{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 322,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03117692907248636,
      "grad_norm": 89.4246826171875,
      "learning_rate": 6.060606060606061e-07,
      "loss": 2.7034,
      "step": 5
    },
    {
      "epoch": 0.06235385814497272,
      "grad_norm": 29.731754302978516,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 2.081,
      "step": 10
    },
    {
      "epoch": 0.09353078721745908,
      "grad_norm": 15.751827239990234,
      "learning_rate": 2.1212121212121216e-06,
      "loss": 1.2438,
      "step": 15
    },
    {
      "epoch": 0.12470771628994544,
      "grad_norm": 9.355167388916016,
      "learning_rate": 2.8787878787878793e-06,
      "loss": 0.7001,
      "step": 20
    },
    {
      "epoch": 0.1558846453624318,
      "grad_norm": 7.231124401092529,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.3416,
      "step": 25
    },
    {
      "epoch": 0.18706157443491817,
      "grad_norm": 4.852797985076904,
      "learning_rate": 4.393939393939394e-06,
      "loss": 0.2417,
      "step": 30
    },
    {
      "epoch": 0.2182385035074045,
      "grad_norm": 3.667936325073242,
      "learning_rate": 4.999852290034674e-06,
      "loss": 0.1518,
      "step": 35
    },
    {
      "epoch": 0.24941543257989088,
      "grad_norm": 2.801180601119995,
      "learning_rate": 4.994684273748881e-06,
      "loss": 0.1622,
      "step": 40
    },
    {
      "epoch": 0.2805923616523772,
      "grad_norm": 3.259979724884033,
      "learning_rate": 4.982148204514838e-06,
      "loss": 0.1173,
      "step": 45
    },
    {
      "epoch": 0.3117692907248636,
      "grad_norm": 3.0364513397216797,
      "learning_rate": 4.962281107627828e-06,
      "loss": 0.1068,
      "step": 50
    },
    {
      "epoch": 0.34294621979734996,
      "grad_norm": 2.9294209480285645,
      "learning_rate": 4.935141660581967e-06,
      "loss": 0.1101,
      "step": 55
    },
    {
      "epoch": 0.37412314886983633,
      "grad_norm": 2.0414984226226807,
      "learning_rate": 4.900810019766151e-06,
      "loss": 0.0871,
      "step": 60
    },
    {
      "epoch": 0.4053000779423227,
      "grad_norm": 1.3614494800567627,
      "learning_rate": 4.859387583722069e-06,
      "loss": 0.0715,
      "step": 65
    },
    {
      "epoch": 0.436477007014809,
      "grad_norm": 1.8288466930389404,
      "learning_rate": 4.8109966936634885e-06,
      "loss": 0.0934,
      "step": 70
    },
    {
      "epoch": 0.4676539360872954,
      "grad_norm": 3.0065410137176514,
      "learning_rate": 4.755780272141338e-06,
      "loss": 0.1147,
      "step": 75
    },
    {
      "epoch": 0.49883086515978176,
      "grad_norm": 1.4399396181106567,
      "learning_rate": 4.693901400921782e-06,
      "loss": 0.0863,
      "step": 80
    },
    {
      "epoch": 0.5300077942322681,
      "grad_norm": 1.658852458000183,
      "learning_rate": 4.625542839324036e-06,
      "loss": 0.0814,
      "step": 85
    },
    {
      "epoch": 0.5611847233047544,
      "grad_norm": 1.3098065853118896,
      "learning_rate": 4.550906484440495e-06,
      "loss": 0.087,
      "step": 90
    },
    {
      "epoch": 0.5923616523772408,
      "grad_norm": 2.129857301712036,
      "learning_rate": 4.4702127748334406e-06,
      "loss": 0.0655,
      "step": 95
    },
    {
      "epoch": 0.6235385814497272,
      "grad_norm": 1.4986920356750488,
      "learning_rate": 4.38370003946948e-06,
      "loss": 0.0769,
      "step": 100
    },
    {
      "epoch": 0.6547155105222136,
      "grad_norm": 1.4480761289596558,
      "learning_rate": 4.2916237938146635e-06,
      "loss": 0.0914,
      "step": 105
    },
    {
      "epoch": 0.6858924395946999,
      "grad_norm": 1.4360016584396362,
      "learning_rate": 4.194255985169259e-06,
      "loss": 0.0808,
      "step": 110
    },
    {
      "epoch": 0.7170693686671863,
      "grad_norm": 1.4440962076187134,
      "learning_rate": 4.091884189471071e-06,
      "loss": 0.0699,
      "step": 115
    },
    {
      "epoch": 0.7482462977396727,
      "grad_norm": 1.1551469564437866,
      "learning_rate": 3.984810761939578e-06,
      "loss": 0.0702,
      "step": 120
    },
    {
      "epoch": 0.779423226812159,
      "grad_norm": 1.4710094928741455,
      "learning_rate": 3.873351944069425e-06,
      "loss": 0.071,
      "step": 125
    },
    {
      "epoch": 0.8106001558846454,
      "grad_norm": 1.350879192352295,
      "learning_rate": 3.757836929610803e-06,
      "loss": 0.0808,
      "step": 130
    },
    {
      "epoch": 0.8417770849571318,
      "grad_norm": 3.6121158599853516,
      "learning_rate": 3.638606892295315e-06,
      "loss": 0.0822,
      "step": 135
    },
    {
      "epoch": 0.872954014029618,
      "grad_norm": 1.3339393138885498,
      "learning_rate": 3.5160139781789615e-06,
      "loss": 0.086,
      "step": 140
    },
    {
      "epoch": 0.9041309431021044,
      "grad_norm": 1.96644926071167,
      "learning_rate": 3.3904202655783426e-06,
      "loss": 0.0712,
      "step": 145
    },
    {
      "epoch": 0.9353078721745908,
      "grad_norm": 2.5620317459106445,
      "learning_rate": 3.2621966956719265e-06,
      "loss": 0.0693,
      "step": 150
    },
    {
      "epoch": 0.9664848012470771,
      "grad_norm": 2.474618673324585,
      "learning_rate": 3.1317219769248373e-06,
      "loss": 0.0902,
      "step": 155
    },
    {
      "epoch": 0.9976617303195635,
      "grad_norm": 1.1106942892074585,
      "learning_rate": 2.9993814665729605e-06,
      "loss": 0.064,
      "step": 160
    },
    {
      "epoch": 1.0249415432579891,
      "grad_norm": 0.7524325847625732,
      "learning_rate": 2.8655660324698887e-06,
      "loss": 0.0573,
      "step": 165
    },
    {
      "epoch": 1.0561184723304755,
      "grad_norm": 0.933891773223877,
      "learning_rate": 2.730670898658255e-06,
      "loss": 0.055,
      "step": 170
    },
    {
      "epoch": 1.0872954014029619,
      "grad_norm": 0.8289980292320251,
      "learning_rate": 2.5950944780750577e-06,
      "loss": 0.0564,
      "step": 175
    },
    {
      "epoch": 1.1184723304754483,
      "grad_norm": 0.7259963750839233,
      "learning_rate": 2.459237195838577e-06,
      "loss": 0.0484,
      "step": 180
    },
    {
      "epoch": 1.1496492595479346,
      "grad_norm": 3.1367318630218506,
      "learning_rate": 2.3235003065923203e-06,
      "loss": 0.0566,
      "step": 185
    },
    {
      "epoch": 1.1808261886204208,
      "grad_norm": 1.2359286546707153,
      "learning_rate": 2.1882847093989544e-06,
      "loss": 0.0633,
      "step": 190
    },
    {
      "epoch": 1.2120031176929071,
      "grad_norm": 0.8619211912155151,
      "learning_rate": 2.0539897636844543e-06,
      "loss": 0.0522,
      "step": 195
    },
    {
      "epoch": 1.2431800467653935,
      "grad_norm": 0.9569171667098999,
      "learning_rate": 1.921012109729562e-06,
      "loss": 0.052,
      "step": 200
    },
    {
      "epoch": 1.2743569758378799,
      "grad_norm": 1.449960708618164,
      "learning_rate": 1.7897444971922559e-06,
      "loss": 0.0579,
      "step": 205
    },
    {
      "epoch": 1.3055339049103662,
      "grad_norm": 0.5719594955444336,
      "learning_rate": 1.6605746251211642e-06,
      "loss": 0.0533,
      "step": 210
    },
    {
      "epoch": 1.3367108339828526,
      "grad_norm": 0.9786936044692993,
      "learning_rate": 1.5338839968859749e-06,
      "loss": 0.0595,
      "step": 215
    },
    {
      "epoch": 1.367887763055339,
      "grad_norm": 1.259804129600525,
      "learning_rate": 1.4100467934067775e-06,
      "loss": 0.059,
      "step": 220
    },
    {
      "epoch": 1.3990646921278254,
      "grad_norm": 0.9070944786071777,
      "learning_rate": 1.2894287680102735e-06,
      "loss": 0.0544,
      "step": 225
    },
    {
      "epoch": 1.4302416212003117,
      "grad_norm": 1.249437689781189,
      "learning_rate": 1.1723861661769e-06,
      "loss": 0.0555,
      "step": 230
    },
    {
      "epoch": 1.461418550272798,
      "grad_norm": 1.1977324485778809,
      "learning_rate": 1.0592646733693748e-06,
      "loss": 0.0597,
      "step": 235
    },
    {
      "epoch": 1.4925954793452845,
      "grad_norm": 1.215759515762329,
      "learning_rate": 9.503983940502795e-07,
      "loss": 0.0511,
      "step": 240
    },
    {
      "epoch": 1.5237724084177708,
      "grad_norm": 0.6760881543159485,
      "learning_rate": 8.461088649041341e-07,
      "loss": 0.0535,
      "step": 245
    },
    {
      "epoch": 1.5549493374902572,
      "grad_norm": 0.9166566729545593,
      "learning_rate": 7.467041051784155e-07,
      "loss": 0.053,
      "step": 250
    },
    {
      "epoch": 1.5861262665627436,
      "grad_norm": 0.8134734630584717,
      "learning_rate": 6.524777069483526e-07,
      "loss": 0.0543,
      "step": 255
    },
    {
      "epoch": 1.61730319563523,
      "grad_norm": 1.4211320877075195,
      "learning_rate": 5.637079679923796e-07,
      "loss": 0.0525,
      "step": 260
    },
    {
      "epoch": 1.6484801247077163,
      "grad_norm": 0.8970643877983093,
      "learning_rate": 4.806570698393232e-07,
      "loss": 0.0488,
      "step": 265
    },
    {
      "epoch": 1.6796570537802027,
      "grad_norm": 0.736966073513031,
      "learning_rate": 4.035703034149463e-07,
      "loss": 0.0511,
      "step": 270
    },
    {
      "epoch": 1.710833982852689,
      "grad_norm": 0.878939151763916,
      "learning_rate": 3.326753445749059e-07,
      "loss": 0.0517,
      "step": 275
    },
    {
      "epoch": 1.7420109119251754,
      "grad_norm": 1.1599282026290894,
      "learning_rate": 2.681815816638503e-07,
      "loss": 0.049,
      "step": 280
    },
    {
      "epoch": 1.7731878409976618,
      "grad_norm": 0.8185058236122131,
      "learning_rate": 2.102794970866917e-07,
      "loss": 0.0508,
      "step": 285
    },
    {
      "epoch": 1.8043647700701482,
      "grad_norm": 0.8199148178100586,
      "learning_rate": 1.5914010471859947e-07,
      "loss": 0.0513,
      "step": 290
    },
    {
      "epoch": 1.8355416991426345,
      "grad_norm": 0.7595412135124207,
      "learning_rate": 1.1491444481531888e-07,
      "loss": 0.0493,
      "step": 295
    },
    {
      "epoch": 1.866718628215121,
      "grad_norm": 0.5946995615959167,
      "learning_rate": 7.773313791559545e-08,
      "loss": 0.0496,
      "step": 300
    },
    {
      "epoch": 1.8978955572876073,
      "grad_norm": 0.7376562356948853,
      "learning_rate": 4.7705999053261895e-08,
      "loss": 0.0516,
      "step": 305
    },
    {
      "epoch": 1.9290724863600937,
      "grad_norm": 0.639217734336853,
      "learning_rate": 2.492171341840405e-08,
      "loss": 0.0459,
      "step": 310
    },
    {
      "epoch": 1.96024941543258,
      "grad_norm": 0.5890375375747681,
      "learning_rate": 9.447574425548444e-09,
      "loss": 0.0551,
      "step": 315
    },
    {
      "epoch": 1.9914263445050664,
      "grad_norm": 0.714377224445343,
      "learning_rate": 1.3292849624821357e-09,
      "loss": 0.0493,
      "step": 320
    }
  ],
  "logging_steps": 5,
  "max_steps": 322,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.581228895897846e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}