{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5538248528902735,
  "eval_steps": 500,
  "global_step": 900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006153609476558594,
      "grad_norm": 26290.87109375,
      "learning_rate": 1.9999999999999996e-07,
      "loss": 8.2122,
      "step": 10
    },
    {
      "epoch": 0.012307218953117188,
      "grad_norm": 1750.26953125,
      "learning_rate": 1.2e-06,
      "loss": 7.6386,
      "step": 20
    },
    {
      "epoch": 0.018460828429675783,
      "grad_norm": 235.5290985107422,
      "learning_rate": 2.1999999999999997e-06,
      "loss": 7.0511,
      "step": 30
    },
    {
      "epoch": 0.024614437906234375,
      "grad_norm": 317.7790222167969,
      "learning_rate": 3.1999999999999994e-06,
      "loss": 6.889,
      "step": 40
    },
    {
      "epoch": 0.03076804738279297,
      "grad_norm": 906.6453247070312,
      "learning_rate": 4.2e-06,
      "loss": 6.8145,
      "step": 50
    },
    {
      "epoch": 0.036921656859351566,
      "grad_norm": 1056.044189453125,
      "learning_rate": 5.199999999999999e-06,
      "loss": 6.7193,
      "step": 60
    },
    {
      "epoch": 0.043075266335910155,
      "grad_norm": 830.0984497070312,
      "learning_rate": 6.199999999999999e-06,
      "loss": 6.6406,
      "step": 70
    },
    {
      "epoch": 0.04922887581246875,
      "grad_norm": 259.50299072265625,
      "learning_rate": 7.2e-06,
      "loss": 6.6149,
      "step": 80
    },
    {
      "epoch": 0.055382485289027346,
      "grad_norm": 143.87599182128906,
      "learning_rate": 8.2e-06,
      "loss": 6.6122,
      "step": 90
    },
    {
      "epoch": 0.06153609476558594,
      "grad_norm": 435.93017578125,
      "learning_rate": 9.199999999999998e-06,
      "loss": 6.5612,
      "step": 100
    },
    {
      "epoch": 0.06768970424214453,
      "grad_norm": 557.260498046875,
      "learning_rate": 1.02e-05,
      "loss": 6.5083,
      "step": 110
    },
    {
      "epoch": 0.07384331371870313,
      "grad_norm": 291.8673095703125,
      "learning_rate": 1.12e-05,
      "loss": 6.5365,
      "step": 120
    },
    {
      "epoch": 0.07999692319526172,
      "grad_norm": 434.8012390136719,
      "learning_rate": 1.2199999999999998e-05,
      "loss": 6.5229,
      "step": 130
    },
    {
      "epoch": 0.08615053267182031,
      "grad_norm": 1767.30908203125,
      "learning_rate": 1.3199999999999997e-05,
      "loss": 6.4961,
      "step": 140
    },
    {
      "epoch": 0.09230414214837891,
      "grad_norm": 548.34228515625,
      "learning_rate": 1.4199999999999998e-05,
      "loss": 6.5135,
      "step": 150
    },
    {
      "epoch": 0.0984577516249375,
      "grad_norm": 212.0412139892578,
      "learning_rate": 1.5199999999999998e-05,
      "loss": 6.47,
      "step": 160
    },
    {
      "epoch": 0.1046113611014961,
      "grad_norm": 220.91732788085938,
      "learning_rate": 1.6199999999999997e-05,
      "loss": 6.4492,
      "step": 170
    },
    {
      "epoch": 0.11076497057805469,
      "grad_norm": 275.8125305175781,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 6.5087,
      "step": 180
    },
    {
      "epoch": 0.11691858005461328,
      "grad_norm": 205.40353393554688,
      "learning_rate": 1.82e-05,
      "loss": 6.4695,
      "step": 190
    },
    {
      "epoch": 0.12307218953117188,
      "grad_norm": 170.36024475097656,
      "learning_rate": 1.92e-05,
      "loss": 6.4184,
      "step": 200
    },
    {
      "epoch": 0.12922579900773049,
      "grad_norm": 240.3199462890625,
      "learning_rate": 2.0199999999999996e-05,
      "loss": 6.4329,
      "step": 210
    },
    {
      "epoch": 0.13537940848428906,
      "grad_norm": 508.4911193847656,
      "learning_rate": 2.12e-05,
      "loss": 6.3826,
      "step": 220
    },
    {
      "epoch": 0.14153301796084766,
      "grad_norm": 117.4170150756836,
      "learning_rate": 2.2199999999999998e-05,
      "loss": 6.394,
      "step": 230
    },
    {
      "epoch": 0.14768662743740626,
      "grad_norm": 76.04037475585938,
      "learning_rate": 2.3199999999999998e-05,
      "loss": 6.3696,
      "step": 240
    },
    {
      "epoch": 0.15384023691396484,
      "grad_norm": 243.4922332763672,
      "learning_rate": 2.42e-05,
      "loss": 6.3843,
      "step": 250
    },
    {
      "epoch": 0.15999384639052344,
      "grad_norm": 84.42725372314453,
      "learning_rate": 2.52e-05,
      "loss": 6.3401,
      "step": 260
    },
    {
      "epoch": 0.16614745586708204,
      "grad_norm": 119.90478515625,
      "learning_rate": 2.6199999999999996e-05,
      "loss": 6.3586,
      "step": 270
    },
    {
      "epoch": 0.17230106534364062,
      "grad_norm": 101.67578125,
      "learning_rate": 2.72e-05,
      "loss": 6.3315,
      "step": 280
    },
    {
      "epoch": 0.17845467482019922,
      "grad_norm": 64.19589233398438,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 6.3033,
      "step": 290
    },
    {
      "epoch": 0.18460828429675782,
      "grad_norm": 43.22764205932617,
      "learning_rate": 2.9199999999999995e-05,
      "loss": 6.3065,
      "step": 300
    },
    {
      "epoch": 0.1907618937733164,
      "grad_norm": 70.01463317871094,
      "learning_rate": 3.02e-05,
      "loss": 6.3044,
      "step": 310
    },
    {
      "epoch": 0.196915503249875,
      "grad_norm": 48.240699768066406,
      "learning_rate": 3.119999999999999e-05,
      "loss": 6.2598,
      "step": 320
    },
    {
      "epoch": 0.2030691127264336,
      "grad_norm": 59.44075012207031,
      "learning_rate": 3.22e-05,
      "loss": 6.2607,
      "step": 330
    },
    {
      "epoch": 0.2092227222029922,
      "grad_norm": 61.2050666809082,
      "learning_rate": 3.3199999999999994e-05,
      "loss": 6.2497,
      "step": 340
    },
    {
      "epoch": 0.21537633167955078,
      "grad_norm": 60.02091598510742,
      "learning_rate": 3.42e-05,
      "loss": 6.2319,
      "step": 350
    },
    {
      "epoch": 0.22152994115610938,
      "grad_norm": 41.094627380371094,
      "learning_rate": 3.5199999999999995e-05,
      "loss": 6.219,
      "step": 360
    },
    {
      "epoch": 0.22768355063266799,
      "grad_norm": 64.33719635009766,
      "learning_rate": 3.62e-05,
      "loss": 6.2147,
      "step": 370
    },
    {
      "epoch": 0.23383716010922656,
      "grad_norm": 62.03013610839844,
      "learning_rate": 3.7199999999999996e-05,
      "loss": 6.1763,
      "step": 380
    },
    {
      "epoch": 0.23999076958578516,
      "grad_norm": 181.71493530273438,
      "learning_rate": 3.8199999999999993e-05,
      "loss": 6.1753,
      "step": 390
    },
    {
      "epoch": 0.24614437906234377,
      "grad_norm": 80.83346557617188,
      "learning_rate": 3.919999999999999e-05,
      "loss": 6.1614,
      "step": 400
    },
    {
      "epoch": 0.25229798853890234,
      "grad_norm": 35.45427703857422,
      "learning_rate": 4.02e-05,
      "loss": 6.1271,
      "step": 410
    },
    {
      "epoch": 0.25845159801546097,
      "grad_norm": 44.923248291015625,
      "learning_rate": 4.12e-05,
      "loss": 6.1036,
      "step": 420
    },
    {
      "epoch": 0.26460520749201955,
      "grad_norm": 53.389827728271484,
      "learning_rate": 4.2199999999999996e-05,
      "loss": 6.0938,
      "step": 430
    },
    {
      "epoch": 0.2707588169685781,
      "grad_norm": 36.64761734008789,
      "learning_rate": 4.319999999999999e-05,
      "loss": 6.1188,
      "step": 440
    },
    {
      "epoch": 0.27691242644513675,
      "grad_norm": 73.00870513916016,
      "learning_rate": 4.42e-05,
      "loss": 6.0647,
      "step": 450
    },
    {
      "epoch": 0.2830660359216953,
      "grad_norm": 54.08327865600586,
      "learning_rate": 4.5199999999999994e-05,
      "loss": 6.0432,
      "step": 460
    },
    {
      "epoch": 0.2892196453982539,
      "grad_norm": 40.13100051879883,
      "learning_rate": 4.62e-05,
      "loss": 6.0383,
      "step": 470
    },
    {
      "epoch": 0.29537325487481253,
      "grad_norm": 39.69797134399414,
      "learning_rate": 4.7199999999999995e-05,
      "loss": 6.0178,
      "step": 480
    },
    {
      "epoch": 0.3015268643513711,
      "grad_norm": 47.64400863647461,
      "learning_rate": 4.82e-05,
      "loss": 6.0412,
      "step": 490
    },
    {
      "epoch": 0.3076804738279297,
      "grad_norm": 59.538387298583984,
      "learning_rate": 4.9199999999999997e-05,
      "loss": 5.9841,
      "step": 500
    },
    {
      "epoch": 0.3138340833044883,
      "grad_norm": 63.603389739990234,
      "learning_rate": 5.0199999999999994e-05,
      "loss": 6.0088,
      "step": 510
    },
    {
      "epoch": 0.3199876927810469,
      "grad_norm": 50.28224563598633,
      "learning_rate": 5.119999999999999e-05,
      "loss": 5.9469,
      "step": 520
    },
    {
      "epoch": 0.32614130225760546,
      "grad_norm": 43.31625747680664,
      "learning_rate": 5.2199999999999995e-05,
      "loss": 5.9464,
      "step": 530
    },
    {
      "epoch": 0.3322949117341641,
      "grad_norm": 56.94966506958008,
      "learning_rate": 5.32e-05,
      "loss": 5.9139,
      "step": 540
    },
    {
      "epoch": 0.33844852121072266,
      "grad_norm": 97.49286651611328,
      "learning_rate": 5.4199999999999996e-05,
      "loss": 5.9177,
      "step": 550
    },
    {
      "epoch": 0.34460213068728124,
      "grad_norm": 57.25851058959961,
      "learning_rate": 5.519999999999999e-05,
      "loss": 5.8818,
      "step": 560
    },
    {
      "epoch": 0.35075574016383987,
      "grad_norm": 40.23714065551758,
      "learning_rate": 5.619999999999999e-05,
      "loss": 5.8755,
      "step": 570
    },
    {
      "epoch": 0.35690934964039844,
      "grad_norm": 39.241878509521484,
      "learning_rate": 5.72e-05,
      "loss": 5.8417,
      "step": 580
    },
    {
      "epoch": 0.363062959116957,
      "grad_norm": 46.391117095947266,
      "learning_rate": 5.82e-05,
      "loss": 5.8327,
      "step": 590
    },
    {
      "epoch": 0.36921656859351565,
      "grad_norm": 51.62196350097656,
      "learning_rate": 5.9199999999999996e-05,
      "loss": 5.7876,
      "step": 600
    },
    {
      "epoch": 0.3753701780700742,
      "grad_norm": 45.93042755126953,
      "learning_rate": 6.019999999999999e-05,
      "loss": 5.7931,
      "step": 610
    },
    {
      "epoch": 0.3815237875466328,
      "grad_norm": 47.86789321899414,
      "learning_rate": 6.12e-05,
      "loss": 5.7766,
      "step": 620
    },
    {
      "epoch": 0.38767739702319143,
      "grad_norm": 51.02197265625,
      "learning_rate": 6.22e-05,
      "loss": 5.7919,
      "step": 630
    },
    {
      "epoch": 0.39383100649975,
      "grad_norm": 43.17515182495117,
      "learning_rate": 6.319999999999999e-05,
      "loss": 5.7334,
      "step": 640
    },
    {
      "epoch": 0.3999846159763086,
      "grad_norm": 38.8260612487793,
      "learning_rate": 6.419999999999999e-05,
      "loss": 5.7261,
      "step": 650
    },
    {
      "epoch": 0.4061382254528672,
      "grad_norm": 46.640254974365234,
      "learning_rate": 6.519999999999999e-05,
      "loss": 5.7115,
      "step": 660
    },
    {
      "epoch": 0.4122918349294258,
      "grad_norm": 44.81010818481445,
      "learning_rate": 6.62e-05,
      "loss": 5.6818,
      "step": 670
    },
    {
      "epoch": 0.4184454444059844,
      "grad_norm": 49.53409957885742,
      "learning_rate": 6.72e-05,
      "loss": 5.6717,
      "step": 680
    },
    {
      "epoch": 0.424599053882543,
      "grad_norm": 43.571102142333984,
      "learning_rate": 6.819999999999999e-05,
      "loss": 5.6472,
      "step": 690
    },
    {
      "epoch": 0.43075266335910156,
      "grad_norm": 52.427913665771484,
      "learning_rate": 6.919999999999999e-05,
      "loss": 5.6295,
      "step": 700
    },
    {
      "epoch": 0.4369062728356602,
      "grad_norm": 41.100425720214844,
      "learning_rate": 7.02e-05,
      "loss": 5.6157,
      "step": 710
    },
    {
      "epoch": 0.44305988231221877,
      "grad_norm": 46.8226432800293,
      "learning_rate": 7.12e-05,
      "loss": 5.5794,
      "step": 720
    },
    {
      "epoch": 0.44921349178877734,
      "grad_norm": 33.2792854309082,
      "learning_rate": 7.219999999999999e-05,
      "loss": 5.5731,
      "step": 730
    },
    {
      "epoch": 0.45536710126533597,
      "grad_norm": 44.90331268310547,
      "learning_rate": 7.319999999999999e-05,
      "loss": 5.5528,
      "step": 740
    },
    {
      "epoch": 0.46152071074189455,
      "grad_norm": 48.1407356262207,
      "learning_rate": 7.419999999999999e-05,
      "loss": 5.5387,
      "step": 750
    },
    {
      "epoch": 0.4676743202184531,
      "grad_norm": 37.969139099121094,
      "learning_rate": 7.519999999999998e-05,
      "loss": 5.5043,
      "step": 760
    },
    {
      "epoch": 0.47382792969501175,
      "grad_norm": 35.248779296875,
      "learning_rate": 7.62e-05,
      "loss": 5.4854,
      "step": 770
    },
    {
      "epoch": 0.4799815391715703,
      "grad_norm": 32.92762756347656,
      "learning_rate": 7.72e-05,
      "loss": 5.4737,
      "step": 780
    },
    {
      "epoch": 0.4861351486481289,
      "grad_norm": 44.820735931396484,
      "learning_rate": 7.819999999999999e-05,
      "loss": 5.4688,
      "step": 790
    },
    {
      "epoch": 0.49228875812468753,
      "grad_norm": 46.39222717285156,
      "learning_rate": 7.92e-05,
      "loss": 5.4436,
      "step": 800
    },
    {
      "epoch": 0.4984423676012461,
      "grad_norm": 32.6016845703125,
      "learning_rate": 8.019999999999998e-05,
      "loss": 5.401,
      "step": 810
    },
    {
      "epoch": 0.5045959770778047,
      "grad_norm": 35.57933044433594,
      "learning_rate": 8.12e-05,
      "loss": 5.3821,
      "step": 820
    },
    {
      "epoch": 0.5107495865543633,
      "grad_norm": 37.136383056640625,
      "learning_rate": 8.22e-05,
      "loss": 5.3444,
      "step": 830
    },
    {
      "epoch": 0.5169031960309219,
      "grad_norm": 34.6875,
      "learning_rate": 8.319999999999999e-05,
      "loss": 5.354,
      "step": 840
    },
    {
      "epoch": 0.5230568055074805,
      "grad_norm": 40.43978500366211,
      "learning_rate": 8.42e-05,
      "loss": 5.3268,
      "step": 850
    },
    {
      "epoch": 0.5292104149840391,
      "grad_norm": 40.99334716796875,
      "learning_rate": 8.519999999999998e-05,
      "loss": 5.3141,
      "step": 860
    },
    {
      "epoch": 0.5353640244605977,
      "grad_norm": 32.84165954589844,
      "learning_rate": 8.62e-05,
      "loss": 5.2781,
      "step": 870
    },
    {
      "epoch": 0.5415176339371562,
      "grad_norm": 35.99928665161133,
      "learning_rate": 8.719999999999999e-05,
      "loss": 5.2663,
      "step": 880
    },
    {
      "epoch": 0.5476712434137149,
      "grad_norm": 36.6622314453125,
      "learning_rate": 8.819999999999999e-05,
      "loss": 5.2482,
      "step": 890
    },
    {
      "epoch": 0.5538248528902735,
      "grad_norm": 34.90840530395508,
      "learning_rate": 8.92e-05,
      "loss": 5.231,
      "step": 900
    }
  ],
  "logging_steps": 10,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 62,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.81982525796352e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}