{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.7181719260065287,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013601741022850925,
      "grad_norm": 2.812274217605591,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 5.2086,
      "step": 10
    },
    {
      "epoch": 0.02720348204570185,
      "grad_norm": 2.162468433380127,
      "learning_rate": 4.000000000000001e-06,
      "loss": 5.2051,
      "step": 20
    },
    {
      "epoch": 0.040805223068552776,
      "grad_norm": 1.7208396196365356,
      "learning_rate": 6e-06,
      "loss": 5.2105,
      "step": 30
    },
    {
      "epoch": 0.0544069640914037,
      "grad_norm": 2.556692361831665,
      "learning_rate": 8.000000000000001e-06,
      "loss": 5.1625,
      "step": 40
    },
    {
      "epoch": 0.06800870511425462,
      "grad_norm": 8.235758781433105,
      "learning_rate": 1e-05,
      "loss": 4.6379,
      "step": 50
    },
    {
      "epoch": 0.08161044613710555,
      "grad_norm": 10.163893699645996,
      "learning_rate": 1.2e-05,
      "loss": 3.8391,
      "step": 60
    },
    {
      "epoch": 0.09521218715995647,
      "grad_norm": 11.116437911987305,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 3.0016,
      "step": 70
    },
    {
      "epoch": 0.1088139281828074,
      "grad_norm": 10.696512222290039,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.4004,
      "step": 80
    },
    {
      "epoch": 0.12241566920565833,
      "grad_norm": 12.367400169372559,
      "learning_rate": 1.8e-05,
      "loss": 1.8969,
      "step": 90
    },
    {
      "epoch": 0.13601741022850924,
      "grad_norm": 13.723505973815918,
      "learning_rate": 2e-05,
      "loss": 1.4616,
      "step": 100
    },
    {
      "epoch": 0.14961915125136016,
      "grad_norm": 22.034290313720703,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 1.2379,
      "step": 110
    },
    {
      "epoch": 0.1632208922742111,
      "grad_norm": 8.55447769165039,
      "learning_rate": 2.4e-05,
      "loss": 1.0951,
      "step": 120
    },
    {
      "epoch": 0.17682263329706202,
      "grad_norm": 11.120705604553223,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.9654,
      "step": 130
    },
    {
      "epoch": 0.19042437431991294,
      "grad_norm": 13.087006568908691,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.8391,
      "step": 140
    },
    {
      "epoch": 0.20402611534276388,
      "grad_norm": 12.556931495666504,
      "learning_rate": 3e-05,
      "loss": 0.7075,
      "step": 150
    },
    {
      "epoch": 0.2176278563656148,
      "grad_norm": 14.245359420776367,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.6853,
      "step": 160
    },
    {
      "epoch": 0.2312295973884657,
      "grad_norm": 12.10800838470459,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.606,
      "step": 170
    },
    {
      "epoch": 0.24483133841131666,
      "grad_norm": 11.180713653564453,
      "learning_rate": 3.6e-05,
      "loss": 0.5734,
      "step": 180
    },
    {
      "epoch": 0.2584330794341676,
      "grad_norm": 12.638975143432617,
      "learning_rate": 3.8e-05,
      "loss": 0.545,
      "step": 190
    },
    {
      "epoch": 0.2720348204570185,
      "grad_norm": 11.650677680969238,
      "learning_rate": 4e-05,
      "loss": 0.5162,
      "step": 200
    },
    {
      "epoch": 0.28563656147986943,
      "grad_norm": 11.61837387084961,
      "learning_rate": 4.2e-05,
      "loss": 0.491,
      "step": 210
    },
    {
      "epoch": 0.2992383025027203,
      "grad_norm": 5.453939914703369,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.4378,
      "step": 220
    },
    {
      "epoch": 0.31284004352557127,
      "grad_norm": 7.908788204193115,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4299,
      "step": 230
    },
    {
      "epoch": 0.3264417845484222,
      "grad_norm": 13.816142082214355,
      "learning_rate": 4.8e-05,
      "loss": 0.4113,
      "step": 240
    },
    {
      "epoch": 0.3400435255712731,
      "grad_norm": 11.430989265441895,
      "learning_rate": 5e-05,
      "loss": 0.4371,
      "step": 250
    },
    {
      "epoch": 0.35364526659412404,
      "grad_norm": 6.766596794128418,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 0.387,
      "step": 260
    },
    {
      "epoch": 0.367247007616975,
      "grad_norm": 10.817462921142578,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.3643,
      "step": 270
    },
    {
      "epoch": 0.3808487486398259,
      "grad_norm": 6.278716564178467,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 0.3477,
      "step": 280
    },
    {
      "epoch": 0.3944504896626768,
      "grad_norm": 13.244380950927734,
      "learning_rate": 5.8e-05,
      "loss": 0.3763,
      "step": 290
    },
    {
      "epoch": 0.40805223068552776,
      "grad_norm": 4.513805866241455,
      "learning_rate": 6e-05,
      "loss": 0.3093,
      "step": 300
    },
    {
      "epoch": 0.42165397170837865,
      "grad_norm": 4.978238582611084,
      "learning_rate": 6.2e-05,
      "loss": 0.306,
      "step": 310
    },
    {
      "epoch": 0.4352557127312296,
      "grad_norm": 9.654162406921387,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.3157,
      "step": 320
    },
    {
      "epoch": 0.44885745375408054,
      "grad_norm": 5.806159496307373,
      "learning_rate": 6.6e-05,
      "loss": 0.302,
      "step": 330
    },
    {
      "epoch": 0.4624591947769314,
      "grad_norm": 6.283710479736328,
      "learning_rate": 6.800000000000001e-05,
      "loss": 0.2885,
      "step": 340
    },
    {
      "epoch": 0.47606093579978237,
      "grad_norm": 6.1870951652526855,
      "learning_rate": 7e-05,
      "loss": 0.289,
      "step": 350
    },
    {
      "epoch": 0.4896626768226333,
      "grad_norm": 3.082080364227295,
      "learning_rate": 7.2e-05,
      "loss": 0.2584,
      "step": 360
    },
    {
      "epoch": 0.5032644178454843,
      "grad_norm": 5.254792213439941,
      "learning_rate": 7.4e-05,
      "loss": 0.2752,
      "step": 370
    },
    {
      "epoch": 0.5168661588683352,
      "grad_norm": 6.462503433227539,
      "learning_rate": 7.6e-05,
      "loss": 0.2639,
      "step": 380
    },
    {
      "epoch": 0.530467899891186,
      "grad_norm": 10.89343547821045,
      "learning_rate": 7.800000000000001e-05,
      "loss": 0.2893,
      "step": 390
    },
    {
      "epoch": 0.544069640914037,
      "grad_norm": 5.2178192138671875,
      "learning_rate": 8e-05,
      "loss": 0.2415,
      "step": 400
    },
    {
      "epoch": 0.5576713819368879,
      "grad_norm": 8.687518119812012,
      "learning_rate": 8.2e-05,
      "loss": 0.2385,
      "step": 410
    },
    {
      "epoch": 0.5712731229597389,
      "grad_norm": 4.670180320739746,
      "learning_rate": 8.4e-05,
      "loss": 0.2297,
      "step": 420
    },
    {
      "epoch": 0.5848748639825898,
      "grad_norm": 3.699598550796509,
      "learning_rate": 8.6e-05,
      "loss": 0.2241,
      "step": 430
    },
    {
      "epoch": 0.5984766050054406,
      "grad_norm": 3.260232925415039,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.216,
      "step": 440
    },
    {
      "epoch": 0.6120783460282916,
      "grad_norm": 3.4559569358825684,
      "learning_rate": 9e-05,
      "loss": 0.2275,
      "step": 450
    },
    {
      "epoch": 0.6256800870511425,
      "grad_norm": 3.7167000770568848,
      "learning_rate": 9.200000000000001e-05,
      "loss": 0.2133,
      "step": 460
    },
    {
      "epoch": 0.6392818280739935,
      "grad_norm": 4.1776628494262695,
      "learning_rate": 9.4e-05,
      "loss": 0.2194,
      "step": 470
    },
    {
      "epoch": 0.6528835690968444,
      "grad_norm": 4.609129428863525,
      "learning_rate": 9.6e-05,
      "loss": 0.2088,
      "step": 480
    },
    {
      "epoch": 0.6664853101196954,
      "grad_norm": 5.062413215637207,
      "learning_rate": 9.8e-05,
      "loss": 0.2185,
      "step": 490
    },
    {
      "epoch": 0.6800870511425462,
      "grad_norm": 3.3144760131835938,
      "learning_rate": 0.0001,
      "loss": 0.2176,
      "step": 500
    },
    {
      "epoch": 0.6936887921653971,
      "grad_norm": 2.5662150382995605,
      "learning_rate": 0.00010200000000000001,
      "loss": 0.2108,
      "step": 510
    },
    {
      "epoch": 0.7072905331882481,
      "grad_norm": 2.718777656555176,
      "learning_rate": 0.00010400000000000001,
      "loss": 0.1734,
      "step": 520
    },
    {
      "epoch": 0.720892274211099,
      "grad_norm": 2.934107780456543,
      "learning_rate": 0.00010600000000000002,
      "loss": 0.209,
      "step": 530
    },
    {
      "epoch": 0.73449401523395,
      "grad_norm": 2.3642656803131104,
      "learning_rate": 0.00010800000000000001,
      "loss": 0.1958,
      "step": 540
    },
    {
      "epoch": 0.7480957562568009,
      "grad_norm": 2.765012502670288,
      "learning_rate": 0.00011000000000000002,
      "loss": 0.1857,
      "step": 550
    },
    {
      "epoch": 0.7616974972796517,
      "grad_norm": 2.482921600341797,
      "learning_rate": 0.00011200000000000001,
      "loss": 0.186,
      "step": 560
    },
    {
      "epoch": 0.7752992383025027,
      "grad_norm": 2.26837420463562,
      "learning_rate": 0.00011399999999999999,
      "loss": 0.1783,
      "step": 570
    },
    {
      "epoch": 0.7889009793253536,
      "grad_norm": 2.1319069862365723,
      "learning_rate": 0.000116,
      "loss": 0.1731,
      "step": 580
    },
    {
      "epoch": 0.8025027203482046,
      "grad_norm": 1.4668622016906738,
      "learning_rate": 0.000118,
      "loss": 0.1854,
      "step": 590
    },
    {
      "epoch": 0.8161044613710555,
      "grad_norm": 2.2079055309295654,
      "learning_rate": 0.00012,
      "loss": 0.1806,
      "step": 600
    },
    {
      "epoch": 0.8297062023939065,
      "grad_norm": 1.9991375207901,
      "learning_rate": 0.000122,
      "loss": 0.1682,
      "step": 610
    },
    {
      "epoch": 0.8433079434167573,
      "grad_norm": 1.889404296875,
      "learning_rate": 0.000124,
      "loss": 0.162,
      "step": 620
    },
    {
      "epoch": 0.8569096844396082,
      "grad_norm": 1.7021093368530273,
      "learning_rate": 0.000126,
      "loss": 0.1521,
      "step": 630
    },
    {
      "epoch": 0.8705114254624592,
      "grad_norm": 2.153775215148926,
      "learning_rate": 0.00012800000000000002,
      "loss": 0.1656,
      "step": 640
    },
    {
      "epoch": 0.8841131664853101,
      "grad_norm": 2.285336494445801,
      "learning_rate": 0.00013000000000000002,
      "loss": 0.1609,
      "step": 650
    },
    {
      "epoch": 0.8977149075081611,
      "grad_norm": 1.8330312967300415,
      "learning_rate": 0.000132,
      "loss": 0.151,
      "step": 660
    },
    {
      "epoch": 0.911316648531012,
      "grad_norm": 1.712404489517212,
      "learning_rate": 0.000134,
      "loss": 0.1577,
      "step": 670
    },
    {
      "epoch": 0.9249183895538629,
      "grad_norm": 1.6210945844650269,
      "learning_rate": 0.00013600000000000003,
      "loss": 0.1641,
      "step": 680
    },
    {
      "epoch": 0.9385201305767138,
      "grad_norm": 1.661000370979309,
      "learning_rate": 0.000138,
      "loss": 0.1543,
      "step": 690
    },
    {
      "epoch": 0.9521218715995647,
      "grad_norm": 1.754515290260315,
      "learning_rate": 0.00014,
      "loss": 0.1621,
      "step": 700
    },
    {
      "epoch": 0.9657236126224157,
      "grad_norm": 1.6415514945983887,
      "learning_rate": 0.000142,
      "loss": 0.144,
      "step": 710
    },
    {
      "epoch": 0.9793253536452666,
      "grad_norm": 2.2498250007629395,
      "learning_rate": 0.000144,
      "loss": 0.1502,
      "step": 720
    },
    {
      "epoch": 0.9929270946681176,
      "grad_norm": 1.6925517320632935,
      "learning_rate": 0.000146,
      "loss": 0.1546,
      "step": 730
    },
    {
      "epoch": 1.0054406964091405,
      "grad_norm": 1.8807954788208008,
      "learning_rate": 0.000148,
      "loss": 0.1409,
      "step": 740
    },
    {
      "epoch": 1.0190424374319913,
      "grad_norm": 1.611089825630188,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.1578,
      "step": 750
    },
    {
      "epoch": 1.0326441784548421,
      "grad_norm": 1.1505298614501953,
      "learning_rate": 0.000152,
      "loss": 0.1371,
      "step": 760
    },
    {
      "epoch": 1.0462459194776932,
      "grad_norm": 1.5354456901550293,
      "learning_rate": 0.000154,
      "loss": 0.1407,
      "step": 770
    },
    {
      "epoch": 1.059847660500544,
      "grad_norm": 1.178806185722351,
      "learning_rate": 0.00015600000000000002,
      "loss": 0.141,
      "step": 780
    },
    {
      "epoch": 1.073449401523395,
      "grad_norm": 1.3755444288253784,
      "learning_rate": 0.00015800000000000002,
      "loss": 0.1287,
      "step": 790
    },
    {
      "epoch": 1.087051142546246,
      "grad_norm": 1.5800870656967163,
      "learning_rate": 0.00016,
      "loss": 0.1382,
      "step": 800
    },
    {
      "epoch": 1.1006528835690967,
      "grad_norm": 1.1685993671417236,
      "learning_rate": 0.000162,
      "loss": 0.1363,
      "step": 810
    },
    {
      "epoch": 1.1142546245919478,
      "grad_norm": 1.1938755512237549,
      "learning_rate": 0.000164,
      "loss": 0.1316,
      "step": 820
    },
    {
      "epoch": 1.1278563656147986,
      "grad_norm": 1.2022426128387451,
      "learning_rate": 0.000166,
      "loss": 0.1313,
      "step": 830
    },
    {
      "epoch": 1.1414581066376497,
      "grad_norm": 1.1900382041931152,
      "learning_rate": 0.000168,
      "loss": 0.1256,
      "step": 840
    },
    {
      "epoch": 1.1550598476605005,
      "grad_norm": 1.067172884941101,
      "learning_rate": 0.00017,
      "loss": 0.1385,
      "step": 850
    },
    {
      "epoch": 1.1686615886833516,
      "grad_norm": 1.4434224367141724,
      "learning_rate": 0.000172,
      "loss": 0.1382,
      "step": 860
    },
    {
      "epoch": 1.1822633297062024,
      "grad_norm": 1.0884168148040771,
      "learning_rate": 0.000174,
      "loss": 0.1266,
      "step": 870
    },
    {
      "epoch": 1.1958650707290532,
      "grad_norm": 1.3909893035888672,
      "learning_rate": 0.00017600000000000002,
      "loss": 0.1215,
      "step": 880
    },
    {
      "epoch": 1.2094668117519043,
      "grad_norm": 0.9344027042388916,
      "learning_rate": 0.00017800000000000002,
      "loss": 0.119,
      "step": 890
    },
    {
      "epoch": 1.2230685527747551,
      "grad_norm": 1.1694083213806152,
      "learning_rate": 0.00018,
      "loss": 0.1192,
      "step": 900
    },
    {
      "epoch": 1.2366702937976062,
      "grad_norm": 0.9874443411827087,
      "learning_rate": 0.000182,
      "loss": 0.1278,
      "step": 910
    },
    {
      "epoch": 1.250272034820457,
      "grad_norm": 1.2893680334091187,
      "learning_rate": 0.00018400000000000003,
      "loss": 0.1318,
      "step": 920
    },
    {
      "epoch": 1.263873775843308,
      "grad_norm": 1.3346811532974243,
      "learning_rate": 0.00018600000000000002,
      "loss": 0.128,
      "step": 930
    },
    {
      "epoch": 1.277475516866159,
      "grad_norm": 0.9889335632324219,
      "learning_rate": 0.000188,
      "loss": 0.1372,
      "step": 940
    },
    {
      "epoch": 1.2910772578890097,
      "grad_norm": 1.2218222618103027,
      "learning_rate": 0.00019,
      "loss": 0.1307,
      "step": 950
    },
    {
      "epoch": 1.3046789989118608,
      "grad_norm": 0.897546112537384,
      "learning_rate": 0.000192,
      "loss": 0.1217,
      "step": 960
    },
    {
      "epoch": 1.3182807399347116,
      "grad_norm": 0.9673519730567932,
      "learning_rate": 0.000194,
      "loss": 0.1094,
      "step": 970
    },
    {
      "epoch": 1.3318824809575625,
      "grad_norm": 0.9263612031936646,
      "learning_rate": 0.000196,
      "loss": 0.1111,
      "step": 980
    },
    {
      "epoch": 1.3454842219804135,
      "grad_norm": 1.0785088539123535,
      "learning_rate": 0.00019800000000000002,
      "loss": 0.12,
      "step": 990
    },
    {
      "epoch": 1.3590859630032643,
      "grad_norm": 0.8844039440155029,
      "learning_rate": 0.0002,
      "loss": 0.1211,
      "step": 1000
    },
    {
      "epoch": 1.3726877040261154,
      "grad_norm": 0.8651229739189148,
      "learning_rate": 0.0001999999906373993,
      "loss": 0.1291,
      "step": 1010
    },
    {
      "epoch": 1.3862894450489662,
      "grad_norm": 1.0854979753494263,
      "learning_rate": 0.000199999962549599,
      "loss": 0.125,
      "step": 1020
    },
    {
      "epoch": 1.3998911860718173,
      "grad_norm": 1.303252935409546,
      "learning_rate": 0.00019999991573660427,
      "loss": 0.1295,
      "step": 1030
    },
    {
      "epoch": 1.4134929270946681,
      "grad_norm": 1.0601307153701782,
      "learning_rate": 0.00019999985019842397,
      "loss": 0.1363,
      "step": 1040
    },
    {
      "epoch": 1.427094668117519,
      "grad_norm": 0.8127331733703613,
      "learning_rate": 0.0001999997659350703,
      "loss": 0.1124,
      "step": 1050
    },
    {
      "epoch": 1.44069640914037,
      "grad_norm": 0.867559015750885,
      "learning_rate": 0.0001999996629465591,
      "loss": 0.1157,
      "step": 1060
    },
    {
      "epoch": 1.4542981501632208,
      "grad_norm": 0.8963221907615662,
      "learning_rate": 0.0001999995412329096,
      "loss": 0.1197,
      "step": 1070
    },
    {
      "epoch": 1.467899891186072,
      "grad_norm": 0.7211653590202332,
      "learning_rate": 0.00019999940079414464,
      "loss": 0.11,
      "step": 1080
    },
    {
      "epoch": 1.4815016322089227,
      "grad_norm": 0.8746846914291382,
      "learning_rate": 0.00019999924163029048,
      "loss": 0.104,
      "step": 1090
    },
    {
      "epoch": 1.4951033732317738,
      "grad_norm": 0.6718381643295288,
      "learning_rate": 0.00019999906374137694,
      "loss": 0.1041,
      "step": 1100
    },
    {
      "epoch": 1.5087051142546246,
      "grad_norm": 0.7146100997924805,
      "learning_rate": 0.00019999886712743732,
      "loss": 0.0894,
      "step": 1110
    },
    {
      "epoch": 1.5223068552774754,
      "grad_norm": 0.784561276435852,
      "learning_rate": 0.00019999865178850845,
      "loss": 0.0885,
      "step": 1120
    },
    {
      "epoch": 1.5359085963003265,
      "grad_norm": 0.610625147819519,
      "learning_rate": 0.00019999841772463066,
      "loss": 0.0853,
      "step": 1130
    },
    {
      "epoch": 1.5495103373231773,
      "grad_norm": 0.5619096159934998,
      "learning_rate": 0.00019999816493584775,
      "loss": 0.0788,
      "step": 1140
    },
    {
      "epoch": 1.5631120783460282,
      "grad_norm": 0.5502200126647949,
      "learning_rate": 0.00019999789342220708,
      "loss": 0.0802,
      "step": 1150
    },
    {
      "epoch": 1.5767138193688792,
      "grad_norm": 0.6030136942863464,
      "learning_rate": 0.0001999976031837595,
      "loss": 0.083,
      "step": 1160
    },
    {
      "epoch": 1.5903155603917303,
      "grad_norm": 0.647160530090332,
      "learning_rate": 0.00019999729422055928,
      "loss": 0.0897,
      "step": 1170
    },
    {
      "epoch": 1.603917301414581,
      "grad_norm": 0.7512218952178955,
      "learning_rate": 0.00019999696653266437,
      "loss": 0.08,
      "step": 1180
    },
    {
      "epoch": 1.617519042437432,
      "grad_norm": 0.5823985934257507,
      "learning_rate": 0.00019999662012013612,
      "loss": 0.0772,
      "step": 1190
    },
    {
      "epoch": 1.631120783460283,
      "grad_norm": 0.6494550108909607,
      "learning_rate": 0.00019999625498303932,
      "loss": 0.0772,
      "step": 1200
    },
    {
      "epoch": 1.6447225244831338,
      "grad_norm": 0.5662053823471069,
      "learning_rate": 0.00019999587112144244,
      "loss": 0.0695,
      "step": 1210
    },
    {
      "epoch": 1.6583242655059847,
      "grad_norm": 0.4981078803539276,
      "learning_rate": 0.0001999954685354173,
      "loss": 0.0742,
      "step": 1220
    },
    {
      "epoch": 1.6719260065288357,
      "grad_norm": 0.5595643520355225,
      "learning_rate": 0.00019999504722503927,
      "loss": 0.0693,
      "step": 1230
    },
    {
      "epoch": 1.6855277475516868,
      "grad_norm": 0.49906110763549805,
      "learning_rate": 0.00019999460719038732,
      "loss": 0.0692,
      "step": 1240
    },
    {
      "epoch": 1.6991294885745374,
      "grad_norm": 0.5288122296333313,
      "learning_rate": 0.00019999414843154375,
      "loss": 0.0689,
      "step": 1250
    },
    {
      "epoch": 1.7127312295973884,
      "grad_norm": 0.48555830121040344,
      "learning_rate": 0.00019999367094859452,
      "loss": 0.0655,
      "step": 1260
    },
    {
      "epoch": 1.7263329706202395,
      "grad_norm": 0.5002060532569885,
      "learning_rate": 0.00019999317474162905,
      "loss": 0.0707,
      "step": 1270
    },
    {
      "epoch": 1.7399347116430903,
      "grad_norm": 0.4510345458984375,
      "learning_rate": 0.0001999926598107402,
      "loss": 0.06,
      "step": 1280
    },
    {
      "epoch": 1.7535364526659412,
      "grad_norm": 0.5075559020042419,
      "learning_rate": 0.00019999212615602445,
      "loss": 0.0675,
      "step": 1290
    },
    {
      "epoch": 1.7671381936887922,
      "grad_norm": 0.5471305251121521,
      "learning_rate": 0.0001999915737775817,
      "loss": 0.0661,
      "step": 1300
    },
    {
      "epoch": 1.780739934711643,
      "grad_norm": 0.6649473905563354,
      "learning_rate": 0.00019999100267551538,
      "loss": 0.0746,
      "step": 1310
    },
    {
      "epoch": 1.7943416757344939,
      "grad_norm": 0.6705607175827026,
      "learning_rate": 0.00019999041284993245,
      "loss": 0.075,
      "step": 1320
    },
    {
      "epoch": 1.807943416757345,
      "grad_norm": 0.645964503288269,
      "learning_rate": 0.00019998980430094334,
      "loss": 0.0825,
      "step": 1330
    },
    {
      "epoch": 1.821545157780196,
      "grad_norm": 0.48304426670074463,
      "learning_rate": 0.00019998917702866202,
      "loss": 0.0726,
      "step": 1340
    },
    {
      "epoch": 1.8351468988030468,
      "grad_norm": 0.5829260349273682,
      "learning_rate": 0.00019998853103320592,
      "loss": 0.0674,
      "step": 1350
    },
    {
      "epoch": 1.8487486398258977,
      "grad_norm": 0.6563496589660645,
      "learning_rate": 0.00019998786631469603,
      "loss": 0.0666,
      "step": 1360
    },
    {
      "epoch": 1.8623503808487487,
      "grad_norm": 0.5735076069831848,
      "learning_rate": 0.00019998718287325676,
      "loss": 0.0686,
      "step": 1370
    },
    {
      "epoch": 1.8759521218715995,
      "grad_norm": 0.5888078212738037,
      "learning_rate": 0.0001999864807090162,
      "loss": 0.0728,
      "step": 1380
    },
    {
      "epoch": 1.8895538628944504,
      "grad_norm": 0.5383855700492859,
      "learning_rate": 0.00019998575982210572,
      "loss": 0.0691,
      "step": 1390
    },
    {
      "epoch": 1.9031556039173014,
      "grad_norm": 0.4812714159488678,
      "learning_rate": 0.0001999850202126604,
      "loss": 0.0591,
      "step": 1400
    },
    {
      "epoch": 1.9167573449401525,
      "grad_norm": 0.5054184794425964,
      "learning_rate": 0.00019998426188081865,
      "loss": 0.0657,
      "step": 1410
    },
    {
      "epoch": 1.9303590859630033,
      "grad_norm": 0.4280984103679657,
      "learning_rate": 0.0001999834848267225,
      "loss": 0.0619,
      "step": 1420
    },
    {
      "epoch": 1.9439608269858542,
      "grad_norm": 0.7245299816131592,
      "learning_rate": 0.0001999826890505175,
      "loss": 0.064,
      "step": 1430
    },
    {
      "epoch": 1.9575625680087052,
      "grad_norm": 0.5748353004455566,
      "learning_rate": 0.0001999818745523526,
      "loss": 0.0646,
      "step": 1440
    },
    {
      "epoch": 1.971164309031556,
      "grad_norm": 0.5662197470664978,
      "learning_rate": 0.00019998104133238034,
      "loss": 0.0646,
      "step": 1450
    },
    {
      "epoch": 1.9847660500544069,
      "grad_norm": 0.47684717178344727,
      "learning_rate": 0.00019998018939075673,
      "loss": 0.0626,
      "step": 1460
    },
    {
      "epoch": 1.998367791077258,
      "grad_norm": 0.4988132417201996,
      "learning_rate": 0.00019997931872764132,
      "loss": 0.0596,
      "step": 1470
    },
    {
      "epoch": 2.010881392818281,
      "grad_norm": 0.48226118087768555,
      "learning_rate": 0.0001999784293431971,
      "loss": 0.0579,
      "step": 1480
    },
    {
      "epoch": 2.0244831338411315,
      "grad_norm": 0.44725948572158813,
      "learning_rate": 0.0001999775212375907,
      "loss": 0.0575,
      "step": 1490
    },
    {
      "epoch": 2.0380848748639826,
      "grad_norm": 0.5599634051322937,
      "learning_rate": 0.00019997659441099206,
      "loss": 0.0594,
      "step": 1500
    },
    {
      "epoch": 2.0516866158868337,
      "grad_norm": 0.47687241435050964,
      "learning_rate": 0.00019997564886357476,
      "loss": 0.0601,
      "step": 1510
    },
    {
      "epoch": 2.0652883569096843,
      "grad_norm": 0.38273605704307556,
      "learning_rate": 0.0001999746845955159,
      "loss": 0.0551,
      "step": 1520
    },
    {
      "epoch": 2.0788900979325353,
      "grad_norm": 0.5022798180580139,
      "learning_rate": 0.00019997370160699602,
      "loss": 0.0628,
      "step": 1530
    },
    {
      "epoch": 2.0924918389553864,
      "grad_norm": 0.4850836992263794,
      "learning_rate": 0.00019997269989819916,
      "loss": 0.0562,
      "step": 1540
    },
    {
      "epoch": 2.1060935799782374,
      "grad_norm": 0.45668402314186096,
      "learning_rate": 0.0001999716794693129,
      "loss": 0.0542,
      "step": 1550
    },
    {
      "epoch": 2.119695321001088,
      "grad_norm": 0.4036906361579895,
      "learning_rate": 0.00019997064032052837,
      "loss": 0.0556,
      "step": 1560
    },
    {
      "epoch": 2.133297062023939,
      "grad_norm": 0.46593400835990906,
      "learning_rate": 0.00019996958245204009,
      "loss": 0.0563,
      "step": 1570
    },
    {
      "epoch": 2.14689880304679,
      "grad_norm": 0.4298928380012512,
      "learning_rate": 0.00019996850586404615,
      "loss": 0.0531,
      "step": 1580
    },
    {
      "epoch": 2.1605005440696408,
      "grad_norm": 0.40037456154823303,
      "learning_rate": 0.00019996741055674816,
      "loss": 0.0519,
      "step": 1590
    },
    {
      "epoch": 2.174102285092492,
      "grad_norm": 0.4456872344017029,
      "learning_rate": 0.00019996629653035126,
      "loss": 0.0477,
      "step": 1600
    },
    {
      "epoch": 2.187704026115343,
      "grad_norm": 0.5066975355148315,
      "learning_rate": 0.000199965163785064,
      "loss": 0.049,
      "step": 1610
    },
    {
      "epoch": 2.2013057671381935,
      "grad_norm": 0.39652055501937866,
      "learning_rate": 0.0001999640123210985,
      "loss": 0.0485,
      "step": 1620
    },
    {
      "epoch": 2.2149075081610445,
      "grad_norm": 0.41883769631385803,
      "learning_rate": 0.00019996284213867033,
      "loss": 0.051,
      "step": 1630
    },
    {
      "epoch": 2.2285092491838956,
      "grad_norm": 0.4462110996246338,
      "learning_rate": 0.0001999616532379987,
      "loss": 0.0515,
      "step": 1640
    },
    {
      "epoch": 2.2421109902067466,
      "grad_norm": 0.5779175758361816,
      "learning_rate": 0.00019996044561930622,
      "loss": 0.0569,
      "step": 1650
    },
    {
      "epoch": 2.2557127312295973,
      "grad_norm": 0.42402154207229614,
      "learning_rate": 0.00019995921928281894,
      "loss": 0.0519,
      "step": 1660
    },
    {
      "epoch": 2.2693144722524483,
      "grad_norm": 0.642371654510498,
      "learning_rate": 0.00019995797422876654,
      "loss": 0.0612,
      "step": 1670
    },
    {
      "epoch": 2.2829162132752994,
      "grad_norm": 0.44712305068969727,
      "learning_rate": 0.0001999567104573822,
      "loss": 0.0555,
      "step": 1680
    },
    {
      "epoch": 2.29651795429815,
      "grad_norm": 0.5107985138893127,
      "learning_rate": 0.0001999554279689025,
      "loss": 0.0629,
      "step": 1690
    },
    {
      "epoch": 2.310119695321001,
      "grad_norm": 0.4252520203590393,
      "learning_rate": 0.00019995412676356762,
      "loss": 0.0511,
      "step": 1700
    },
    {
      "epoch": 2.323721436343852,
      "grad_norm": 0.413331001996994,
      "learning_rate": 0.0001999528068416212,
      "loss": 0.0504,
      "step": 1710
    },
    {
      "epoch": 2.337323177366703,
      "grad_norm": 0.4216318726539612,
      "learning_rate": 0.0001999514682033104,
      "loss": 0.0539,
      "step": 1720
    },
    {
      "epoch": 2.3509249183895538,
      "grad_norm": 0.5241215825080872,
      "learning_rate": 0.0001999501108488859,
      "loss": 0.0551,
      "step": 1730
    },
    {
      "epoch": 2.364526659412405,
      "grad_norm": 0.48192480206489563,
      "learning_rate": 0.00019994873477860185,
      "loss": 0.0591,
      "step": 1740
    },
    {
      "epoch": 2.378128400435256,
      "grad_norm": 0.5567557215690613,
      "learning_rate": 0.00019994733999271596,
      "loss": 0.0522,
      "step": 1750
    },
    {
      "epoch": 2.3917301414581065,
      "grad_norm": 0.4027807414531708,
      "learning_rate": 0.00019994592649148933,
      "loss": 0.0524,
      "step": 1760
    },
    {
      "epoch": 2.4053318824809575,
      "grad_norm": 0.40923604369163513,
      "learning_rate": 0.0001999444942751867,
      "loss": 0.0533,
      "step": 1770
    },
    {
      "epoch": 2.4189336235038086,
      "grad_norm": 0.37654557824134827,
      "learning_rate": 0.00019994304334407622,
      "loss": 0.0482,
      "step": 1780
    },
    {
      "epoch": 2.432535364526659,
      "grad_norm": 0.37279561161994934,
      "learning_rate": 0.00019994157369842964,
      "loss": 0.0448,
      "step": 1790
    },
    {
      "epoch": 2.4461371055495102,
      "grad_norm": 0.34426528215408325,
      "learning_rate": 0.0001999400853385221,
      "loss": 0.0467,
      "step": 1800
    },
    {
      "epoch": 2.4597388465723613,
      "grad_norm": 0.4583146870136261,
      "learning_rate": 0.00019993857826463231,
      "loss": 0.0501,
      "step": 1810
    },
    {
      "epoch": 2.4733405875952124,
      "grad_norm": 0.4296802878379822,
      "learning_rate": 0.00019993705247704245,
      "loss": 0.0475,
      "step": 1820
    },
    {
      "epoch": 2.486942328618063,
      "grad_norm": 0.3417333960533142,
      "learning_rate": 0.00019993550797603828,
      "loss": 0.0475,
      "step": 1830
    },
    {
      "epoch": 2.500544069640914,
      "grad_norm": 0.3391024172306061,
      "learning_rate": 0.000199933944761909,
      "loss": 0.0428,
      "step": 1840
    },
    {
      "epoch": 2.514145810663765,
      "grad_norm": 0.4173300862312317,
      "learning_rate": 0.00019993236283494728,
      "loss": 0.0487,
      "step": 1850
    },
    {
      "epoch": 2.527747551686616,
      "grad_norm": 0.41140830516815186,
      "learning_rate": 0.00019993076219544938,
      "loss": 0.0499,
      "step": 1860
    },
    {
      "epoch": 2.5413492927094667,
      "grad_norm": 0.4615647494792938,
      "learning_rate": 0.00019992914284371497,
      "loss": 0.0485,
      "step": 1870
    },
    {
      "epoch": 2.554951033732318,
      "grad_norm": 0.4390008747577667,
      "learning_rate": 0.00019992750478004738,
      "loss": 0.0476,
      "step": 1880
    },
    {
      "epoch": 2.568552774755169,
      "grad_norm": 0.36641502380371094,
      "learning_rate": 0.00019992584800475322,
      "loss": 0.0421,
      "step": 1890
    },
    {
      "epoch": 2.5821545157780195,
      "grad_norm": 0.43274542689323425,
      "learning_rate": 0.00019992417251814282,
      "loss": 0.048,
      "step": 1900
    },
    {
      "epoch": 2.5957562568008705,
      "grad_norm": 0.39262470602989197,
      "learning_rate": 0.0001999224783205299,
      "loss": 0.0482,
      "step": 1910
    },
    {
      "epoch": 2.6093579978237216,
      "grad_norm": 0.3618634343147278,
      "learning_rate": 0.0001999207654122316,
      "loss": 0.0464,
      "step": 1920
    },
    {
      "epoch": 2.622959738846572,
      "grad_norm": 0.3481171727180481,
      "learning_rate": 0.0001999190337935688,
      "loss": 0.0478,
      "step": 1930
    },
    {
      "epoch": 2.6365614798694232,
      "grad_norm": 0.33801841735839844,
      "learning_rate": 0.0001999172834648657,
      "loss": 0.0444,
      "step": 1940
    },
    {
      "epoch": 2.6501632208922743,
      "grad_norm": 0.337028443813324,
      "learning_rate": 0.00019991551442645006,
      "loss": 0.0403,
      "step": 1950
    },
    {
      "epoch": 2.663764961915125,
      "grad_norm": 0.3380604684352875,
      "learning_rate": 0.0001999137266786531,
      "loss": 0.0483,
      "step": 1960
    },
    {
      "epoch": 2.677366702937976,
      "grad_norm": 0.44155481457710266,
      "learning_rate": 0.0001999119202218096,
      "loss": 0.045,
      "step": 1970
    },
    {
      "epoch": 2.690968443960827,
      "grad_norm": 0.43442797660827637,
      "learning_rate": 0.00019991009505625784,
      "loss": 0.0442,
      "step": 1980
    },
    {
      "epoch": 2.704570184983678,
      "grad_norm": 0.4352855980396271,
      "learning_rate": 0.00019990825118233957,
      "loss": 0.0463,
      "step": 1990
    },
    {
      "epoch": 2.7181719260065287,
      "grad_norm": 0.5073570609092712,
      "learning_rate": 0.00019990638860040006,
      "loss": 0.0492,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 73600,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.54846097526789e+21,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}