,
      "loss": 0.0226,
      "step": 612
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.8367198110469669,
      "learning_rate": 9.327970174164409e-07,
      "loss": 0.0146,
      "step": 613
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.8193340863552715,
      "learning_rate": 9.221764136524202e-07,
      "loss": 0.0192,
      "step": 614
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.7265415845223846,
      "learning_rate": 9.116074370191705e-07,
      "loss": 0.0142,
      "step": 615
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.69231773481907,
      "learning_rate": 9.010902991859196e-07,
      "loss": 0.0127,
      "step": 616
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.7678911416679844,
      "learning_rate": 8.906252107837054e-07,
      "loss": 0.0135,
      "step": 617
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.9669858607123597,
      "learning_rate": 8.802123814011458e-07,
      "loss": 0.0203,
      "step": 618
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.7067832443423266,
      "learning_rate": 8.698520195802499e-07,
      "loss": 0.0216,
      "step": 619
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.9954743141268118,
      "learning_rate": 8.595443328122345e-07,
      "loss": 0.0185,
      "step": 620
    },
    {
      "epoch": 3.98,
      "grad_norm": 0.9698572023393062,
      "learning_rate": 8.492895275333705e-07,
      "loss": 0.0258,
      "step": 621
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.8472663094138484,
      "learning_rate": 8.390878091208544e-07,
      "loss": 0.0175,
      "step": 622
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.6974557493903148,
      "learning_rate": 8.289393818886837e-07,
      "loss": 0.0148,
      "step": 623
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.8038533284087307,
      "learning_rate": 8.188444490835774e-07,
      "loss": 0.0214,
      "step": 624
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.7210389335963133,
      "learning_rate": 8.088032128808952e-07,
      "loss": 0.0196,
      "step": 625
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.7781631119204625,
      "learning_rate": 7.988158743805973e-07,
      "loss": 0.0127,
      "step": 626
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.8340070618057464,
      "learning_rate": 7.888826336032093e-07,
      "loss": 0.0146,
      "step": 627
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.8166490795788324,
      "learning_rate": 7.790036894858198e-07,
      "loss": 0.0145,
      "step": 628
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.7195731647426649,
      "learning_rate": 7.691792398780962e-07,
      "loss": 0.0119,
      "step": 629
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.7441257497162197,
      "learning_rate": 7.594094815383223e-07,
      "loss": 0.0125,
      "step": 630
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.8162804472924826,
      "learning_rate": 7.496946101294585e-07,
      "loss": 0.0096,
      "step": 631
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.7311628316123926,
      "learning_rate": 7.400348202152192e-07,
      "loss": 0.0191,
      "step": 632
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.6708461799554153,
      "learning_rate": 7.304303052561841e-07,
      "loss": 0.0128,
      "step": 633
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.6419666578612593,
      "learning_rate": 7.208812576059113e-07,
      "loss": 0.014,
      "step": 634
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.571786193933189,
      "learning_rate": 7.113878685070994e-07,
      "loss": 0.0126,
      "step": 635
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.47401578964273894,
      "learning_rate": 7.019503280877466e-07,
      "loss": 0.0089,
      "step": 636
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.6164893786390164,
      "learning_rate": 6.925688253573465e-07,
      "loss": 0.0145,
      "step": 637
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.5908416108258375,
      "learning_rate": 6.832435482031064e-07,
      "loss": 0.0102,
      "step": 638
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.8989690237248916,
      "learning_rate": 6.73974683386176e-07,
      "loss": 0.0121,
      "step": 639
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.5982510810482049,
      "learning_rate": 6.647624165379173e-07,
      "loss": 0.0103,
      "step": 640
    },
    {
      "epoch": 4.11,
      "grad_norm": 0.5954747425687978,
      "learning_rate": 6.55606932156175e-07,
      "loss": 0.0147,
      "step": 641
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.5239184374248897,
      "learning_rate": 6.465084136015951e-07,
      "loss": 0.0119,
      "step": 642
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.7241845644184443,
      "learning_rate": 6.374670430939404e-07,
      "loss": 0.0134,
      "step": 643
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.6317510180810488,
      "learning_rate": 6.284830017084488e-07,
      "loss": 0.0103,
      "step": 644
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.6913499172359965,
      "learning_rate": 6.195564693722027e-07,
      "loss": 0.0118,
      "step": 645
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.5052384619401886,
      "learning_rate": 6.106876248605299e-07,
      "loss": 0.0113,
      "step": 646
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.5731989344051904,
      "learning_rate": 6.018766457934177e-07,
      "loss": 0.0074,
      "step": 647
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.6142999396736151,
      "learning_rate": 5.931237086319592e-07,
      "loss": 0.0084,
      "step": 648
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.823064223506891,
      "learning_rate": 5.844289886748196e-07,
      "loss": 0.0105,
      "step": 649
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.4979417009348811,
      "learning_rate": 5.757926600547231e-07,
      "loss": 0.0084,
      "step": 650
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.5818651933396465,
      "learning_rate": 5.672148957349661e-07,
      "loss": 0.0122,
      "step": 651
    },
    {
      "epoch": 4.18,
      "grad_norm": 0.6507789337199134,
      "learning_rate": 5.586958675059548e-07,
      "loss": 0.0122,
      "step": 652
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.6049124608829605,
      "learning_rate": 5.502357459817639e-07,
      "loss": 0.013,
      "step": 653
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.7753742285832561,
      "learning_rate": 5.418347005967189e-07,
      "loss": 0.0091,
      "step": 654
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.7274448172473815,
      "learning_rate": 5.334928996020013e-07,
      "loss": 0.0087,
      "step": 655
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.6502094548231069,
      "learning_rate": 5.252105100622848e-07,
      "loss": 0.0106,
      "step": 656
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.40178486721608847,
      "learning_rate": 5.169876978523828e-07,
      "loss": 0.007,
      "step": 657
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.659442151178035,
      "learning_rate": 5.088246276539292e-07,
      "loss": 0.0113,
      "step": 658
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.46880646234385925,
      "learning_rate": 5.0072146295208e-07,
      "loss": 0.0073,
      "step": 659
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.7953540459085934,
      "learning_rate": 4.926783660322411e-07,
      "loss": 0.0085,
      "step": 660
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.48150007551722795,
      "learning_rate": 4.846954979768149e-07,
      "loss": 0.0057,
      "step": 661
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.6684968423071089,
      "learning_rate": 4.7677301866197455e-07,
      "loss": 0.0072,
      "step": 662
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.6745759023213915,
      "learning_rate": 4.6891108675446453e-07,
      "loss": 0.0076,
      "step": 663
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.43582709520930096,
      "learning_rate": 4.611098597084226e-07,
      "loss": 0.0074,
      "step": 664
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.5144595833892877,
      "learning_rate": 4.533694937622227e-07,
      "loss": 0.0088,
      "step": 665
    },
    {
      "epoch": 4.27,
      "grad_norm": 0.46455722697398655,
      "learning_rate": 4.456901439353499e-07,
      "loss": 0.0089,
      "step": 666
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.38039382254115717,
      "learning_rate": 4.3807196402529535e-07,
      "loss": 0.0059,
      "step": 667
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.5159499420296817,
      "learning_rate": 4.3051510660447336e-07,
      "loss": 0.0087,
      "step": 668
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.7569018705460623,
      "learning_rate": 4.2301972301716934e-07,
      "loss": 0.0108,
      "step": 669
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.3133032924877106,
      "learning_rate": 4.155859633765044e-07,
      "loss": 0.0067,
      "step": 670
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.44083154398560964,
      "learning_rate": 4.0821397656143503e-07,
      "loss": 0.0054,
      "step": 671
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.4568048829370673,
      "learning_rate": 4.009039102137657e-07,
      "loss": 0.0077,
      "step": 672
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.3862359262132665,
      "learning_rate": 3.9365591073519387e-07,
      "loss": 0.0082,
      "step": 673
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.41096373612079085,
      "learning_rate": 3.8647012328438085e-07,
      "loss": 0.0103,
      "step": 674
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.4363393359078843,
      "learning_rate": 3.793466917740402e-07,
      "loss": 0.0049,
      "step": 675
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.5352286390397114,
      "learning_rate": 3.7228575886805744e-07,
      "loss": 0.007,
      "step": 676
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.4879159236521443,
      "learning_rate": 3.6528746597863283e-07,
      "loss": 0.0061,
      "step": 677
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.4921806354878511,
      "learning_rate": 3.583519532634516e-07,
      "loss": 0.0085,
      "step": 678
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.45260045929318704,
      "learning_rate": 3.514793596228702e-07,
      "loss": 0.0096,
      "step": 679
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.5256395174007804,
      "learning_rate": 3.44669822697144e-07,
      "loss": 0.0078,
      "step": 680
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.47644883575785074,
      "learning_rate": 3.3792347886366265e-07,
      "loss": 0.0057,
      "step": 681
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.48359867196539846,
      "learning_rate": 3.31240463234221e-07,
      "loss": 0.0077,
      "step": 682
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.4273121166226652,
      "learning_rate": 3.2462090965231767e-07,
      "loss": 0.0025,
      "step": 683
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.37391829672343596,
      "learning_rate": 3.180649506904667e-07,
      "loss": 0.0089,
      "step": 684
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.29650839546439783,
      "learning_rate": 3.1157271764755085e-07,
      "loss": 0.006,
      "step": 685
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.4198572707034165,
      "learning_rate": 3.0514434054618216e-07,
      "loss": 0.0055,
      "step": 686
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.6603558563893405,
      "learning_rate": 2.987799481301091e-07,
      "loss": 0.0088,
      "step": 687
    },
    {
      "epoch": 4.41,
      "grad_norm": 0.4060296688138882,
      "learning_rate": 2.924796678616297e-07,
      "loss": 0.0113,
      "step": 688
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.39194103781295037,
      "learning_rate": 2.862436259190414e-07,
      "loss": 0.0073,
      "step": 689
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.4346352843060614,
      "learning_rate": 2.800719471941152e-07,
      "loss": 0.0059,
      "step": 690
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.3183987621587136,
      "learning_rate": 2.739647552895949e-07,
      "loss": 0.0069,
      "step": 691
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.4423423752806832,
      "learning_rate": 2.6792217251671744e-07,
      "loss": 0.0059,
      "step": 692
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.3864494243096827,
      "learning_rate": 2.619443198927677e-07,
      "loss": 0.006,
      "step": 693
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.322457016979029,
      "learning_rate": 2.5603131713865374e-07,
      "loss": 0.0069,
      "step": 694
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.4345066616936449,
      "learning_rate": 2.50183282676508e-07,
      "loss": 0.0091,
      "step": 695
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.34388210592322604,
      "learning_rate": 2.444003336273163e-07,
      "loss": 0.0032,
      "step": 696
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.2991066001834922,
      "learning_rate": 2.3868258580857164e-07,
      "loss": 0.0046,
      "step": 697
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.2816700072013969,
      "learning_rate": 2.3303015373195713e-07,
      "loss": 0.0054,
      "step": 698
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.2947482752142615,
      "learning_rate": 2.2744315060104846e-07,
      "loss": 0.0047,
      "step": 699
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.3450546396283738,
      "learning_rate": 2.2192168830904963e-07,
      "loss": 0.006,
      "step": 700
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.3822670216774389,
      "learning_rate": 2.1646587743655287e-07,
      "loss": 0.0062,
      "step": 701
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.36389056812509235,
      "learning_rate": 2.1107582724932088e-07,
      "loss": 0.0074,
      "step": 702
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.2241853060874336,
      "learning_rate": 2.0575164569610016e-07,
      "loss": 0.0031,
      "step": 703
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.3284239718422547,
      "learning_rate": 2.0049343940645935e-07,
      "loss": 0.006,
      "step": 704
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.27416640208522525,
      "learning_rate": 1.953013136886541e-07,
      "loss": 0.0043,
      "step": 705
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.48892924275252236,
      "learning_rate": 1.901753725275166e-07,
      "loss": 0.0054,
      "step": 706
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.3072478855542186,
      "learning_rate": 1.8511571858237357e-07,
      "loss": 0.0073,
      "step": 707
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.2971096027391294,
      "learning_rate": 1.801224531849908e-07,
      "loss": 0.0045,
      "step": 708
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.41409498453561594,
      "learning_rate": 1.7519567633754352e-07,
      "loss": 0.0072,
      "step": 709
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.4406136254485443,
      "learning_rate": 1.70335486710614e-07,
      "loss": 0.0103,
      "step": 710
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.42158029678243275,
      "learning_rate": 1.6554198164121265e-07,
      "loss": 0.0052,
      "step": 711
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.4431835034360088,
      "learning_rate": 1.6081525713083428e-07,
      "loss": 0.0065,
      "step": 712
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.3242131925310783,
      "learning_rate": 1.561554078435296e-07,
      "loss": 0.0049,
      "step": 713
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.30171217982310045,
      "learning_rate": 1.5156252710401207e-07,
      "loss": 0.0055,
      "step": 714
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.310486222626081,
      "learning_rate": 1.4703670689578884e-07,
      "loss": 0.0058,
      "step": 715
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.3339175632499386,
      "learning_rate": 1.4257803785931926e-07,
      "loss": 0.0068,
      "step": 716
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.3042650666913759,
      "learning_rate": 1.3818660929019717e-07,
      "loss": 0.0071,
      "step": 717
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.41181026400823106,
      "learning_rate": 1.3386250913736408e-07,
      "loss": 0.0075,
      "step": 718
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.33075765366170357,
      "learning_rate": 1.296058240013491e-07,
      "loss": 0.0075,
      "step": 719
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.313752325847203,
      "learning_rate": 1.2541663913253191e-07,
      "loss": 0.0059,
      "step": 720
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.3825929288696057,
      "learning_rate": 1.2129503842943645e-07,
      "loss": 0.0067,
      "step": 721
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2587277820133266,
      "learning_rate": 1.1724110443705115e-07,
      "loss": 0.0047,
      "step": 722
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.2844582633390916,
      "learning_rate": 1.1325491834517676e-07,
      "loss": 0.0059,
      "step": 723
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.3590012407056503,
      "learning_rate": 1.0933655998679653e-07,
      "loss": 0.0035,
      "step": 724
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.29637783616327534,
      "learning_rate": 1.0548610783648199e-07,
      "loss": 0.0052,
      "step": 725
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.468740471256091,
      "learning_rate": 1.0170363900881795e-07,
      "loss": 0.0089,
      "step": 726
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.3818989755993541,
      "learning_rate": 9.798922925685994e-08,
      "loss": 0.0066,
      "step": 727
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.19662730812801094,
      "learning_rate": 9.434295297061668e-08,
      "loss": 0.0029,
      "step": 728
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.4179664781848645,
      "learning_rate": 9.076488317555886e-08,
      "loss": 0.0041,
      "step": 729
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.3006062367530363,
      "learning_rate": 8.725509153115918e-08,
      "loss": 0.0053,
      "step": 730
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.3304263245245509,
      "learning_rate": 8.38136483294546e-08,
      "loss": 0.0063,
      "step": 731
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.33457641464311794,
      "learning_rate": 8.044062249364048e-08,
      "loss": 0.0042,
      "step": 732
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.5879062825513777,
      "learning_rate": 7.713608157668921e-08,
      "loss": 0.0215,
      "step": 733
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.3577419876216768,
      "learning_rate": 7.390009175999835e-08,
      "loss": 0.0082,
      "step": 734
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.3770406600422441,
      "learning_rate": 7.073271785206314e-08,
      "loss": 0.0079,
      "step": 735
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.33565852275052394,
      "learning_rate": 6.763402328718116e-08,
      "loss": 0.0062,
      "step": 736
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.3033892045381189,
      "learning_rate": 6.460407012417918e-08,
      "loss": 0.0053,
      "step": 737
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.30091494834673244,
      "learning_rate": 6.164291904517333e-08,
      "loss": 0.0054,
      "step": 738
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.3765855319872929,
      "learning_rate": 5.875062935435121e-08,
      "loss": 0.0054,
      "step": 739
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.32640485923404255,
      "learning_rate": 5.592725897678446e-08,
      "loss": 0.0078,
      "step": 740
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.43041557447303824,
      "learning_rate": 5.3172864457271926e-08,
      "loss": 0.0073,
      "step": 741
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.3145742514405517,
      "learning_rate": 5.048750095920151e-08,
      "loss": 0.0063,
      "step": 742
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.298005394632711,
      "learning_rate": 4.787122226345014e-08,
      "loss": 0.0065,
      "step": 743
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.3474563462151658,
      "learning_rate": 4.532408076730504e-08,
      "loss": 0.008,
      "step": 744
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.29691793848547526,
      "learning_rate": 4.2846127483414206e-08,
      "loss": 0.005,
      "step": 745
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.47996246825010264,
      "learning_rate": 4.043741203876483e-08,
      "loss": 0.0065,
      "step": 746
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.2723339869975387,
      "learning_rate": 3.80979826736893e-08,
      "loss": 0.0038,
      "step": 747
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.3193070254481293,
      "learning_rate": 3.58278862409e-08,
      "loss": 0.0064,
      "step": 748
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.3658645249659323,
      "learning_rate": 3.3627168204549306e-08,
      "loss": 0.0063,
      "step": 749
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.4539998378845374,
      "learning_rate": 3.1495872639320357e-08,
      "loss": 0.0146,
      "step": 750
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.33061360188059097,
      "learning_rate": 2.9434042229544543e-08,
      "loss": 0.0056,
      "step": 751
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.37571330796736685,
      "learning_rate": 2.7441718268344737e-08,
      "loss": 0.0042,
      "step": 752
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.24563965048779188,
      "learning_rate": 2.5518940656811095e-08,
      "loss": 0.0035,
      "step": 753
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.39737175448371265,
      "learning_rate": 2.3665747903199418e-08,
      "loss": 0.008,
      "step": 754
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.36816172196985625,
      "learning_rate": 2.1882177122162173e-08,
      "loss": 0.0069,
      "step": 755
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.32357455960281883,
      "learning_rate": 2.0168264034002404e-08,
      "loss": 0.0072,
      "step": 756
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.327715642412912,
      "learning_rate": 1.8524042963961095e-08,
      "loss": 0.0068,
      "step": 757
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.6097749311012022,
      "learning_rate": 1.6949546841528607e-08,
      "loss": 0.013,
      "step": 758
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.4321595864495073,
      "learning_rate": 1.544480719978447e-08,
      "loss": 0.0091,
      "step": 759
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.522892152816329,
      "learning_rate": 1.4009854174767521e-08,
      "loss": 0.0095,
      "step": 760
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.43910025769757355,
      "learning_rate": 1.2644716504870091e-08,
      "loss": 0.0092,
      "step": 761
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.3819688600545869,
      "learning_rate": 1.1349421530265246e-08,
      "loss": 0.0057,
      "step": 762
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.34216774949023093,
      "learning_rate": 1.0123995192356183e-08,
      "loss": 0.0061,
      "step": 763
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.32918386903466523,
      "learning_rate": 8.968462033259405e-09,
      "loss": 0.0056,
      "step": 764
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.32782865250712573,
      "learning_rate": 7.882845195312016e-09,
      "loss": 0.0069,
      "step": 765
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.32386570553933425,
      "learning_rate": 6.8671664206073625e-09,
      "loss": 0.0085,
      "step": 766
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.29040869309413037,
      "learning_rate": 5.921446050561386e-09,
      "loss": 0.0065,
      "step": 767
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.39088721124754344,
      "learning_rate": 5.0457030255038334e-09,
      "loss": 0.0104,
      "step": 768
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.35735404978476537,
      "learning_rate": 4.239954884299401e-09,
      "loss": 0.0053,
      "step": 769
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.3435391858174354,
      "learning_rate": 3.5042177639972304e-09,
      "loss": 0.0091,
      "step": 770
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.30805175946860186,
      "learning_rate": 2.838506399506446e-09,
      "loss": 0.006,
      "step": 771
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.31450891017452326,
      "learning_rate": 2.2428341233012294e-09,
      "loss": 0.0052,
      "step": 772
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.29550957672641986,
      "learning_rate": 1.7172128651554152e-09,
      "loss": 0.0053,
      "step": 773
    },
    {
      "epoch": 4.96,
      "grad_norm": 0.38035918617041065,
      "learning_rate": 1.2616531519011874e-09,
      "loss": 0.0082,
      "step": 774
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.3753382938936097,
      "learning_rate": 8.761641072196346e-10,
      "loss": 0.0117,
      "step": 775
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.34899017006560656,
      "learning_rate": 5.607534514585066e-10,
      "loss": 0.009,
      "step": 776
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.4164441780723029,
      "learning_rate": 3.1542750147639517e-10,
      "loss": 0.0117,
      "step": 777
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.34196331948745795,
      "learning_rate": 1.401911705168346e-10,
      "loss": 0.0072,
      "step": 778
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.30827744639368837,
      "learning_rate": 3.5047968109214176e-11,
      "loss": 0.0079,
      "step": 779
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3980736767828406,
      "learning_rate": 0.0,
      "loss": 0.0105,
      "step": 780
    },
    {
      "epoch": 5.0,
      "step": 780,
      "total_flos": 0.0,
      "train_loss": 0.17049677305622027,
      "train_runtime": 2176.8836,
      "train_samples_per_second": 11.484,
      "train_steps_per_second": 0.358
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 780,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}