{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9968798751950079,
  "eval_steps": 500,
  "global_step": 960,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02080083203328133,
      "grad_norm": 2.954287528991699,
      "learning_rate": 6.25e-06,
      "loss": 0.3901,
      "step": 10
    },
    {
      "epoch": 0.04160166406656266,
      "grad_norm": 0.9206520318984985,
      "learning_rate": 1.25e-05,
      "loss": 0.2205,
      "step": 20
    },
    {
      "epoch": 0.062402496099843996,
      "grad_norm": 0.7870351076126099,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.1896,
      "step": 30
    },
    {
      "epoch": 0.08320332813312532,
      "grad_norm": 0.8045545220375061,
      "learning_rate": 2.5e-05,
      "loss": 0.1908,
      "step": 40
    },
    {
      "epoch": 0.10400416016640665,
      "grad_norm": 0.6852442622184753,
      "learning_rate": 2.9996796251818968e-05,
      "loss": 0.1623,
      "step": 50
    },
    {
      "epoch": 0.12480499219968799,
      "grad_norm": 0.6457542181015015,
      "learning_rate": 2.9884808696055675e-05,
      "loss": 0.164,
      "step": 60
    },
    {
      "epoch": 0.1456058242329693,
      "grad_norm": 0.6373040676116943,
      "learning_rate": 2.9613999639484314e-05,
      "loss": 0.1737,
      "step": 70
    },
    {
      "epoch": 0.16640665626625065,
      "grad_norm": 0.8400445580482483,
      "learning_rate": 2.9187258625509518e-05,
      "loss": 0.1637,
      "step": 80
    },
    {
      "epoch": 0.187207488299532,
      "grad_norm": 0.5773469805717468,
      "learning_rate": 2.86091389977234e-05,
      "loss": 0.1647,
      "step": 90
    },
    {
      "epoch": 0.2080083203328133,
      "grad_norm": 0.599101722240448,
      "learning_rate": 2.788580931554828e-05,
      "loss": 0.1653,
      "step": 100
    },
    {
      "epoch": 0.22880915236609464,
      "grad_norm": 1.818597435951233,
      "learning_rate": 2.7024987535462327e-05,
      "loss": 0.165,
      "step": 110
    },
    {
      "epoch": 0.24960998439937598,
      "grad_norm": 0.5002009272575378,
      "learning_rate": 2.6035858660096975e-05,
      "loss": 0.1638,
      "step": 120
    },
    {
      "epoch": 0.2704108164326573,
      "grad_norm": 0.44137582182884216,
      "learning_rate": 2.49289767338935e-05,
      "loss": 0.1355,
      "step": 130
    },
    {
      "epoch": 0.2912116484659386,
      "grad_norm": 0.49017268419265747,
      "learning_rate": 2.3716152231029077e-05,
      "loss": 0.1621,
      "step": 140
    },
    {
      "epoch": 0.31201248049922,
      "grad_norm": 0.42483949661254883,
      "learning_rate": 2.2410326037187558e-05,
      "loss": 0.1518,
      "step": 150
    },
    {
      "epoch": 0.3328133125325013,
      "grad_norm": 0.4569384455680847,
      "learning_rate": 2.1025431369794546e-05,
      "loss": 0.1346,
      "step": 160
    },
    {
      "epoch": 0.3536141445657826,
      "grad_norm": 0.4882504642009735,
      "learning_rate": 1.9576245110033233e-05,
      "loss": 0.1534,
      "step": 170
    },
    {
      "epoch": 0.374414976599064,
      "grad_norm": 0.4138477146625519,
      "learning_rate": 1.8078230132934514e-05,
      "loss": 0.1443,
      "step": 180
    },
    {
      "epoch": 0.3952158086323453,
      "grad_norm": 0.43387019634246826,
      "learning_rate": 1.6547370317885354e-05,
      "loss": 0.148,
      "step": 190
    },
    {
      "epoch": 0.4160166406656266,
      "grad_norm": 0.414253294467926,
      "learning_rate": 1.5e-05,
      "loss": 0.1412,
      "step": 200
    },
    {
      "epoch": 0.43681747269890797,
      "grad_norm": 0.43514806032180786,
      "learning_rate": 1.3452629682114646e-05,
      "loss": 0.1331,
      "step": 210
    },
    {
      "epoch": 0.4576183047321893,
      "grad_norm": 0.3686197102069855,
      "learning_rate": 1.1921769867065483e-05,
      "loss": 0.1317,
      "step": 220
    },
    {
      "epoch": 0.4784191367654706,
      "grad_norm": 0.4298912584781647,
      "learning_rate": 1.042375488996677e-05,
      "loss": 0.1332,
      "step": 230
    },
    {
      "epoch": 0.49921996879875197,
      "grad_norm": 0.4293174743652344,
      "learning_rate": 8.974568630205462e-06,
      "loss": 0.1402,
      "step": 240
    },
    {
      "epoch": 0.5200208008320333,
      "grad_norm": 0.4327556788921356,
      "learning_rate": 7.589673962812442e-06,
      "loss": 0.1433,
      "step": 250
    },
    {
      "epoch": 0.5408216328653146,
      "grad_norm": 0.3953148424625397,
      "learning_rate": 6.283847768970927e-06,
      "loss": 0.1416,
      "step": 260
    },
    {
      "epoch": 0.5616224648985959,
      "grad_norm": 0.3659592866897583,
      "learning_rate": 5.071023266106502e-06,
      "loss": 0.1262,
      "step": 270
    },
    {
      "epoch": 0.5824232969318772,
      "grad_norm": 0.36430105566978455,
      "learning_rate": 3.964141339903026e-06,
      "loss": 0.1153,
      "step": 280
    },
    {
      "epoch": 0.6032241289651586,
      "grad_norm": 0.38800716400146484,
      "learning_rate": 2.975012464537676e-06,
      "loss": 0.1293,
      "step": 290
    },
    {
      "epoch": 0.62402496099844,
      "grad_norm": 0.3738207519054413,
      "learning_rate": 2.1141906844517207e-06,
      "loss": 0.1308,
      "step": 300
    },
    {
      "epoch": 0.6448257930317213,
      "grad_norm": 0.3933092951774597,
      "learning_rate": 1.390861002276602e-06,
      "loss": 0.122,
      "step": 310
    },
    {
      "epoch": 0.6656266250650026,
      "grad_norm": 0.37102991342544556,
      "learning_rate": 8.127413744904805e-07,
      "loss": 0.1165,
      "step": 320
    },
    {
      "epoch": 0.6864274570982839,
      "grad_norm": 0.37362316250801086,
      "learning_rate": 3.860003605156881e-07,
      "loss": 0.1344,
      "step": 330
    },
    {
      "epoch": 0.7072282891315652,
      "grad_norm": 0.3693016469478607,
      "learning_rate": 1.1519130394432476e-07,
      "loss": 0.1144,
      "step": 340
    },
    {
      "epoch": 0.7280291211648466,
      "grad_norm": 0.3761012852191925,
      "learning_rate": 3.20374818103486e-09,
      "loss": 0.1156,
      "step": 350
    },
    {
      "epoch": 0.748829953198128,
      "grad_norm": 0.3676697909832001,
      "learning_rate": 5.1232604899952296e-08,
      "loss": 0.1182,
      "step": 360
    },
    {
      "epoch": 0.7696307852314093,
      "grad_norm": 0.3497138023376465,
      "learning_rate": 2.587654042896087e-07,
      "loss": 0.1222,
      "step": 370
    },
    {
      "epoch": 0.7904316172646906,
      "grad_norm": 0.35691773891448975,
      "learning_rate": 6.23587763126211e-07,
      "loss": 0.1155,
      "step": 380
    },
    {
      "epoch": 0.8112324492979719,
      "grad_norm": 0.3574424386024475,
      "learning_rate": 1.1418070123306974e-06,
      "loss": 0.1258,
      "step": 390
    },
    {
      "epoch": 0.8320332813312532,
      "grad_norm": 0.38320979475975037,
      "learning_rate": 1.8078937319026607e-06,
      "loss": 0.1278,
      "step": 400
    },
    {
      "epoch": 0.8528341133645346,
      "grad_norm": 0.36120763421058655,
      "learning_rate": 2.614740750051663e-06,
      "loss": 0.1271,
      "step": 410
    },
    {
      "epoch": 0.8736349453978159,
      "grad_norm": 0.409435510635376,
      "learning_rate": 3.5537389770028983e-06,
      "loss": 0.1471,
      "step": 420
    },
    {
      "epoch": 0.8944357774310973,
      "grad_norm": 0.386115163564682,
      "learning_rate": 4.614869264327555e-06,
      "loss": 0.1139,
      "step": 430
    },
    {
      "epoch": 0.9152366094643786,
      "grad_norm": 0.35207945108413696,
      "learning_rate": 5.786809309654979e-06,
      "loss": 0.1101,
      "step": 440
    },
    {
      "epoch": 0.9360374414976599,
      "grad_norm": 0.44765156507492065,
      "learning_rate": 7.0570544660893674e-06,
      "loss": 0.1315,
      "step": 450
    },
    {
      "epoch": 0.9568382735309412,
      "grad_norm": 0.41203129291534424,
      "learning_rate": 8.412051167289444e-06,
      "loss": 0.1179,
      "step": 460
    },
    {
      "epoch": 0.9776391055642226,
      "grad_norm": 0.41569167375564575,
      "learning_rate": 9.837341544560421e-06,
      "loss": 0.1252,
      "step": 470
    },
    {
      "epoch": 0.9984399375975039,
      "grad_norm": 0.46924883127212524,
      "learning_rate": 1.1317717692888014e-05,
      "loss": 0.1205,
      "step": 480
    },
    {
      "epoch": 1.0192407696307852,
      "grad_norm": 0.42828407883644104,
      "learning_rate": 1.28373839398898e-05,
      "loss": 0.0949,
      "step": 490
    },
    {
      "epoch": 1.0400416016640666,
      "grad_norm": 0.3947116434574127,
      "learning_rate": 1.4380125386267791e-05,
      "loss": 0.0831,
      "step": 500
    },
    {
      "epoch": 1.0608424336973479,
      "grad_norm": 0.38285788893699646,
      "learning_rate": 1.592948091942891e-05,
      "loss": 0.1127,
      "step": 510
    },
    {
      "epoch": 1.0816432657306292,
      "grad_norm": 0.4146939218044281,
      "learning_rate": 1.7468918854211013e-05,
      "loss": 0.1104,
      "step": 520
    },
    {
      "epoch": 1.1024440977639105,
      "grad_norm": 0.41506752371788025,
      "learning_rate": 1.898201332662109e-05,
      "loss": 0.1132,
      "step": 530
    },
    {
      "epoch": 1.1232449297971918,
      "grad_norm": 0.42785730957984924,
      "learning_rate": 2.045261955845744e-05,
      "loss": 0.1006,
      "step": 540
    },
    {
      "epoch": 1.1440457618304731,
      "grad_norm": 0.3749031126499176,
      "learning_rate": 2.186504612273522e-05,
      "loss": 0.1012,
      "step": 550
    },
    {
      "epoch": 1.1648465938637544,
      "grad_norm": 0.39465105533599854,
      "learning_rate": 2.3204222371836402e-05,
      "loss": 0.1132,
      "step": 560
    },
    {
      "epoch": 1.185647425897036,
      "grad_norm": 0.4099496006965637,
      "learning_rate": 2.4455859241919324e-05,
      "loss": 0.0986,
      "step": 570
    },
    {
      "epoch": 1.2064482579303173,
      "grad_norm": 0.41943153738975525,
      "learning_rate": 2.5606601717798212e-05,
      "loss": 0.1272,
      "step": 580
    },
    {
      "epoch": 1.2272490899635986,
      "grad_norm": 0.40004250407218933,
      "learning_rate": 2.6644171331486363e-05,
      "loss": 0.1255,
      "step": 590
    },
    {
      "epoch": 1.24804992199688,
      "grad_norm": 0.5004564523696899,
      "learning_rate": 2.7557497173937928e-05,
      "loss": 0.1217,
      "step": 600
    },
    {
      "epoch": 1.2688507540301612,
      "grad_norm": 0.4227616488933563,
      "learning_rate": 2.8336834022087772e-05,
      "loss": 0.1198,
      "step": 610
    },
    {
      "epoch": 1.2896515860634425,
      "grad_norm": 0.41835710406303406,
      "learning_rate": 2.8973866320769182e-05,
      "loss": 0.1207,
      "step": 620
    },
    {
      "epoch": 1.3104524180967239,
      "grad_norm": 0.40431123971939087,
      "learning_rate": 2.9461796910018204e-05,
      "loss": 0.1186,
      "step": 630
    },
    {
      "epoch": 1.3312532501300052,
      "grad_norm": 0.41547176241874695,
      "learning_rate": 2.979541955104084e-05,
      "loss": 0.1254,
      "step": 640
    },
    {
      "epoch": 1.3520540821632865,
      "grad_norm": 0.3922775387763977,
      "learning_rate": 2.997117447698802e-05,
      "loss": 0.1285,
      "step": 650
    },
    {
      "epoch": 1.3728549141965678,
      "grad_norm": 0.3523492217063904,
      "learning_rate": 2.9987186375809513e-05,
      "loss": 0.1148,
      "step": 660
    },
    {
      "epoch": 1.3936557462298491,
      "grad_norm": 0.44066208600997925,
      "learning_rate": 2.984328439990804e-05,
      "loss": 0.1137,
      "step": 670
    },
    {
      "epoch": 1.4144565782631306,
      "grad_norm": 0.4978786110877991,
      "learning_rate": 2.9541003989089956e-05,
      "loss": 0.1197,
      "step": 680
    },
    {
      "epoch": 1.435257410296412,
      "grad_norm": 0.4303026795387268,
      "learning_rate": 2.9083570487361445e-05,
      "loss": 0.1332,
      "step": 690
    },
    {
      "epoch": 1.4560582423296933,
      "grad_norm": 0.34892433881759644,
      "learning_rate": 2.8475864728379682e-05,
      "loss": 0.1246,
      "step": 700
    },
    {
      "epoch": 1.4768590743629746,
      "grad_norm": 0.41033118963241577,
      "learning_rate": 2.772437095676361e-05,
      "loss": 0.1384,
      "step": 710
    },
    {
      "epoch": 1.497659906396256,
      "grad_norm": 0.4040924310684204,
      "learning_rate": 2.683710764094591e-05,
      "loss": 0.1151,
      "step": 720
    },
    {
      "epoch": 1.5184607384295372,
      "grad_norm": 0.33467575907707214,
      "learning_rate": 2.5823541915795936e-05,
      "loss": 0.1196,
      "step": 730
    },
    {
      "epoch": 1.5392615704628185,
      "grad_norm": 0.3744722008705139,
      "learning_rate": 2.4694488567914113e-05,
      "loss": 0.107,
      "step": 740
    },
    {
      "epoch": 1.5600624024960998,
      "grad_norm": 0.3619571626186371,
      "learning_rate": 2.3461994641428768e-05,
      "loss": 0.1324,
      "step": 750
    },
    {
      "epoch": 1.5808632345293812,
      "grad_norm": 0.39949360489845276,
      "learning_rate": 2.213921089555611e-05,
      "loss": 0.1197,
      "step": 760
    },
    {
      "epoch": 1.6016640665626625,
      "grad_norm": 0.35969704389572144,
      "learning_rate": 2.074025148547636e-05,
      "loss": 0.1209,
      "step": 770
    },
    {
      "epoch": 1.6224648985959438,
      "grad_norm": 0.36468222737312317,
      "learning_rate": 1.9280043363736577e-05,
      "loss": 0.113,
      "step": 780
    },
    {
      "epoch": 1.643265730629225,
      "grad_norm": 0.3423938453197479,
      "learning_rate": 1.7774167009073373e-05,
      "loss": 0.1149,
      "step": 790
    },
    {
      "epoch": 1.6640665626625064,
      "grad_norm": 0.38151758909225464,
      "learning_rate": 1.6238690182084996e-05,
      "loss": 0.1083,
      "step": 800
    },
    {
      "epoch": 1.6848673946957877,
      "grad_norm": 0.36454877257347107,
      "learning_rate": 1.4689996481586692e-05,
      "loss": 0.1013,
      "step": 810
    },
    {
      "epoch": 1.705668226729069,
      "grad_norm": 0.3282698690891266,
      "learning_rate": 1.3144610530959793e-05,
      "loss": 0.1004,
      "step": 820
    },
    {
      "epoch": 1.7264690587623504,
      "grad_norm": 0.313281774520874,
      "learning_rate": 1.1619021659762912e-05,
      "loss": 0.1045,
      "step": 830
    },
    {
      "epoch": 1.7472698907956317,
      "grad_norm": 0.332051157951355,
      "learning_rate": 1.0129507961929739e-05,
      "loss": 0.0952,
      "step": 840
    },
    {
      "epoch": 1.7680707228289132,
      "grad_norm": 0.3139604330062866,
      "learning_rate": 8.69196260785939e-06,
      "loss": 0.0919,
      "step": 850
    },
    {
      "epoch": 1.7888715548621945,
      "grad_norm": 0.35599997639656067,
      "learning_rate": 7.321724263655999e-06,
      "loss": 0.1008,
      "step": 860
    },
    {
      "epoch": 1.8096723868954758,
      "grad_norm": 0.30996373295783997,
      "learning_rate": 6.033413426951388e-06,
      "loss": 0.0986,
      "step": 870
    },
    {
      "epoch": 1.8304732189287571,
      "grad_norm": 0.33788877725601196,
      "learning_rate": 4.840776425613894e-06,
      "loss": 0.1144,
      "step": 880
    },
    {
      "epoch": 1.8512740509620385,
      "grad_norm": 0.3212425708770752,
      "learning_rate": 3.7565387438831093e-06,
      "loss": 0.105,
      "step": 890
    },
    {
      "epoch": 1.8720748829953198,
      "grad_norm": 0.30934497714042664,
      "learning_rate": 2.792269240947083e-06,
      "loss": 0.1027,
      "step": 900
    },
    {
      "epoch": 1.8928757150286013,
      "grad_norm": 0.3211897015571594,
      "learning_rate": 1.958256710754496e-06,
      "loss": 0.1049,
      "step": 910
    },
    {
      "epoch": 1.9136765470618826,
      "grad_norm": 0.3027827739715576,
      "learning_rate": 1.2634001001741423e-06,
      "loss": 0.0987,
      "step": 920
    },
    {
      "epoch": 1.934477379095164,
      "grad_norm": 0.28623566031455994,
      "learning_rate": 7.151135568777889e-07,
      "loss": 0.1019,
      "step": 930
    },
    {
      "epoch": 1.9552782111284452,
      "grad_norm": 0.3042950928211212,
      "learning_rate": 3.192473200896845e-07,
      "loss": 0.1044,
      "step": 940
    },
    {
      "epoch": 1.9760790431617266,
      "grad_norm": 0.32153618335723877,
      "learning_rate": 8.002529830135996e-08,
      "loss": 0.1004,
      "step": 950
    },
    {
      "epoch": 1.9968798751950079,
      "grad_norm": 0.3439652621746063,
      "learning_rate": 0.0,
      "loss": 0.0966,
      "step": 960
    },
    {
      "epoch": 1.9968798751950079,
      "step": 960,
      "total_flos": 2.5563512438029025e+18,
      "train_loss": 0.12861698282261688,
      "train_runtime": 5934.2279,
      "train_samples_per_second": 10.368,
      "train_steps_per_second": 0.162
    }
  ],
  "logging_steps": 10,
  "max_steps": 960,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.5563512438029025e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}