{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 116,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 2.0672700448576804,
      "learning_rate": 0.0,
      "loss": 0.866,
      "step": 1
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 2.2902360202645,
      "learning_rate": 7.142857142857143e-07,
      "loss": 1.043,
      "step": 2
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 2.064587564315747,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.9479,
      "step": 3
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 2.1456251662755625,
      "learning_rate": 2.1428571428571427e-06,
      "loss": 1.0053,
      "step": 4
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 2.118290312395124,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 1.0228,
      "step": 5
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 1.9727110398188104,
      "learning_rate": 3.5714285714285718e-06,
      "loss": 0.9994,
      "step": 6
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 1.811780274017235,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 0.9832,
      "step": 7
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 1.60280335465616,
      "learning_rate": 5e-06,
      "loss": 1.0057,
      "step": 8
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 1.4946232465242921,
      "learning_rate": 4.999302593030069e-06,
      "loss": 1.0198,
      "step": 9
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 1.219848341043154,
      "learning_rate": 4.99721076122146e-06,
      "loss": 0.9945,
      "step": 10
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 1.0507879904789832,
      "learning_rate": 4.99372567166064e-06,
      "loss": 0.8896,
      "step": 11
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 1.2565080469257828,
      "learning_rate": 4.98884926876821e-06,
      "loss": 0.9066,
      "step": 12
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 1.3416635061508715,
      "learning_rate": 4.982584273214061e-06,
      "loss": 0.973,
      "step": 13
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 1.1635476799198454,
      "learning_rate": 4.974934180399447e-06,
      "loss": 0.8802,
      "step": 14
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 1.2273198837996009,
      "learning_rate": 4.965903258506806e-06,
      "loss": 0.9058,
      "step": 15
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 0.9397916351800223,
      "learning_rate": 4.955496546118439e-06,
      "loss": 0.851,
      "step": 16
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 0.8970172914695361,
      "learning_rate": 4.943719849405347e-06,
      "loss": 0.959,
      "step": 17
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.8557735235888474,
      "learning_rate": 4.930579738887827e-06,
      "loss": 0.8591,
      "step": 18
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.8228693160594547,
      "learning_rate": 4.9160835457696075e-06,
      "loss": 0.9227,
      "step": 19
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.9746558682735013,
      "learning_rate": 4.900239357847582e-06,
      "loss": 0.9185,
      "step": 20
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.783405238430667,
      "learning_rate": 4.883056014999423e-06,
      "loss": 0.8853,
      "step": 21
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 1.1560714233050355,
      "learning_rate": 4.864543104251587e-06,
      "loss": 0.837,
      "step": 22
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 0.6779891999963303,
      "learning_rate": 4.844710954430464e-06,
      "loss": 0.9006,
      "step": 23
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.6140368603132732,
      "learning_rate": 4.823570630399665e-06,
      "loss": 0.8713,
      "step": 24
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.5959623784138675,
      "learning_rate": 4.8011339268866505e-06,
      "loss": 0.8573,
      "step": 25
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.6645642684704514,
      "learning_rate": 4.777413361902152e-06,
      "loss": 0.9244,
      "step": 26
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 0.5749254662708289,
      "learning_rate": 4.752422169756048e-06,
      "loss": 0.8214,
      "step": 27
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.49471199829315,
      "learning_rate": 4.726174293673612e-06,
      "loss": 0.8501,
      "step": 28
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.49471199829315,
      "learning_rate": 4.698684378016223e-06,
      "loss": 0.7761,
      "step": 29
    },
    {
      "epoch": 1.0350877192982457,
      "grad_norm": 0.8408225702174914,
      "learning_rate": 4.669967760110908e-06,
      "loss": 0.8702,
      "step": 30
    },
    {
      "epoch": 1.0701754385964912,
      "grad_norm": 0.5019249070368396,
      "learning_rate": 4.6400404616932505e-06,
      "loss": 0.8376,
      "step": 31
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 0.4303555090359822,
      "learning_rate": 4.608919179968457e-06,
      "loss": 0.8834,
      "step": 32
    },
    {
      "epoch": 1.1403508771929824,
      "grad_norm": 0.446333451933402,
      "learning_rate": 4.576621278295558e-06,
      "loss": 0.8631,
      "step": 33
    },
    {
      "epoch": 1.1754385964912282,
      "grad_norm": 0.4783577116779664,
      "learning_rate": 4.543164776499945e-06,
      "loss": 0.8283,
      "step": 34
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 0.4079076293014929,
      "learning_rate": 4.508568340819654e-06,
      "loss": 0.8868,
      "step": 35
    },
    {
      "epoch": 1.2456140350877192,
      "grad_norm": 0.40536005006600634,
      "learning_rate": 4.472851273490985e-06,
      "loss": 0.8674,
      "step": 36
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 0.4152500781803562,
      "learning_rate": 4.436033501979299e-06,
      "loss": 0.859,
      "step": 37
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 0.4347010646426558,
      "learning_rate": 4.398135567860972e-06,
      "loss": 0.8254,
      "step": 38
    },
    {
      "epoch": 1.3508771929824561,
      "grad_norm": 0.3788287405374111,
      "learning_rate": 4.359178615362725e-06,
      "loss": 0.7829,
      "step": 39
    },
    {
      "epoch": 1.3859649122807016,
      "grad_norm": 0.4670693473167291,
      "learning_rate": 4.319184379564716e-06,
      "loss": 0.8918,
      "step": 40
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 0.47835984972537143,
      "learning_rate": 4.278175174273989e-06,
      "loss": 0.8179,
      "step": 41
    },
    {
      "epoch": 1.456140350877193,
      "grad_norm": 0.4499444593333873,
      "learning_rate": 4.236173879575022e-06,
      "loss": 0.7871,
      "step": 42
    },
    {
      "epoch": 1.4912280701754386,
      "grad_norm": 0.3878622142611574,
      "learning_rate": 4.1932039290643534e-06,
      "loss": 0.7789,
      "step": 43
    },
    {
      "epoch": 1.526315789473684,
      "grad_norm": 0.3316720314372251,
      "learning_rate": 4.149289296776369e-06,
      "loss": 0.91,
      "step": 44
    },
    {
      "epoch": 1.5614035087719298,
      "grad_norm": 0.36941265676632506,
      "learning_rate": 4.104454483807579e-06,
      "loss": 0.8607,
      "step": 45
    },
    {
      "epoch": 1.5964912280701755,
      "grad_norm": 0.3620626145776103,
      "learning_rate": 4.058724504646834e-06,
      "loss": 0.7937,
      "step": 46
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 0.3338291240702826,
      "learning_rate": 4.012124873219094e-06,
      "loss": 0.7815,
      "step": 47
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.472336322290524,
      "learning_rate": 3.964681588650562e-06,
      "loss": 0.8394,
      "step": 48
    },
    {
      "epoch": 1.7017543859649122,
      "grad_norm": 0.3420578655212163,
      "learning_rate": 3.916421120763106e-06,
      "loss": 0.8502,
      "step": 49
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 0.38000589787590366,
      "learning_rate": 3.8673703953060685e-06,
      "loss": 0.8347,
      "step": 50
    },
    {
      "epoch": 1.7719298245614035,
      "grad_norm": 0.3654714309302888,
      "learning_rate": 3.817556778933697e-06,
      "loss": 0.8581,
      "step": 51
    },
    {
      "epoch": 1.807017543859649,
      "grad_norm": 0.3918540158095545,
      "learning_rate": 3.7670080639366e-06,
      "loss": 0.8769,
      "step": 52
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 0.35070825923075866,
      "learning_rate": 3.7157524527357036e-06,
      "loss": 0.7163,
      "step": 53
    },
    {
      "epoch": 1.8771929824561404,
      "grad_norm": 0.35412184971484173,
      "learning_rate": 3.663818542147409e-06,
      "loss": 0.8756,
      "step": 54
    },
    {
      "epoch": 1.912280701754386,
      "grad_norm": 0.3580769918085387,
      "learning_rate": 3.61123530742869e-06,
      "loss": 0.8221,
      "step": 55
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 0.33901277879515357,
      "learning_rate": 3.5580320861110627e-06,
      "loss": 0.8703,
      "step": 56
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 0.3463865269661244,
      "learning_rate": 3.5042385616324243e-06,
      "loss": 0.8721,
      "step": 57
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5122368904802833,
      "learning_rate": 3.4498847467759e-06,
      "loss": 0.8607,
      "step": 58
    },
    {
      "epoch": 2.0350877192982457,
      "grad_norm": 0.32344238659596863,
      "learning_rate": 3.3950009669249502e-06,
      "loss": 0.7811,
      "step": 59
    },
    {
      "epoch": 2.0701754385964914,
      "grad_norm": 0.35108823806290596,
      "learning_rate": 3.3396178431440572e-06,
      "loss": 0.9217,
      "step": 60
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 0.30984522709441403,
      "learning_rate": 3.283766275094454e-06,
      "loss": 0.7985,
      "step": 61
    },
    {
      "epoch": 2.1403508771929824,
      "grad_norm": 0.34747230260102224,
      "learning_rate": 3.227477423794412e-06,
      "loss": 0.7911,
      "step": 62
    },
    {
      "epoch": 2.175438596491228,
      "grad_norm": 0.3139893028606255,
      "learning_rate": 3.1707826942337124e-06,
      "loss": 0.8368,
      "step": 63
    },
    {
      "epoch": 2.2105263157894735,
      "grad_norm": 0.32074434133002094,
      "learning_rate": 3.1137137178519983e-06,
      "loss": 0.8602,
      "step": 64
    },
    {
      "epoch": 2.245614035087719,
      "grad_norm": 0.33796734180843035,
      "learning_rate": 3.056302334890786e-06,
      "loss": 0.7477,
      "step": 65
    },
    {
      "epoch": 2.280701754385965,
      "grad_norm": 0.308432104905017,
      "learning_rate": 2.9985805766289815e-06,
      "loss": 0.808,
      "step": 66
    },
    {
      "epoch": 2.3157894736842106,
      "grad_norm": 0.3123660916517458,
      "learning_rate": 2.940580647511805e-06,
      "loss": 0.7912,
      "step": 67
    },
    {
      "epoch": 2.3508771929824563,
      "grad_norm": 0.3406699289606851,
      "learning_rate": 2.8823349071831154e-06,
      "loss": 0.857,
      "step": 68
    },
    {
      "epoch": 2.3859649122807016,
      "grad_norm": 0.3063945394190306,
      "learning_rate": 2.8238758524311316e-06,
      "loss": 0.7817,
      "step": 69
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 0.31851738409569996,
      "learning_rate": 2.7652360990576457e-06,
      "loss": 0.9036,
      "step": 70
    },
    {
      "epoch": 2.456140350877193,
      "grad_norm": 0.3145120103359264,
      "learning_rate": 2.7064483636808314e-06,
      "loss": 0.8149,
      "step": 71
    },
    {
      "epoch": 2.4912280701754383,
      "grad_norm": 0.3050076260239972,
      "learning_rate": 2.6475454454818072e-06,
      "loss": 0.8372,
      "step": 72
    },
    {
      "epoch": 2.526315789473684,
      "grad_norm": 0.30681790274996473,
      "learning_rate": 2.5885602079051354e-06,
      "loss": 0.7604,
      "step": 73
    },
    {
      "epoch": 2.56140350877193,
      "grad_norm": 0.314110412232156,
      "learning_rate": 2.529525560323462e-06,
      "loss": 0.8005,
      "step": 74
    },
    {
      "epoch": 2.5964912280701755,
      "grad_norm": 0.3236785830234266,
      "learning_rate": 2.470474439676539e-06,
      "loss": 0.8168,
      "step": 75
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.359656431397916,
      "learning_rate": 2.411439792094866e-06,
      "loss": 0.8402,
      "step": 76
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.32366656133415905,
      "learning_rate": 2.3524545545181936e-06,
      "loss": 0.8668,
      "step": 77
    },
    {
      "epoch": 2.7017543859649122,
      "grad_norm": 0.3262664895219039,
      "learning_rate": 2.2935516363191695e-06,
      "loss": 0.8343,
      "step": 78
    },
    {
      "epoch": 2.736842105263158,
      "grad_norm": 0.3081673146324169,
      "learning_rate": 2.234763900942355e-06,
      "loss": 0.8084,
      "step": 79
    },
    {
      "epoch": 2.7719298245614032,
      "grad_norm": 0.32923845454809364,
      "learning_rate": 2.1761241475688697e-06,
      "loss": 0.7564,
      "step": 80
    },
    {
      "epoch": 2.807017543859649,
      "grad_norm": 0.3103638917941097,
      "learning_rate": 2.117665092816885e-06,
      "loss": 0.7423,
      "step": 81
    },
    {
      "epoch": 2.8421052631578947,
      "grad_norm": 0.32527138845292275,
      "learning_rate": 2.059419352488196e-06,
      "loss": 0.8292,
      "step": 82
    },
    {
      "epoch": 2.8771929824561404,
      "grad_norm": 0.31114889374241994,
      "learning_rate": 2.0014194233710193e-06,
      "loss": 0.8229,
      "step": 83
    },
    {
      "epoch": 2.912280701754386,
      "grad_norm": 0.3119716096545865,
      "learning_rate": 1.9436976651092143e-06,
      "loss": 0.829,
      "step": 84
    },
    {
      "epoch": 2.9473684210526314,
      "grad_norm": 0.31910821914661397,
      "learning_rate": 1.8862862821480023e-06,
      "loss": 0.7703,
      "step": 85
    },
    {
      "epoch": 2.982456140350877,
      "grad_norm": 0.31035281270338383,
      "learning_rate": 1.829217305766289e-06,
      "loss": 0.8182,
      "step": 86
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.31035281270338383,
      "learning_rate": 1.772522576205589e-06,
      "loss": 0.6834,
      "step": 87
    },
    {
      "epoch": 3.0350877192982457,
      "grad_norm": 0.6381501820358972,
      "learning_rate": 1.7162337249055478e-06,
      "loss": 0.7985,
      "step": 88
    },
    {
      "epoch": 3.0701754385964914,
      "grad_norm": 0.34108098057965985,
      "learning_rate": 1.6603821568559436e-06,
      "loss": 0.8813,
      "step": 89
    },
    {
      "epoch": 3.1052631578947367,
      "grad_norm": 0.3119218127003693,
      "learning_rate": 1.6049990330750508e-06,
      "loss": 0.8332,
      "step": 90
    },
    {
      "epoch": 3.1403508771929824,
      "grad_norm": 0.31753762794893836,
      "learning_rate": 1.5501152532241003e-06,
      "loss": 0.8085,
      "step": 91
    },
    {
      "epoch": 3.175438596491228,
      "grad_norm": 0.26210667369890633,
      "learning_rate": 1.495761438367577e-06,
      "loss": 0.7937,
      "step": 92
    },
    {
      "epoch": 3.2105263157894735,
      "grad_norm": 0.3412780118816683,
      "learning_rate": 1.4419679138889379e-06,
      "loss": 0.9124,
      "step": 93
    },
    {
      "epoch": 3.245614035087719,
      "grad_norm": 0.3273535475843529,
      "learning_rate": 1.3887646925713116e-06,
      "loss": 0.771,
      "step": 94
    },
    {
      "epoch": 3.280701754385965,
      "grad_norm": 0.301251838670515,
      "learning_rate": 1.3361814578525922e-06,
      "loss": 0.7657,
      "step": 95
    },
    {
      "epoch": 3.3157894736842106,
      "grad_norm": 0.362998311924545,
      "learning_rate": 1.2842475472642969e-06,
      "loss": 0.8489,
      "step": 96
    },
    {
      "epoch": 3.3508771929824563,
      "grad_norm": 0.29122608306651043,
      "learning_rate": 1.2329919360634003e-06,
      "loss": 0.8135,
      "step": 97
    },
    {
      "epoch": 3.3859649122807016,
      "grad_norm": 0.29475940071935713,
      "learning_rate": 1.182443221066303e-06,
      "loss": 0.7059,
      "step": 98
    },
    {
      "epoch": 3.4210526315789473,
      "grad_norm": 0.28973717877011734,
      "learning_rate": 1.1326296046939334e-06,
      "loss": 0.7342,
      "step": 99
    },
    {
      "epoch": 3.456140350877193,
      "grad_norm": 0.30995625777137437,
      "learning_rate": 1.083578879236895e-06,
      "loss": 0.7411,
      "step": 100
    },
    {
      "epoch": 3.4912280701754383,
      "grad_norm": 0.2859432013165172,
      "learning_rate": 1.0353184113494386e-06,
      "loss": 0.772,
      "step": 101
    },
    {
      "epoch": 3.526315789473684,
      "grad_norm": 0.30700142822195303,
      "learning_rate": 9.878751267809069e-07,
      "loss": 0.7557,
      "step": 102
    },
    {
      "epoch": 3.56140350877193,
      "grad_norm": 0.27176920609301375,
      "learning_rate": 9.412754953531664e-07,
      "loss": 0.7729,
      "step": 103
    },
    {
      "epoch": 3.5964912280701755,
      "grad_norm": 0.36537897752504017,
      "learning_rate": 8.955455161924217e-07,
      "loss": 0.7803,
      "step": 104
    },
    {
      "epoch": 3.6315789473684212,
      "grad_norm": 0.32335411827201516,
      "learning_rate": 8.507107032236323e-07,
      "loss": 0.6832,
      "step": 105
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.31630275273581865,
      "learning_rate": 8.067960709356479e-07,
      "loss": 0.7744,
      "step": 106
    },
    {
      "epoch": 3.7017543859649122,
      "grad_norm": 0.3116184912362651,
      "learning_rate": 7.638261204249783e-07,
      "loss": 0.7719,
      "step": 107
    },
    {
      "epoch": 3.736842105263158,
      "grad_norm": 0.31857809248659025,
      "learning_rate": 7.218248257260127e-07,
      "loss": 0.8335,
      "step": 108
    },
    {
      "epoch": 3.7719298245614032,
      "grad_norm": 0.28756539469934045,
      "learning_rate": 6.808156204352845e-07,
      "loss": 0.857,
      "step": 109
    },
    {
      "epoch": 3.807017543859649,
      "grad_norm": 0.30352272859730434,
      "learning_rate": 6.40821384637276e-07,
      "loss": 0.8382,
      "step": 110
    },
    {
      "epoch": 3.8421052631578947,
      "grad_norm": 0.31531467423763393,
      "learning_rate": 6.018644321390288e-07,
      "loss": 0.823,
      "step": 111
    },
    {
      "epoch": 3.8771929824561404,
      "grad_norm": 0.2935691411718206,
      "learning_rate": 5.639664980207024e-07,
      "loss": 0.7658,
      "step": 112
    },
    {
      "epoch": 3.912280701754386,
      "grad_norm": 0.3148692436878785,
      "learning_rate": 5.271487265090163e-07,
      "loss": 0.7759,
      "step": 113
    },
    {
      "epoch": 3.9473684210526314,
      "grad_norm": 0.28756831417145856,
      "learning_rate": 4.914316591803475e-07,
      "loss": 0.7931,
      "step": 114
    },
    {
      "epoch": 3.982456140350877,
      "grad_norm": 0.2919237741849252,
      "learning_rate": 4.5683522350005505e-07,
      "loss": 0.81,
      "step": 115
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.4951765857412171,
      "learning_rate": 4.23378721704443e-07,
      "loss": 0.7612,
      "step": 116
    }
  ],
  "logging_steps": 1,
  "max_steps": 140,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 52227047817216.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}