{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.15292113438220262,
  "eval_steps": 500,
  "global_step": 6104,
  "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
    },
    {
      "epoch": 0.07688646156929552,
      "grad_norm": 1.2765967845916748,
      "learning_rate": 4.965133917685858e-05,
      "loss": 1.941,
      "step": 3069
    },
    {
      "epoch": 0.0776630924942379,
      "grad_norm": 2.622286796569824,
      "learning_rate": 4.9637223062514714e-05,
      "loss": 1.9426,
      "step": 3100
    },
    {
      "epoch": 0.07843972341918028,
      "grad_norm": 1.231387972831726,
      "learning_rate": 4.962282892045718e-05,
      "loss": 1.9416,
      "step": 3131
    },
    {
      "epoch": 0.07921635434412266,
      "grad_norm": 1.2499909400939941,
      "learning_rate": 4.9608156913121904e-05,
      "loss": 1.9226,
      "step": 3162
    },
    {
      "epoch": 0.07999298526906504,
      "grad_norm": 1.296276330947876,
      "learning_rate": 4.959320720608049e-05,
      "loss": 1.935,
      "step": 3193
    },
    {
      "epoch": 0.08076961619400741,
      "grad_norm": 1.2642258405685425,
      "learning_rate": 4.9577979968038354e-05,
      "loss": 1.9185,
      "step": 3224
    },
    {
      "epoch": 0.0815462471189498,
      "grad_norm": 1.2396386861801147,
      "learning_rate": 4.956247537083282e-05,
      "loss": 1.9361,
      "step": 3255
    },
    {
      "epoch": 0.08232287804389217,
      "grad_norm": 1.1811271905899048,
      "learning_rate": 4.9546693589431145e-05,
      "loss": 1.9276,
      "step": 3286
    },
    {
      "epoch": 0.08309950896883456,
      "grad_norm": 1.2897577285766602,
      "learning_rate": 4.9530634801928595e-05,
      "loss": 1.9277,
      "step": 3317
    },
    {
      "epoch": 0.08387613989377693,
      "grad_norm": 1.1453903913497925,
      "learning_rate": 4.9514299189546395e-05,
      "loss": 1.9109,
      "step": 3348
    },
    {
      "epoch": 0.08465277081871932,
      "grad_norm": 1.1600992679595947,
      "learning_rate": 4.949768693662973e-05,
      "loss": 1.9248,
      "step": 3379
    },
    {
      "epoch": 0.08542940174366169,
      "grad_norm": 1.234217882156372,
      "learning_rate": 4.948079823064559e-05,
      "loss": 1.9236,
      "step": 3410
    },
    {
      "epoch": 0.08620603266860406,
      "grad_norm": 1.1933766603469849,
      "learning_rate": 4.946363326218074e-05,
      "loss": 1.9284,
      "step": 3441
    },
    {
      "epoch": 0.08698266359354645,
      "grad_norm": 1.2087618112564087,
      "learning_rate": 4.9446192224939525e-05,
      "loss": 1.8971,
      "step": 3472
    },
    {
      "epoch": 0.08775929451848882,
      "grad_norm": 1.227864146232605,
      "learning_rate": 4.942847531574167e-05,
      "loss": 1.9353,
      "step": 3503
    },
    {
      "epoch": 0.08853592544343121,
      "grad_norm": 1.2828229665756226,
      "learning_rate": 4.941048273452008e-05,
      "loss": 1.9085,
      "step": 3534
    },
    {
      "epoch": 0.08931255636837358,
      "grad_norm": 1.2086588144302368,
      "learning_rate": 4.9392214684318605e-05,
      "loss": 1.9044,
      "step": 3565
    },
    {
      "epoch": 0.09008918729331597,
      "grad_norm": 1.1942152976989746,
      "learning_rate": 4.93736713712897e-05,
      "loss": 1.9182,
      "step": 3596
    },
    {
      "epoch": 0.09086581821825834,
      "grad_norm": 1.2399394512176514,
      "learning_rate": 4.9354853004692124e-05,
      "loss": 1.8943,
      "step": 3627
    },
    {
      "epoch": 0.09164244914320072,
      "grad_norm": 1.227376937866211,
      "learning_rate": 4.93357597968886e-05,
      "loss": 1.8881,
      "step": 3658
    },
    {
      "epoch": 0.0924190800681431,
      "grad_norm": 1.1314102411270142,
      "learning_rate": 4.931639196334338e-05,
      "loss": 1.8878,
      "step": 3689
    },
    {
      "epoch": 0.09319571099308548,
      "grad_norm": 1.192134141921997,
      "learning_rate": 4.9296749722619826e-05,
      "loss": 1.8857,
      "step": 3720
    },
    {
      "epoch": 0.09397234191802786,
      "grad_norm": 1.1905503273010254,
      "learning_rate": 4.9276833296377966e-05,
      "loss": 1.8701,
      "step": 3751
    },
    {
      "epoch": 0.09474897284297024,
      "grad_norm": 1.2335184812545776,
      "learning_rate": 4.925664290937196e-05,
      "loss": 1.8741,
      "step": 3782
    },
    {
      "epoch": 0.09552560376791262,
      "grad_norm": 1.1403234004974365,
      "learning_rate": 4.9236178789447576e-05,
      "loss": 1.8658,
      "step": 3813
    },
    {
      "epoch": 0.096302234692855,
      "grad_norm": 1.1734315156936646,
      "learning_rate": 4.921544116753962e-05,
      "loss": 1.8724,
      "step": 3844
    },
    {
      "epoch": 0.09707886561779737,
      "grad_norm": 1.2637231349945068,
      "learning_rate": 4.919443027766935e-05,
      "loss": 1.867,
      "step": 3875
    },
    {
      "epoch": 0.09785549654273976,
      "grad_norm": 1.1608805656433105,
      "learning_rate": 4.91731463569418e-05,
      "loss": 1.8764,
      "step": 3906
    },
    {
      "epoch": 0.09863212746768213,
      "grad_norm": 1.1562029123306274,
      "learning_rate": 4.915158964554312e-05,
      "loss": 1.8612,
      "step": 3937
    },
    {
      "epoch": 0.09940875839262452,
      "grad_norm": 1.1918668746948242,
      "learning_rate": 4.912976038673786e-05,
      "loss": 1.866,
      "step": 3968
    },
    {
      "epoch": 0.10018538931756689,
      "grad_norm": 1.180206537246704,
      "learning_rate": 4.9107658826866254e-05,
      "loss": 1.849,
      "step": 3999
    },
    {
      "epoch": 0.10096202024250928,
      "grad_norm": 1.0747199058532715,
      "learning_rate": 4.908528521534139e-05,
      "loss": 1.8547,
      "step": 4030
    },
    {
      "epoch": 0.10173865116745165,
      "grad_norm": 1.103786826133728,
      "learning_rate": 4.906263980464644e-05,
      "loss": 1.864,
      "step": 4061
    },
    {
      "epoch": 0.10251528209239402,
      "grad_norm": 1.1390912532806396,
      "learning_rate": 4.903972285033178e-05,
      "loss": 1.8669,
      "step": 4092
    },
    {
      "epoch": 0.10329191301733641,
      "grad_norm": 1.1488468647003174,
      "learning_rate": 4.901653461101213e-05,
      "loss": 1.8445,
      "step": 4123
    },
    {
      "epoch": 0.10406854394227878,
      "grad_norm": 1.1589927673339844,
      "learning_rate": 4.8993075348363626e-05,
      "loss": 1.8529,
      "step": 4154
    },
    {
      "epoch": 0.10484517486722117,
      "grad_norm": 1.1793478727340698,
      "learning_rate": 4.896934532712084e-05,
      "loss": 1.8818,
      "step": 4185
    },
    {
      "epoch": 0.10562180579216354,
      "grad_norm": 1.165292739868164,
      "learning_rate": 4.8945344815073846e-05,
      "loss": 1.8425,
      "step": 4216
    },
    {
      "epoch": 0.10639843671710593,
      "grad_norm": 1.105946660041809,
      "learning_rate": 4.892107408306516e-05,
      "loss": 1.8327,
      "step": 4247
    },
    {
      "epoch": 0.1071750676420483,
      "grad_norm": 1.1690527200698853,
      "learning_rate": 4.889653340498669e-05,
      "loss": 1.8428,
      "step": 4278
    },
    {
      "epoch": 0.10795169856699068,
      "grad_norm": 1.1106913089752197,
      "learning_rate": 4.8871723057776664e-05,
      "loss": 1.839,
      "step": 4309
    },
    {
      "epoch": 0.10872832949193306,
      "grad_norm": 1.1590090990066528,
      "learning_rate": 4.8846643321416476e-05,
      "loss": 1.8658,
      "step": 4340
    },
    {
      "epoch": 0.10950496041687544,
      "grad_norm": 1.1315875053405762,
      "learning_rate": 4.882129447892753e-05,
      "loss": 1.8282,
      "step": 4371
    },
    {
      "epoch": 0.11028159134181782,
      "grad_norm": 1.6055009365081787,
      "learning_rate": 4.8795676816368076e-05,
      "loss": 1.8294,
      "step": 4402
    },
    {
      "epoch": 0.1110582222667602,
      "grad_norm": 1.1236201524734497,
      "learning_rate": 4.876979062282995e-05,
      "loss": 1.8325,
      "step": 4433
    },
    {
      "epoch": 0.11183485319170258,
      "grad_norm": 1.0830161571502686,
      "learning_rate": 4.8743636190435325e-05,
      "loss": 1.8451,
      "step": 4464
    },
    {
      "epoch": 0.11261148411664496,
      "grad_norm": 1.1663005352020264,
      "learning_rate": 4.871721381433344e-05,
      "loss": 1.8349,
      "step": 4495
    },
    {
      "epoch": 0.11338811504158733,
      "grad_norm": 1.2522406578063965,
      "learning_rate": 4.869052379269719e-05,
      "loss": 1.8162,
      "step": 4526
    },
    {
      "epoch": 0.11416474596652972,
      "grad_norm": 1.1271142959594727,
      "learning_rate": 4.866356642671985e-05,
      "loss": 1.8348,
      "step": 4557
    },
    {
      "epoch": 0.11494137689147209,
      "grad_norm": 1.053753137588501,
      "learning_rate": 4.8636342020611634e-05,
      "loss": 1.8376,
      "step": 4588
    },
    {
      "epoch": 0.11571800781641448,
      "grad_norm": 1.1469056606292725,
      "learning_rate": 4.860885088159626e-05,
      "loss": 1.8248,
      "step": 4619
    },
    {
      "epoch": 0.11649463874135685,
      "grad_norm": 1.116128921508789,
      "learning_rate": 4.858109331990751e-05,
      "loss": 1.8454,
      "step": 4650
    },
    {
      "epoch": 0.11727126966629922,
      "grad_norm": 1.0742651224136353,
      "learning_rate": 4.855306964878567e-05,
      "loss": 1.8231,
      "step": 4681
    },
    {
      "epoch": 0.11804790059124161,
      "grad_norm": 1.101657748222351,
      "learning_rate": 4.8524780184474084e-05,
      "loss": 1.8421,
      "step": 4712
    },
    {
      "epoch": 0.11882453151618398,
      "grad_norm": 1.0891473293304443,
      "learning_rate": 4.8496225246215496e-05,
      "loss": 1.8053,
      "step": 4743
    },
    {
      "epoch": 0.11960116244112637,
      "grad_norm": 1.1044034957885742,
      "learning_rate": 4.8467405156248505e-05,
      "loss": 1.8184,
      "step": 4774
    },
    {
      "epoch": 0.12037779336606874,
      "grad_norm": 1.0644783973693848,
      "learning_rate": 4.843832023980392e-05,
      "loss": 1.8237,
      "step": 4805
    },
    {
      "epoch": 0.12115442429101113,
      "grad_norm": 1.0427252054214478,
      "learning_rate": 4.840897082510106e-05,
      "loss": 1.8189,
      "step": 4836
    },
    {
      "epoch": 0.1219310552159535,
      "grad_norm": 1.0806516408920288,
      "learning_rate": 4.8379357243344084e-05,
      "loss": 1.7958,
      "step": 4867
    },
    {
      "epoch": 0.12270768614089587,
      "grad_norm": 1.092581868171692,
      "learning_rate": 4.8349479828718236e-05,
      "loss": 1.8229,
      "step": 4898
    },
    {
      "epoch": 0.12348431706583826,
      "grad_norm": 1.0858657360076904,
      "learning_rate": 4.8319338918386075e-05,
      "loss": 1.8165,
      "step": 4929
    },
    {
      "epoch": 0.12426094799078063,
      "grad_norm": 1.073818325996399,
      "learning_rate": 4.828893485248369e-05,
      "loss": 1.8352,
      "step": 4960
    },
    {
      "epoch": 0.125037578915723,
      "grad_norm": 1.069009780883789,
      "learning_rate": 4.825826797411682e-05,
      "loss": 1.8,
      "step": 4991
    },
    {
      "epoch": 0.1258142098406654,
      "grad_norm": 1.0416266918182373,
      "learning_rate": 4.822733862935702e-05,
      "loss": 1.8224,
      "step": 5022
    },
    {
      "epoch": 0.12659084076560778,
      "grad_norm": 1.093551516532898,
      "learning_rate": 4.819614716723775e-05,
      "loss": 1.8005,
      "step": 5053
    },
    {
      "epoch": 0.12736747169055015,
      "grad_norm": 1.026353359222412,
      "learning_rate": 4.8164693939750425e-05,
      "loss": 1.844,
      "step": 5084
    },
    {
      "epoch": 0.12814410261549253,
      "grad_norm": 1.102858304977417,
      "learning_rate": 4.813297930184042e-05,
      "loss": 1.8045,
      "step": 5115
    },
    {
      "epoch": 0.1289207335404349,
      "grad_norm": 1.0537272691726685,
      "learning_rate": 4.810100361140314e-05,
      "loss": 1.8054,
      "step": 5146
    },
    {
      "epoch": 0.1296973644653773,
      "grad_norm": 1.0389387607574463,
      "learning_rate": 4.8068767229279885e-05,
      "loss": 1.7734,
      "step": 5177
    },
    {
      "epoch": 0.13047399539031967,
      "grad_norm": 1.0274319648742676,
      "learning_rate": 4.8036270519253854e-05,
      "loss": 1.788,
      "step": 5208
    },
    {
      "epoch": 0.13125062631526205,
      "grad_norm": 1.0932128429412842,
      "learning_rate": 4.8003513848046e-05,
      "loss": 1.7813,
      "step": 5239
    },
    {
      "epoch": 0.13202725724020442,
      "grad_norm": 1.0294640064239502,
      "learning_rate": 4.79704975853109e-05,
      "loss": 1.81,
      "step": 5270
    },
    {
      "epoch": 0.13280388816514682,
      "grad_norm": 1.0617210865020752,
      "learning_rate": 4.793722210363262e-05,
      "loss": 1.8017,
      "step": 5301
    },
    {
      "epoch": 0.1335805190900892,
      "grad_norm": 1.0657798051834106,
      "learning_rate": 4.7903687778520414e-05,
      "loss": 1.7884,
      "step": 5332
    },
    {
      "epoch": 0.13435715001503157,
      "grad_norm": 0.9954230189323425,
      "learning_rate": 4.7869894988404593e-05,
      "loss": 1.7848,
      "step": 5363
    },
    {
      "epoch": 0.13513378093997394,
      "grad_norm": 1.117066740989685,
      "learning_rate": 4.783584411463221e-05,
      "loss": 1.7984,
      "step": 5394
    },
    {
      "epoch": 0.13591041186491631,
      "grad_norm": 1.0706870555877686,
      "learning_rate": 4.780153554146274e-05,
      "loss": 1.8008,
      "step": 5425
    },
    {
      "epoch": 0.13668704278985871,
      "grad_norm": 1.063238501548767,
      "learning_rate": 4.7766969656063766e-05,
      "loss": 1.7754,
      "step": 5456
    },
    {
      "epoch": 0.1374636737148011,
      "grad_norm": 1.0702152252197266,
      "learning_rate": 4.773214684850662e-05,
      "loss": 1.8037,
      "step": 5487
    },
    {
      "epoch": 0.13824030463974346,
      "grad_norm": 1.0318788290023804,
      "learning_rate": 4.769706751176193e-05,
      "loss": 1.8174,
      "step": 5518
    },
    {
      "epoch": 0.13901693556468583,
      "grad_norm": 1.0097755193710327,
      "learning_rate": 4.7661732041695264e-05,
      "loss": 1.7648,
      "step": 5549
    },
    {
      "epoch": 0.1397935664896282,
      "grad_norm": 1.063873291015625,
      "learning_rate": 4.762614083706258e-05,
      "loss": 1.809,
      "step": 5580
    },
    {
      "epoch": 0.1405701974145706,
      "grad_norm": 1.0409623384475708,
      "learning_rate": 4.759029429950581e-05,
      "loss": 1.7853,
      "step": 5611
    },
    {
      "epoch": 0.14134682833951298,
      "grad_norm": 1.0131454467773438,
      "learning_rate": 4.7554192833548235e-05,
      "loss": 1.7759,
      "step": 5642
    },
    {
      "epoch": 0.14212345926445535,
      "grad_norm": 1.1056182384490967,
      "learning_rate": 4.751783684659e-05,
      "loss": 1.7959,
      "step": 5673
    },
    {
      "epoch": 0.14290009018939773,
      "grad_norm": 1.0572400093078613,
      "learning_rate": 4.748122674890348e-05,
      "loss": 1.7597,
      "step": 5704
    },
    {
      "epoch": 0.14367672111434013,
      "grad_norm": 1.0375362634658813,
      "learning_rate": 4.7444362953628654e-05,
      "loss": 1.7719,
      "step": 5735
    },
    {
      "epoch": 0.1444533520392825,
      "grad_norm": 1.0238806009292603,
      "learning_rate": 4.7407245876768424e-05,
      "loss": 1.7966,
      "step": 5766
    },
    {
      "epoch": 0.14522998296422487,
      "grad_norm": 1.0335805416107178,
      "learning_rate": 4.736987593718397e-05,
      "loss": 1.7805,
      "step": 5797
    },
    {
      "epoch": 0.14600661388916725,
      "grad_norm": 1.08907151222229,
      "learning_rate": 4.733225355658999e-05,
      "loss": 1.7905,
      "step": 5828
    },
    {
      "epoch": 0.14678324481410962,
      "grad_norm": 1.0151680707931519,
      "learning_rate": 4.7294379159549926e-05,
      "loss": 1.7766,
      "step": 5859
    },
    {
      "epoch": 0.14755987573905202,
      "grad_norm": 0.9963664412498474,
      "learning_rate": 4.725625317347119e-05,
      "loss": 1.7708,
      "step": 5890
    },
    {
      "epoch": 0.1483365066639944,
      "grad_norm": 1.0762684345245361,
      "learning_rate": 4.7217876028600374e-05,
      "loss": 1.7998,
      "step": 5921
    },
    {
      "epoch": 0.14911313758893677,
      "grad_norm": 1.1151126623153687,
      "learning_rate": 4.717924815801832e-05,
      "loss": 1.782,
      "step": 5952
    },
    {
      "epoch": 0.14988976851387914,
      "grad_norm": 0.9970519542694092,
      "learning_rate": 4.714036999763532e-05,
      "loss": 1.7779,
      "step": 5983
    },
    {
      "epoch": 0.1506663994388215,
      "grad_norm": 1.0146524906158447,
      "learning_rate": 4.7101241986186116e-05,
      "loss": 1.7631,
      "step": 6014
    },
    {
      "epoch": 0.15144303036376391,
      "grad_norm": 1.0554558038711548,
      "learning_rate": 4.7061864565225e-05,
      "loss": 1.7729,
      "step": 6045
    },
    {
      "epoch": 0.1522196612887063,
      "grad_norm": 1.0413624048233032,
      "learning_rate": 4.702223817912081e-05,
      "loss": 1.79,
      "step": 6076
    }
  ],
  "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": 4.263370839061168e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}