{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.07646056719110131,
  "eval_steps": 500,
  "global_step": 3052,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000776630924942379,
      "grad_norm": 23.55321502685547,
      "learning_rate": 1.0157273918741808e-06,
      "loss": 8.3443,
      "step": 31
    },
    {
      "epoch": 0.001553261849884758,
      "grad_norm": 14.414477348327637,
      "learning_rate": 2.0314547837483616e-06,
      "loss": 7.198,
      "step": 62
    },
    {
      "epoch": 0.002329892774827137,
      "grad_norm": 10.739773750305176,
      "learning_rate": 3.0471821756225426e-06,
      "loss": 5.8639,
      "step": 93
    },
    {
      "epoch": 0.003106523699769516,
      "grad_norm": 5.378795623779297,
      "learning_rate": 4.062909567496723e-06,
      "loss": 4.8222,
      "step": 124
    },
    {
      "epoch": 0.003883154624711895,
      "grad_norm": 8.51831340789795,
      "learning_rate": 5.078636959370905e-06,
      "loss": 4.1983,
      "step": 155
    },
    {
      "epoch": 0.004659785549654274,
      "grad_norm": 6.32632303237915,
      "learning_rate": 6.094364351245085e-06,
      "loss": 3.8598,
      "step": 186
    },
    {
      "epoch": 0.005436416474596653,
      "grad_norm": 5.590794563293457,
      "learning_rate": 7.110091743119267e-06,
      "loss": 3.6646,
      "step": 217
    },
    {
      "epoch": 0.006213047399539032,
      "grad_norm": 5.9275383949279785,
      "learning_rate": 8.125819134993446e-06,
      "loss": 3.5212,
      "step": 248
    },
    {
      "epoch": 0.006989678324481411,
      "grad_norm": 5.041291236877441,
      "learning_rate": 9.141546526867629e-06,
      "loss": 3.3936,
      "step": 279
    },
    {
      "epoch": 0.00776630924942379,
      "grad_norm": 4.923459529876709,
      "learning_rate": 1.015727391874181e-05,
      "loss": 3.2713,
      "step": 310
    },
    {
      "epoch": 0.00854294017436617,
      "grad_norm": 8.011459350585938,
      "learning_rate": 1.117300131061599e-05,
      "loss": 3.2,
      "step": 341
    },
    {
      "epoch": 0.009319571099308548,
      "grad_norm": 5.261505603790283,
      "learning_rate": 1.218872870249017e-05,
      "loss": 3.1255,
      "step": 372
    },
    {
      "epoch": 0.010096202024250926,
      "grad_norm": 5.004406452178955,
      "learning_rate": 1.3204456094364351e-05,
      "loss": 3.0726,
      "step": 403
    },
    {
      "epoch": 0.010872832949193306,
      "grad_norm": 4.326263427734375,
      "learning_rate": 1.4220183486238533e-05,
      "loss": 2.9682,
      "step": 434
    },
    {
      "epoch": 0.011649463874135685,
      "grad_norm": 4.251937389373779,
      "learning_rate": 1.5235910878112714e-05,
      "loss": 2.9215,
      "step": 465
    },
    {
      "epoch": 0.012426094799078065,
      "grad_norm": 4.35053825378418,
      "learning_rate": 1.6251638269986893e-05,
      "loss": 2.8902,
      "step": 496
    },
    {
      "epoch": 0.013202725724020443,
      "grad_norm": 4.316098690032959,
      "learning_rate": 1.7267365661861077e-05,
      "loss": 2.8521,
      "step": 527
    },
    {
      "epoch": 0.013979356648962823,
      "grad_norm": 5.536881446838379,
      "learning_rate": 1.8283093053735257e-05,
      "loss": 2.8058,
      "step": 558
    },
    {
      "epoch": 0.014755987573905201,
      "grad_norm": 3.759843111038208,
      "learning_rate": 1.9298820445609438e-05,
      "loss": 2.7438,
      "step": 589
    },
    {
      "epoch": 0.01553261849884758,
      "grad_norm": 4.057703018188477,
      "learning_rate": 2.031454783748362e-05,
      "loss": 2.7508,
      "step": 620
    },
    {
      "epoch": 0.01630924942378996,
      "grad_norm": 3.427708625793457,
      "learning_rate": 2.13302752293578e-05,
      "loss": 2.7046,
      "step": 651
    },
    {
      "epoch": 0.01708588034873234,
      "grad_norm": 3.9012484550476074,
      "learning_rate": 2.234600262123198e-05,
      "loss": 2.6706,
      "step": 682
    },
    {
      "epoch": 0.017862511273674716,
      "grad_norm": 3.388504981994629,
      "learning_rate": 2.336173001310616e-05,
      "loss": 2.6377,
      "step": 713
    },
    {
      "epoch": 0.018639142198617096,
      "grad_norm": 3.2011327743530273,
      "learning_rate": 2.437745740498034e-05,
      "loss": 2.6214,
      "step": 744
    },
    {
      "epoch": 0.019415773123559476,
      "grad_norm": 3.5897209644317627,
      "learning_rate": 2.5393184796854525e-05,
      "loss": 2.6045,
      "step": 775
    },
    {
      "epoch": 0.020192404048501852,
      "grad_norm": 3.386697769165039,
      "learning_rate": 2.6408912188728702e-05,
      "loss": 2.6066,
      "step": 806
    },
    {
      "epoch": 0.020969034973444232,
      "grad_norm": 2.9135866165161133,
      "learning_rate": 2.7424639580602886e-05,
      "loss": 2.5697,
      "step": 837
    },
    {
      "epoch": 0.021745665898386612,
      "grad_norm": 9.212065696716309,
      "learning_rate": 2.8440366972477066e-05,
      "loss": 2.5773,
      "step": 868
    },
    {
      "epoch": 0.022522296823328992,
      "grad_norm": 3.0339765548706055,
      "learning_rate": 2.9456094364351244e-05,
      "loss": 2.5453,
      "step": 899
    },
    {
      "epoch": 0.02329892774827137,
      "grad_norm": 2.8007607460021973,
      "learning_rate": 3.0471821756225428e-05,
      "loss": 2.5297,
      "step": 930
    },
    {
      "epoch": 0.02407555867321375,
      "grad_norm": 3.251768112182617,
      "learning_rate": 3.148754914809961e-05,
      "loss": 2.5273,
      "step": 961
    },
    {
      "epoch": 0.02485218959815613,
      "grad_norm": 3.357011079788208,
      "learning_rate": 3.2503276539973785e-05,
      "loss": 2.4896,
      "step": 992
    },
    {
      "epoch": 0.025628820523098506,
      "grad_norm": 2.6069107055664062,
      "learning_rate": 3.351900393184797e-05,
      "loss": 2.4718,
      "step": 1023
    },
    {
      "epoch": 0.026405451448040886,
      "grad_norm": 4.663620948791504,
      "learning_rate": 3.453473132372215e-05,
      "loss": 2.4338,
      "step": 1054
    },
    {
      "epoch": 0.027182082372983266,
      "grad_norm": 3.819638252258301,
      "learning_rate": 3.555045871559633e-05,
      "loss": 2.4848,
      "step": 1085
    },
    {
      "epoch": 0.027958713297925646,
      "grad_norm": 2.6367456912994385,
      "learning_rate": 3.6566186107470514e-05,
      "loss": 2.4415,
      "step": 1116
    },
    {
      "epoch": 0.028735344222868022,
      "grad_norm": 2.6118173599243164,
      "learning_rate": 3.7581913499344695e-05,
      "loss": 2.4341,
      "step": 1147
    },
    {
      "epoch": 0.029511975147810402,
      "grad_norm": 9.820366859436035,
      "learning_rate": 3.8597640891218876e-05,
      "loss": 2.3773,
      "step": 1178
    },
    {
      "epoch": 0.030288606072752782,
      "grad_norm": 38.890506744384766,
      "learning_rate": 3.9613368283093056e-05,
      "loss": 2.3966,
      "step": 1209
    },
    {
      "epoch": 0.03106523699769516,
      "grad_norm": 2.5037872791290283,
      "learning_rate": 4.062909567496724e-05,
      "loss": 2.3421,
      "step": 1240
    },
    {
      "epoch": 0.03184186792263754,
      "grad_norm": 2.426544189453125,
      "learning_rate": 4.164482306684142e-05,
      "loss": 2.3568,
      "step": 1271
    },
    {
      "epoch": 0.03261849884757992,
      "grad_norm": 2.5375614166259766,
      "learning_rate": 4.26605504587156e-05,
      "loss": 2.3314,
      "step": 1302
    },
    {
      "epoch": 0.0333951297725223,
      "grad_norm": 3.555330276489258,
      "learning_rate": 4.367627785058978e-05,
      "loss": 2.3262,
      "step": 1333
    },
    {
      "epoch": 0.03417176069746468,
      "grad_norm": 2.2571263313293457,
      "learning_rate": 4.469200524246396e-05,
      "loss": 2.3167,
      "step": 1364
    },
    {
      "epoch": 0.03494839162240705,
      "grad_norm": 2.184305429458618,
      "learning_rate": 4.570773263433814e-05,
      "loss": 2.2936,
      "step": 1395
    },
    {
      "epoch": 0.03572502254734943,
      "grad_norm": 2.8295061588287354,
      "learning_rate": 4.672346002621232e-05,
      "loss": 2.3033,
      "step": 1426
    },
    {
      "epoch": 0.03650165347229181,
      "grad_norm": 2.0241498947143555,
      "learning_rate": 4.77391874180865e-05,
      "loss": 2.2833,
      "step": 1457
    },
    {
      "epoch": 0.03727828439723419,
      "grad_norm": 2.0029919147491455,
      "learning_rate": 4.875491480996068e-05,
      "loss": 2.2663,
      "step": 1488
    },
    {
      "epoch": 0.03805491532217657,
      "grad_norm": 2.033773899078369,
      "learning_rate": 4.977064220183487e-05,
      "loss": 2.2597,
      "step": 1519
    },
    {
      "epoch": 0.03883154624711895,
      "grad_norm": 2.047210216522217,
      "learning_rate": 4.9999915451558777e-05,
      "loss": 2.2231,
      "step": 1550
    },
    {
      "epoch": 0.03960817717206133,
      "grad_norm": 2.032099485397339,
      "learning_rate": 4.999955597496219e-05,
      "loss": 2.2411,
      "step": 1581
    },
    {
      "epoch": 0.040384808097003705,
      "grad_norm": 1.8384301662445068,
      "learning_rate": 4.9998914381774255e-05,
      "loss": 2.2409,
      "step": 1612
    },
    {
      "epoch": 0.041161439021946085,
      "grad_norm": 1.721187710762024,
      "learning_rate": 4.999799067923527e-05,
      "loss": 2.193,
      "step": 1643
    },
    {
      "epoch": 0.041938069946888465,
      "grad_norm": 1.8869743347167969,
      "learning_rate": 4.999678487776908e-05,
      "loss": 2.1917,
      "step": 1674
    },
    {
      "epoch": 0.042714700871830845,
      "grad_norm": 1.796583652496338,
      "learning_rate": 4.9995296990983006e-05,
      "loss": 2.1794,
      "step": 1705
    },
    {
      "epoch": 0.043491331796773225,
      "grad_norm": 1.6308804750442505,
      "learning_rate": 4.999352703566763e-05,
      "loss": 2.1854,
      "step": 1736
    },
    {
      "epoch": 0.044267962721715605,
      "grad_norm": 1.683474063873291,
      "learning_rate": 4.999147503179668e-05,
      "loss": 2.1797,
      "step": 1767
    },
    {
      "epoch": 0.045044593646657985,
      "grad_norm": 1.7442835569381714,
      "learning_rate": 4.998914100252672e-05,
      "loss": 2.1695,
      "step": 1798
    },
    {
      "epoch": 0.04582122457160036,
      "grad_norm": 1.704970359802246,
      "learning_rate": 4.998652497419696e-05,
      "loss": 2.1365,
      "step": 1829
    },
    {
      "epoch": 0.04659785549654274,
      "grad_norm": 1.7213023900985718,
      "learning_rate": 4.9983626976328927e-05,
      "loss": 2.1524,
      "step": 1860
    },
    {
      "epoch": 0.04737448642148512,
      "grad_norm": 4.01851224899292,
      "learning_rate": 4.998044704162613e-05,
      "loss": 2.1171,
      "step": 1891
    },
    {
      "epoch": 0.0481511173464275,
      "grad_norm": 1.5619858503341675,
      "learning_rate": 4.9976985205973705e-05,
      "loss": 2.1608,
      "step": 1922
    },
    {
      "epoch": 0.04892774827136988,
      "grad_norm": 1.684493064880371,
      "learning_rate": 4.997324150843799e-05,
      "loss": 2.1278,
      "step": 1953
    },
    {
      "epoch": 0.04970437919631226,
      "grad_norm": 1.660992980003357,
      "learning_rate": 4.99692159912661e-05,
      "loss": 2.1136,
      "step": 1984
    },
    {
      "epoch": 0.05048101012125464,
      "grad_norm": 1.5491065979003906,
      "learning_rate": 4.996490869988546e-05,
      "loss": 2.1086,
      "step": 2015
    },
    {
      "epoch": 0.05125764104619701,
      "grad_norm": 1.5400840044021606,
      "learning_rate": 4.996031968290326e-05,
      "loss": 2.1069,
      "step": 2046
    },
    {
      "epoch": 0.05203427197113939,
      "grad_norm": 1.5687416791915894,
      "learning_rate": 4.995544899210594e-05,
      "loss": 2.0888,
      "step": 2077
    },
    {
      "epoch": 0.05281090289608177,
      "grad_norm": 1.5218935012817383,
      "learning_rate": 4.9950296682458583e-05,
      "loss": 2.1032,
      "step": 2108
    },
    {
      "epoch": 0.05358753382102415,
      "grad_norm": 1.763666033744812,
      "learning_rate": 4.994486281210429e-05,
      "loss": 2.0799,
      "step": 2139
    },
    {
      "epoch": 0.05436416474596653,
      "grad_norm": 1.5606274604797363,
      "learning_rate": 4.9939147442363566e-05,
      "loss": 2.0614,
      "step": 2170
    },
    {
      "epoch": 0.05514079567090891,
      "grad_norm": 1.4900213479995728,
      "learning_rate": 4.9933150637733574e-05,
      "loss": 2.0565,
      "step": 2201
    },
    {
      "epoch": 0.05591742659585129,
      "grad_norm": 1.4454214572906494,
      "learning_rate": 4.992687246588743e-05,
      "loss": 2.0381,
      "step": 2232
    },
    {
      "epoch": 0.056694057520793664,
      "grad_norm": 1.4872742891311646,
      "learning_rate": 4.992031299767347e-05,
      "loss": 2.0227,
      "step": 2263
    },
    {
      "epoch": 0.057470688445736044,
      "grad_norm": 1.4538527727127075,
      "learning_rate": 4.9913472307114386e-05,
      "loss": 2.0395,
      "step": 2294
    },
    {
      "epoch": 0.058247319370678424,
      "grad_norm": 1.432769536972046,
      "learning_rate": 4.9906350471406446e-05,
      "loss": 2.0352,
      "step": 2325
    },
    {
      "epoch": 0.059023950295620804,
      "grad_norm": 1.4535977840423584,
      "learning_rate": 4.989894757091861e-05,
      "loss": 2.0341,
      "step": 2356
    },
    {
      "epoch": 0.059800581220563184,
      "grad_norm": 1.4004398584365845,
      "learning_rate": 4.989126368919158e-05,
      "loss": 2.0185,
      "step": 2387
    },
    {
      "epoch": 0.060577212145505564,
      "grad_norm": 1.433273434638977,
      "learning_rate": 4.988329891293693e-05,
      "loss": 2.036,
      "step": 2418
    },
    {
      "epoch": 0.06135384307044794,
      "grad_norm": 1.3791290521621704,
      "learning_rate": 4.987505333203608e-05,
      "loss": 2.0421,
      "step": 2449
    },
    {
      "epoch": 0.06213047399539032,
      "grad_norm": 2.008694648742676,
      "learning_rate": 4.9866527039539276e-05,
      "loss": 1.9948,
      "step": 2480
    },
    {
      "epoch": 0.0629071049203327,
      "grad_norm": 1.3773164749145508,
      "learning_rate": 4.9857720131664594e-05,
      "loss": 2.0158,
      "step": 2511
    },
    {
      "epoch": 0.06368373584527508,
      "grad_norm": 1.4489777088165283,
      "learning_rate": 4.9848632707796773e-05,
      "loss": 2.0216,
      "step": 2542
    },
    {
      "epoch": 0.06446036677021745,
      "grad_norm": 2.1266908645629883,
      "learning_rate": 4.9839264870486155e-05,
      "loss": 1.9962,
      "step": 2573
    },
    {
      "epoch": 0.06523699769515984,
      "grad_norm": 1.3704332113265991,
      "learning_rate": 4.9829616725447526e-05,
      "loss": 1.9801,
      "step": 2604
    },
    {
      "epoch": 0.06601362862010221,
      "grad_norm": 1.3643279075622559,
      "learning_rate": 4.981968838155888e-05,
      "loss": 1.9902,
      "step": 2635
    },
    {
      "epoch": 0.0667902595450446,
      "grad_norm": 1.3495409488677979,
      "learning_rate": 4.980947995086024e-05,
      "loss": 1.993,
      "step": 2666
    },
    {
      "epoch": 0.06756689046998697,
      "grad_norm": 1.3918874263763428,
      "learning_rate": 4.979899154855234e-05,
      "loss": 1.9782,
      "step": 2697
    },
    {
      "epoch": 0.06834352139492936,
      "grad_norm": 1.3205535411834717,
      "learning_rate": 4.9788223292995386e-05,
      "loss": 1.9638,
      "step": 2728
    },
    {
      "epoch": 0.06912015231987173,
      "grad_norm": 1.3123114109039307,
      "learning_rate": 4.977717530570768e-05,
      "loss": 1.9826,
      "step": 2759
    },
    {
      "epoch": 0.0698967832448141,
      "grad_norm": 1.3138995170593262,
      "learning_rate": 4.976584771136425e-05,
      "loss": 1.9938,
      "step": 2790
    },
    {
      "epoch": 0.07067341416975649,
      "grad_norm": 1.2964009046554565,
      "learning_rate": 4.975424063779547e-05,
      "loss": 1.9924,
      "step": 2821
    },
    {
      "epoch": 0.07145004509469886,
      "grad_norm": 1.3010739088058472,
      "learning_rate": 4.974235421598557e-05,
      "loss": 1.9592,
      "step": 2852
    },
    {
      "epoch": 0.07222667601964125,
      "grad_norm": 1.2549103498458862,
      "learning_rate": 4.973018858007122e-05,
      "loss": 1.9394,
      "step": 2883
    },
    {
      "epoch": 0.07300330694458362,
      "grad_norm": 1.3198585510253906,
      "learning_rate": 4.9717743867339963e-05,
      "loss": 1.989,
      "step": 2914
    },
    {
      "epoch": 0.07377993786952601,
      "grad_norm": 1.3653398752212524,
      "learning_rate": 4.9705020218228695e-05,
      "loss": 1.9499,
      "step": 2945
    },
    {
      "epoch": 0.07455656879446838,
      "grad_norm": 4.606905937194824,
      "learning_rate": 4.969201777632205e-05,
      "loss": 1.9504,
      "step": 2976
    },
    {
      "epoch": 0.07533319971941076,
      "grad_norm": 1.2823129892349243,
      "learning_rate": 4.9678736688350846e-05,
      "loss": 1.9483,
      "step": 3007
    },
    {
      "epoch": 0.07610983064435314,
      "grad_norm": 1.2803308963775635,
      "learning_rate": 4.966517710419033e-05,
      "loss": 1.9604,
      "step": 3038
    }
  ],
  "logging_steps": 31,
  "max_steps": 30517,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3052,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.131685419530584e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}