{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 160,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "grad_norm": 0.43827545642852783,
      "learning_rate": 0.0001,
      "loss": 0.9667,
      "mean_token_accuracy": 0.7421795547008514,
      "num_tokens": 18527.0,
      "step": 10
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.39240607619285583,
      "learning_rate": 0.0001,
      "loss": 1.0018,
      "mean_token_accuracy": 0.728947314620018,
      "num_tokens": 37412.0,
      "step": 20
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.38427555561065674,
      "learning_rate": 0.0001,
      "loss": 1.0093,
      "mean_token_accuracy": 0.7267461150884629,
      "num_tokens": 56239.0,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.42408332228660583,
      "learning_rate": 0.0001,
      "loss": 0.9559,
      "mean_token_accuracy": 0.7414480298757553,
      "num_tokens": 75307.0,
      "step": 40
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.46401479840278625,
      "learning_rate": 0.0001,
      "loss": 1.0349,
      "mean_token_accuracy": 0.7188403815031051,
      "num_tokens": 93699.0,
      "step": 50
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4213346838951111,
      "learning_rate": 0.0001,
      "loss": 0.9546,
      "mean_token_accuracy": 0.7375499308109283,
      "num_tokens": 112724.0,
      "step": 60
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.40055790543556213,
      "learning_rate": 0.0001,
      "loss": 0.9743,
      "mean_token_accuracy": 0.7378433406352997,
      "num_tokens": 131831.0,
      "step": 70
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.45163631439208984,
      "learning_rate": 0.0001,
      "loss": 1.0189,
      "mean_token_accuracy": 0.7263641938567161,
      "num_tokens": 151163.0,
      "step": 80
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4101933538913727,
      "learning_rate": 0.0001,
      "loss": 0.9672,
      "mean_token_accuracy": 0.7346487194299698,
      "num_tokens": 169973.0,
      "step": 90
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.44324934482574463,
      "learning_rate": 0.0001,
      "loss": 1.0263,
      "mean_token_accuracy": 0.7225876390933991,
      "num_tokens": 188998.0,
      "step": 100
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.47263675928115845,
      "learning_rate": 0.0001,
      "loss": 0.9865,
      "mean_token_accuracy": 0.7291134670376778,
      "num_tokens": 207403.0,
      "step": 110
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.418999046087265,
      "learning_rate": 0.0001,
      "loss": 0.9624,
      "mean_token_accuracy": 0.7337665259838104,
      "num_tokens": 225944.0,
      "step": 120
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.48299098014831543,
      "learning_rate": 0.0001,
      "loss": 1.0097,
      "mean_token_accuracy": 0.7260663375258446,
      "num_tokens": 244001.0,
      "step": 130
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.45383280515670776,
      "learning_rate": 0.0001,
      "loss": 0.9367,
      "mean_token_accuracy": 0.7450008347630501,
      "num_tokens": 262380.0,
      "step": 140
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.3950103521347046,
      "learning_rate": 0.0001,
      "loss": 0.9511,
      "mean_token_accuracy": 0.7413627594709397,
      "num_tokens": 281319.0,
      "step": 150
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5174494981765747,
      "learning_rate": 0.0001,
      "loss": 0.8982,
      "mean_token_accuracy": 0.7503622397780418,
      "num_tokens": 301270.0,
      "step": 160
    },
    {
      "epoch": 0.8,
      "eval_loss": 1.0890034437179565,
      "eval_mean_token_accuracy": 0.7177740011364222,
      "eval_num_tokens": 301270.0,
      "eval_runtime": 18.3264,
      "eval_samples_per_second": 17.461,
      "eval_steps_per_second": 8.731,
      "step": 160
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.5073578953742981,
      "learning_rate": 0.0001,
      "loss": 0.9498,
      "mean_token_accuracy": 0.7435929998755455,
      "num_tokens": 320325.0,
      "step": 170
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4688466489315033,
      "learning_rate": 0.0001,
      "loss": 0.957,
      "mean_token_accuracy": 0.7374263137578965,
      "num_tokens": 339313.0,
      "step": 180
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.5929768085479736,
      "learning_rate": 0.0001,
      "loss": 1.02,
      "mean_token_accuracy": 0.7251972034573555,
      "num_tokens": 357224.0,
      "step": 190
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.5086683630943298,
      "learning_rate": 0.0001,
      "loss": 0.992,
      "mean_token_accuracy": 0.7353695958852768,
      "num_tokens": 376643.0,
      "step": 200
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.4679737985134125,
      "learning_rate": 0.0001,
      "loss": 0.8852,
      "mean_token_accuracy": 0.7536615505814552,
      "num_tokens": 395121.0,
      "step": 210
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.4789438843727112,
      "learning_rate": 0.0001,
      "loss": 0.9377,
      "mean_token_accuracy": 0.7422916859388351,
      "num_tokens": 413895.0,
      "step": 220
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.5487397313117981,
      "learning_rate": 0.0001,
      "loss": 0.8488,
      "mean_token_accuracy": 0.7590507015585899,
      "num_tokens": 432763.0,
      "step": 230
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.5337480306625366,
      "learning_rate": 0.0001,
      "loss": 0.8666,
      "mean_token_accuracy": 0.7579806029796601,
      "num_tokens": 451608.0,
      "step": 240
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.49510490894317627,
      "learning_rate": 0.0001,
      "loss": 0.9104,
      "mean_token_accuracy": 0.7497919097542762,
      "num_tokens": 470554.0,
      "step": 250
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.657990038394928,
      "learning_rate": 0.0001,
      "loss": 0.8951,
      "mean_token_accuracy": 0.7483552470803261,
      "num_tokens": 488654.0,
      "step": 260
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.6424926519393921,
      "learning_rate": 0.0001,
      "loss": 0.8812,
      "mean_token_accuracy": 0.7524763226509095,
      "num_tokens": 507895.0,
      "step": 270
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.556791365146637,
      "learning_rate": 0.0001,
      "loss": 0.8923,
      "mean_token_accuracy": 0.753830075263977,
      "num_tokens": 526416.0,
      "step": 280
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.5518726110458374,
      "learning_rate": 0.0001,
      "loss": 0.919,
      "mean_token_accuracy": 0.7403171643614769,
      "num_tokens": 545947.0,
      "step": 290
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.5887808799743652,
      "learning_rate": 0.0001,
      "loss": 0.9512,
      "mean_token_accuracy": 0.741435107588768,
      "num_tokens": 564070.0,
      "step": 300
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.4931485056877136,
      "learning_rate": 0.0001,
      "loss": 0.8012,
      "mean_token_accuracy": 0.775481553375721,
      "num_tokens": 583519.0,
      "step": 310
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.5791897177696228,
      "learning_rate": 0.0001,
      "loss": 0.8834,
      "mean_token_accuracy": 0.7559772610664368,
      "num_tokens": 602601.0,
      "step": 320
    },
    {
      "epoch": 1.6,
      "eval_loss": 1.0932101011276245,
      "eval_mean_token_accuracy": 0.7169803511351347,
      "eval_num_tokens": 602601.0,
      "eval_runtime": 18.3278,
      "eval_samples_per_second": 17.46,
      "eval_steps_per_second": 8.73,
      "step": 320
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.5467931032180786,
      "learning_rate": 0.0001,
      "loss": 0.9429,
      "mean_token_accuracy": 0.7412704437971115,
      "num_tokens": 621579.0,
      "step": 330
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.5784661173820496,
      "learning_rate": 0.0001,
      "loss": 0.8529,
      "mean_token_accuracy": 0.7641629755496979,
      "num_tokens": 640332.0,
      "step": 340
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.4568150043487549,
      "learning_rate": 0.0001,
      "loss": 0.8716,
      "mean_token_accuracy": 0.7545537993311882,
      "num_tokens": 659812.0,
      "step": 350
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5498734712600708,
      "learning_rate": 0.0001,
      "loss": 0.8971,
      "mean_token_accuracy": 0.752374118566513,
      "num_tokens": 678799.0,
      "step": 360
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.5936706066131592,
      "learning_rate": 0.0001,
      "loss": 0.9918,
      "mean_token_accuracy": 0.7298657700419426,
      "num_tokens": 696438.0,
      "step": 370
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.5029641389846802,
      "learning_rate": 0.0001,
      "loss": 0.9371,
      "mean_token_accuracy": 0.7457024067640304,
      "num_tokens": 715459.0,
      "step": 380
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.4978192448616028,
      "learning_rate": 0.0001,
      "loss": 0.9429,
      "mean_token_accuracy": 0.7435312032699585,
      "num_tokens": 733927.0,
      "step": 390
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5036607384681702,
      "learning_rate": 0.0001,
      "loss": 0.9169,
      "mean_token_accuracy": 0.7488818317651749,
      "num_tokens": 753286.0,
      "step": 400
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.6137449741363525,
      "learning_rate": 0.0001,
      "loss": 0.7911,
      "mean_token_accuracy": 0.7788830995559692,
      "num_tokens": 772442.0,
      "step": 410
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.5912206172943115,
      "learning_rate": 0.0001,
      "loss": 0.7946,
      "mean_token_accuracy": 0.7781713724136352,
      "num_tokens": 790693.0,
      "step": 420
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.6016007661819458,
      "learning_rate": 0.0001,
      "loss": 0.7889,
      "mean_token_accuracy": 0.7735364139080048,
      "num_tokens": 809966.0,
      "step": 430
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.5786302089691162,
      "learning_rate": 0.0001,
      "loss": 0.8027,
      "mean_token_accuracy": 0.7732229202985763,
      "num_tokens": 829851.0,
      "step": 440
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5368697047233582,
      "learning_rate": 0.0001,
      "loss": 0.8208,
      "mean_token_accuracy": 0.7645123034715653,
      "num_tokens": 848464.0,
      "step": 450
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.712200403213501,
      "learning_rate": 0.0001,
      "loss": 0.8047,
      "mean_token_accuracy": 0.7703322827816009,
      "num_tokens": 867916.0,
      "step": 460
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.7731253504753113,
      "learning_rate": 0.0001,
      "loss": 0.8328,
      "mean_token_accuracy": 0.7687120333313942,
      "num_tokens": 886785.0,
      "step": 470
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.629067599773407,
      "learning_rate": 0.0001,
      "loss": 0.8092,
      "mean_token_accuracy": 0.7728568613529205,
      "num_tokens": 905795.0,
      "step": 480
    },
    {
      "epoch": 2.4,
      "eval_loss": 1.125347375869751,
      "eval_mean_token_accuracy": 0.711500994861126,
      "eval_num_tokens": 905795.0,
      "eval_runtime": 18.3351,
      "eval_samples_per_second": 17.453,
      "eval_steps_per_second": 8.726,
      "step": 480
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.6760151982307434,
      "learning_rate": 0.0001,
      "loss": 0.8256,
      "mean_token_accuracy": 0.7666228592395783,
      "num_tokens": 925208.0,
      "step": 490
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.6286083459854126,
      "learning_rate": 0.0001,
      "loss": 0.7471,
      "mean_token_accuracy": 0.7834655106067657,
      "num_tokens": 944333.0,
      "step": 500
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.6746655702590942,
      "learning_rate": 0.0001,
      "loss": 0.7931,
      "mean_token_accuracy": 0.7782243847846985,
      "num_tokens": 962570.0,
      "step": 510
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.5208497643470764,
      "learning_rate": 0.0001,
      "loss": 0.8246,
      "mean_token_accuracy": 0.7665972113609314,
      "num_tokens": 981133.0,
      "step": 520
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.7782342433929443,
      "learning_rate": 0.0001,
      "loss": 0.9398,
      "mean_token_accuracy": 0.743607884645462,
      "num_tokens": 997998.0,
      "step": 530
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.6813472509384155,
      "learning_rate": 0.0001,
      "loss": 0.777,
      "mean_token_accuracy": 0.7756537094712257,
      "num_tokens": 1016188.0,
      "step": 540
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.6890157461166382,
      "learning_rate": 0.0001,
      "loss": 0.8493,
      "mean_token_accuracy": 0.7614479854702949,
      "num_tokens": 1035174.0,
      "step": 550
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.6942604780197144,
      "learning_rate": 0.0001,
      "loss": 0.8153,
      "mean_token_accuracy": 0.7713455215096474,
      "num_tokens": 1054954.0,
      "step": 560
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.6722708344459534,
      "learning_rate": 0.0001,
      "loss": 0.8811,
      "mean_token_accuracy": 0.7551201656460762,
      "num_tokens": 1073025.0,
      "step": 570
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.6637946963310242,
      "learning_rate": 0.0001,
      "loss": 0.7876,
      "mean_token_accuracy": 0.7747491240501404,
      "num_tokens": 1092124.0,
      "step": 580
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.6677461862564087,
      "learning_rate": 0.0001,
      "loss": 0.8539,
      "mean_token_accuracy": 0.762774932384491,
      "num_tokens": 1110204.0,
      "step": 590
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4877650737762451,
      "learning_rate": 0.0001,
      "loss": 0.8402,
      "mean_token_accuracy": 0.7657597422599792,
      "num_tokens": 1129929.0,
      "step": 600
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.6816550493240356,
      "learning_rate": 0.0001,
      "loss": 0.7319,
      "mean_token_accuracy": 0.7943529888987542,
      "num_tokens": 1149086.0,
      "step": 610
    },
    {
      "epoch": 3.1,
      "grad_norm": 0.7251542210578918,
      "learning_rate": 0.0001,
      "loss": 0.7057,
      "mean_token_accuracy": 0.7959758445620537,
      "num_tokens": 1168048.0,
      "step": 620
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.7146134376525879,
      "learning_rate": 0.0001,
      "loss": 0.7221,
      "mean_token_accuracy": 0.7933965891599655,
      "num_tokens": 1186753.0,
      "step": 630
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.6925667524337769,
      "learning_rate": 0.0001,
      "loss": 0.7238,
      "mean_token_accuracy": 0.7958982989192009,
      "num_tokens": 1205471.0,
      "step": 640
    },
    {
      "epoch": 3.2,
      "eval_loss": 1.1755249500274658,
      "eval_mean_token_accuracy": 0.7095268920063973,
      "eval_num_tokens": 1205471.0,
      "eval_runtime": 18.3277,
      "eval_samples_per_second": 17.46,
      "eval_steps_per_second": 8.73,
      "step": 640
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.6561554074287415,
      "learning_rate": 0.0001,
      "loss": 0.7076,
      "mean_token_accuracy": 0.7985713094472885,
      "num_tokens": 1225099.0,
      "step": 650
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.6188570857048035,
      "learning_rate": 0.0001,
      "loss": 0.7698,
      "mean_token_accuracy": 0.7828378319740296,
      "num_tokens": 1244074.0,
      "step": 660
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.7147638201713562,
      "learning_rate": 0.0001,
      "loss": 0.6592,
      "mean_token_accuracy": 0.8094674125313759,
      "num_tokens": 1263781.0,
      "step": 670
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.761284589767456,
      "learning_rate": 0.0001,
      "loss": 0.7407,
      "mean_token_accuracy": 0.7847389072179795,
      "num_tokens": 1282554.0,
      "step": 680
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.7286498546600342,
      "learning_rate": 0.0001,
      "loss": 0.7481,
      "mean_token_accuracy": 0.7902553915977478,
      "num_tokens": 1300680.0,
      "step": 690
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.8410418033599854,
      "learning_rate": 0.0001,
      "loss": 0.7742,
      "mean_token_accuracy": 0.7788675397634506,
      "num_tokens": 1318486.0,
      "step": 700
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.7173243165016174,
      "learning_rate": 0.0001,
      "loss": 0.7424,
      "mean_token_accuracy": 0.784093214571476,
      "num_tokens": 1337891.0,
      "step": 710
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.7642121911048889,
      "learning_rate": 0.0001,
      "loss": 0.7925,
      "mean_token_accuracy": 0.7764820963144302,
      "num_tokens": 1355920.0,
      "step": 720
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.5757165551185608,
      "learning_rate": 0.0001,
      "loss": 0.7781,
      "mean_token_accuracy": 0.7775384098291397,
      "num_tokens": 1374735.0,
      "step": 730
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.8661559820175171,
      "learning_rate": 0.0001,
      "loss": 0.7351,
      "mean_token_accuracy": 0.7928005129098892,
      "num_tokens": 1393708.0,
      "step": 740
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.6611236333847046,
      "learning_rate": 0.0001,
      "loss": 0.7353,
      "mean_token_accuracy": 0.7899313852190971,
      "num_tokens": 1412985.0,
      "step": 750
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.787530779838562,
      "learning_rate": 0.0001,
      "loss": 0.758,
      "mean_token_accuracy": 0.7841869756579399,
      "num_tokens": 1431040.0,
      "step": 760
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.8499215841293335,
      "learning_rate": 0.0001,
      "loss": 0.6919,
      "mean_token_accuracy": 0.8019102051854133,
      "num_tokens": 1450564.0,
      "step": 770
    },
    {
      "epoch": 3.9,
      "grad_norm": 0.6646876335144043,
      "learning_rate": 0.0001,
      "loss": 0.7811,
      "mean_token_accuracy": 0.7774537593126297,
      "num_tokens": 1468469.0,
      "step": 780
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.692534863948822,
      "learning_rate": 0.0001,
      "loss": 0.6728,
      "mean_token_accuracy": 0.8075501427054406,
      "num_tokens": 1487203.0,
      "step": 790
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.673222541809082,
      "learning_rate": 0.0001,
      "loss": 0.6851,
      "mean_token_accuracy": 0.7982731312513351,
      "num_tokens": 1506572.0,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.1678171157836914,
      "eval_mean_token_accuracy": 0.708535186573863,
      "eval_num_tokens": 1506572.0,
      "eval_runtime": 18.3587,
      "eval_samples_per_second": 17.43,
      "eval_steps_per_second": 8.715,
      "step": 800
    }
  ],
  "logging_steps": 10,
  "max_steps": 800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 160,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4794283939332096.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}