{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0016,
      "grad_norm": 1.146405816078186,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 2.7179,
      "step": 10
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.8417394161224365,
      "learning_rate": 2.857142857142857e-05,
      "loss": 2.5949,
      "step": 20
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.8995881080627441,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 2.332,
      "step": 30
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.7769910097122192,
      "learning_rate": 5.714285714285714e-05,
      "loss": 2.203,
      "step": 40
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.612125039100647,
      "learning_rate": 7.142857142857143e-05,
      "loss": 1.9872,
      "step": 50
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.6728251576423645,
      "learning_rate": 8.571428571428571e-05,
      "loss": 2.0172,
      "step": 60
    },
    {
      "epoch": 0.0112,
      "grad_norm": 0.6434890627861023,
      "learning_rate": 0.0001,
      "loss": 2.0561,
      "step": 70
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.7109495997428894,
      "learning_rate": 9.979716024340772e-05,
      "loss": 1.9218,
      "step": 80
    },
    {
      "epoch": 0.0144,
      "grad_norm": 0.6287526488304138,
      "learning_rate": 9.959432048681541e-05,
      "loss": 1.9663,
      "step": 90
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.761013388633728,
      "learning_rate": 9.939148073022312e-05,
      "loss": 1.8372,
      "step": 100
    },
    {
      "epoch": 0.0176,
      "grad_norm": 0.771938681602478,
      "learning_rate": 9.918864097363083e-05,
      "loss": 1.8148,
      "step": 110
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.7518746256828308,
      "learning_rate": 9.898580121703854e-05,
      "loss": 1.8446,
      "step": 120
    },
    {
      "epoch": 0.0208,
      "grad_norm": 0.7022605538368225,
      "learning_rate": 9.878296146044626e-05,
      "loss": 1.789,
      "step": 130
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.9093043208122253,
      "learning_rate": 9.858012170385395e-05,
      "loss": 1.8224,
      "step": 140
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.8738862872123718,
      "learning_rate": 9.837728194726166e-05,
      "loss": 1.7464,
      "step": 150
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.941779613494873,
      "learning_rate": 9.817444219066937e-05,
      "loss": 1.7306,
      "step": 160
    },
    {
      "epoch": 0.0272,
      "grad_norm": 0.9313611388206482,
      "learning_rate": 9.797160243407709e-05,
      "loss": 1.7978,
      "step": 170
    },
    {
      "epoch": 0.0288,
      "grad_norm": 1.0731656551361084,
      "learning_rate": 9.77687626774848e-05,
      "loss": 1.731,
      "step": 180
    },
    {
      "epoch": 0.0304,
      "grad_norm": 0.8247014880180359,
      "learning_rate": 9.756592292089249e-05,
      "loss": 1.8481,
      "step": 190
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.9673528671264648,
      "learning_rate": 9.73630831643002e-05,
      "loss": 1.7895,
      "step": 200
    },
    {
      "epoch": 0.0336,
      "grad_norm": 0.9702100157737732,
      "learning_rate": 9.716024340770791e-05,
      "loss": 1.7249,
      "step": 210
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.7754144668579102,
      "learning_rate": 9.695740365111563e-05,
      "loss": 1.7188,
      "step": 220
    },
    {
      "epoch": 0.0368,
      "grad_norm": 0.9529830813407898,
      "learning_rate": 9.675456389452334e-05,
      "loss": 1.6541,
      "step": 230
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.926566481590271,
      "learning_rate": 9.655172413793105e-05,
      "loss": 1.6983,
      "step": 240
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8942977786064148,
      "learning_rate": 9.634888438133874e-05,
      "loss": 1.6702,
      "step": 250
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.8158455491065979,
      "learning_rate": 9.614604462474645e-05,
      "loss": 1.6604,
      "step": 260
    },
    {
      "epoch": 0.0432,
      "grad_norm": 0.7894754409790039,
      "learning_rate": 9.594320486815415e-05,
      "loss": 1.6727,
      "step": 270
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.6904724836349487,
      "learning_rate": 9.574036511156188e-05,
      "loss": 1.6826,
      "step": 280
    },
    {
      "epoch": 0.0464,
      "grad_norm": 0.8520828485488892,
      "learning_rate": 9.553752535496959e-05,
      "loss": 1.6916,
      "step": 290
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.9699738025665283,
      "learning_rate": 9.533468559837728e-05,
      "loss": 1.7246,
      "step": 300
    },
    {
      "epoch": 0.0496,
      "grad_norm": 0.8913508057594299,
      "learning_rate": 9.5131845841785e-05,
      "loss": 1.7523,
      "step": 310
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.8721765875816345,
      "learning_rate": 9.49290060851927e-05,
      "loss": 1.7196,
      "step": 320
    },
    {
      "epoch": 0.0528,
      "grad_norm": 0.7151852250099182,
      "learning_rate": 9.47261663286004e-05,
      "loss": 1.7053,
      "step": 330
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.6914452314376831,
      "learning_rate": 9.452332657200811e-05,
      "loss": 1.6875,
      "step": 340
    },
    {
      "epoch": 0.056,
      "grad_norm": 1.2169272899627686,
      "learning_rate": 9.432048681541582e-05,
      "loss": 1.6677,
      "step": 350
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.8138052821159363,
      "learning_rate": 9.411764705882353e-05,
      "loss": 1.6198,
      "step": 360
    },
    {
      "epoch": 0.0592,
      "grad_norm": 0.8007165789604187,
      "learning_rate": 9.391480730223125e-05,
      "loss": 1.6212,
      "step": 370
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.8502349853515625,
      "learning_rate": 9.371196754563894e-05,
      "loss": 1.6224,
      "step": 380
    },
    {
      "epoch": 0.0624,
      "grad_norm": 0.8362089395523071,
      "learning_rate": 9.350912778904665e-05,
      "loss": 1.6112,
      "step": 390
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.9215638041496277,
      "learning_rate": 9.330628803245436e-05,
      "loss": 1.5754,
      "step": 400
    },
    {
      "epoch": 0.0656,
      "grad_norm": 1.6188653707504272,
      "learning_rate": 9.310344827586207e-05,
      "loss": 1.6563,
      "step": 410
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.8163923621177673,
      "learning_rate": 9.290060851926979e-05,
      "loss": 1.661,
      "step": 420
    },
    {
      "epoch": 0.0688,
      "grad_norm": 0.9351512789726257,
      "learning_rate": 9.269776876267748e-05,
      "loss": 1.6473,
      "step": 430
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.8003765940666199,
      "learning_rate": 9.24949290060852e-05,
      "loss": 1.6461,
      "step": 440
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.8877068161964417,
      "learning_rate": 9.22920892494929e-05,
      "loss": 1.6587,
      "step": 450
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.8028745651245117,
      "learning_rate": 9.208924949290061e-05,
      "loss": 1.6517,
      "step": 460
    },
    {
      "epoch": 0.0752,
      "grad_norm": 0.9326333403587341,
      "learning_rate": 9.188640973630833e-05,
      "loss": 1.5998,
      "step": 470
    },
    {
      "epoch": 0.0768,
      "grad_norm": 1.0918183326721191,
      "learning_rate": 9.168356997971604e-05,
      "loss": 1.5878,
      "step": 480
    },
    {
      "epoch": 0.0784,
      "grad_norm": 0.8588346838951111,
      "learning_rate": 9.148073022312373e-05,
      "loss": 1.7285,
      "step": 490
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.67298823595047,
      "learning_rate": 9.127789046653144e-05,
      "loss": 1.6274,
      "step": 500
    },
    {
      "epoch": 0.0816,
      "grad_norm": 0.7446144819259644,
      "learning_rate": 9.107505070993914e-05,
      "loss": 1.6256,
      "step": 510
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.8986618518829346,
      "learning_rate": 9.087221095334687e-05,
      "loss": 1.5674,
      "step": 520
    },
    {
      "epoch": 0.0848,
      "grad_norm": 0.6667838096618652,
      "learning_rate": 9.066937119675458e-05,
      "loss": 1.6515,
      "step": 530
    },
    {
      "epoch": 0.0864,
      "grad_norm": 1.1137733459472656,
      "learning_rate": 9.046653144016227e-05,
      "loss": 1.6212,
      "step": 540
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.7851391434669495,
      "learning_rate": 9.026369168356998e-05,
      "loss": 1.6381,
      "step": 550
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.7694608569145203,
      "learning_rate": 9.00608519269777e-05,
      "loss": 1.4762,
      "step": 560
    },
    {
      "epoch": 0.0912,
      "grad_norm": 0.7087869048118591,
      "learning_rate": 8.985801217038539e-05,
      "loss": 1.5247,
      "step": 570
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.9042320251464844,
      "learning_rate": 8.96551724137931e-05,
      "loss": 1.6496,
      "step": 580
    },
    {
      "epoch": 0.0944,
      "grad_norm": 0.6515716314315796,
      "learning_rate": 8.945233265720081e-05,
      "loss": 1.621,
      "step": 590
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.78260737657547,
      "learning_rate": 8.924949290060852e-05,
      "loss": 1.5592,
      "step": 600
    },
    {
      "epoch": 0.0976,
      "grad_norm": 0.7633717060089111,
      "learning_rate": 8.904665314401624e-05,
      "loss": 1.5134,
      "step": 610
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.731940746307373,
      "learning_rate": 8.884381338742393e-05,
      "loss": 1.6229,
      "step": 620
    },
    {
      "epoch": 0.1008,
      "grad_norm": 0.710590660572052,
      "learning_rate": 8.864097363083164e-05,
      "loss": 1.589,
      "step": 630
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.8609699606895447,
      "learning_rate": 8.843813387423935e-05,
      "loss": 1.5154,
      "step": 640
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.8874902129173279,
      "learning_rate": 8.823529411764706e-05,
      "loss": 1.5012,
      "step": 650
    },
    {
      "epoch": 0.1056,
      "grad_norm": 1.0511506795883179,
      "learning_rate": 8.803245436105478e-05,
      "loss": 1.5922,
      "step": 660
    },
    {
      "epoch": 0.1072,
      "grad_norm": 1.154720425605774,
      "learning_rate": 8.782961460446247e-05,
      "loss": 1.5793,
      "step": 670
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.8440026640892029,
      "learning_rate": 8.762677484787018e-05,
      "loss": 1.5934,
      "step": 680
    },
    {
      "epoch": 0.1104,
      "grad_norm": 0.7727090716362,
      "learning_rate": 8.74239350912779e-05,
      "loss": 1.5149,
      "step": 690
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.8490658402442932,
      "learning_rate": 8.72210953346856e-05,
      "loss": 1.6429,
      "step": 700
    },
    {
      "epoch": 0.1136,
      "grad_norm": 1.0228997468948364,
      "learning_rate": 8.701825557809332e-05,
      "loss": 1.5675,
      "step": 710
    },
    {
      "epoch": 0.1152,
      "grad_norm": 1.192125678062439,
      "learning_rate": 8.681541582150103e-05,
      "loss": 1.4715,
      "step": 720
    },
    {
      "epoch": 0.1168,
      "grad_norm": 0.8283504843711853,
      "learning_rate": 8.661257606490872e-05,
      "loss": 1.5983,
      "step": 730
    },
    {
      "epoch": 0.1184,
      "grad_norm": 1.1739537715911865,
      "learning_rate": 8.640973630831643e-05,
      "loss": 1.4758,
      "step": 740
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6867349743843079,
      "learning_rate": 8.620689655172413e-05,
      "loss": 1.5578,
      "step": 750
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.8097574710845947,
      "learning_rate": 8.600405679513184e-05,
      "loss": 1.6312,
      "step": 760
    },
    {
      "epoch": 0.1232,
      "grad_norm": 0.8128074407577515,
      "learning_rate": 8.580121703853957e-05,
      "loss": 1.5523,
      "step": 770
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.9583007097244263,
      "learning_rate": 8.559837728194726e-05,
      "loss": 1.5231,
      "step": 780
    },
    {
      "epoch": 0.1264,
      "grad_norm": 0.782313883304596,
      "learning_rate": 8.539553752535497e-05,
      "loss": 1.4811,
      "step": 790
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.6528032422065735,
      "learning_rate": 8.519269776876268e-05,
      "loss": 1.4877,
      "step": 800
    },
    {
      "epoch": 0.1296,
      "grad_norm": 0.8802277445793152,
      "learning_rate": 8.498985801217038e-05,
      "loss": 1.6482,
      "step": 810
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.8377550840377808,
      "learning_rate": 8.478701825557809e-05,
      "loss": 1.4813,
      "step": 820
    },
    {
      "epoch": 0.1328,
      "grad_norm": 0.7418074011802673,
      "learning_rate": 8.45841784989858e-05,
      "loss": 1.5555,
      "step": 830
    },
    {
      "epoch": 0.1344,
      "grad_norm": 1.0339664220809937,
      "learning_rate": 8.438133874239351e-05,
      "loss": 1.4302,
      "step": 840
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.8376950025558472,
      "learning_rate": 8.417849898580123e-05,
      "loss": 1.5217,
      "step": 850
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.7994508147239685,
      "learning_rate": 8.397565922920892e-05,
      "loss": 1.5267,
      "step": 860
    },
    {
      "epoch": 0.1392,
      "grad_norm": 0.8002382516860962,
      "learning_rate": 8.377281947261663e-05,
      "loss": 1.5817,
      "step": 870
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.7412007451057434,
      "learning_rate": 8.356997971602434e-05,
      "loss": 1.6293,
      "step": 880
    },
    {
      "epoch": 0.1424,
      "grad_norm": 1.1818407773971558,
      "learning_rate": 8.336713995943205e-05,
      "loss": 1.4277,
      "step": 890
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.8380255699157715,
      "learning_rate": 8.316430020283977e-05,
      "loss": 1.4584,
      "step": 900
    },
    {
      "epoch": 0.1456,
      "grad_norm": 0.7898982167243958,
      "learning_rate": 8.296146044624746e-05,
      "loss": 1.6421,
      "step": 910
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.7636155486106873,
      "learning_rate": 8.275862068965517e-05,
      "loss": 1.5505,
      "step": 920
    },
    {
      "epoch": 0.1488,
      "grad_norm": 0.6694974303245544,
      "learning_rate": 8.255578093306288e-05,
      "loss": 1.5199,
      "step": 930
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.8093082904815674,
      "learning_rate": 8.23529411764706e-05,
      "loss": 1.543,
      "step": 940
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.7055695652961731,
      "learning_rate": 8.21501014198783e-05,
      "loss": 1.5538,
      "step": 950
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.7818977236747742,
      "learning_rate": 8.194726166328602e-05,
      "loss": 1.5042,
      "step": 960
    },
    {
      "epoch": 0.1552,
      "grad_norm": 0.9525614380836487,
      "learning_rate": 8.174442190669371e-05,
      "loss": 1.5782,
      "step": 970
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.8219505548477173,
      "learning_rate": 8.154158215010142e-05,
      "loss": 1.6347,
      "step": 980
    },
    {
      "epoch": 0.1584,
      "grad_norm": 0.8257455825805664,
      "learning_rate": 8.133874239350912e-05,
      "loss": 1.4861,
      "step": 990
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7356287240982056,
      "learning_rate": 8.113590263691683e-05,
      "loss": 1.5001,
      "step": 1000
    },
    {
      "epoch": 0.1616,
      "grad_norm": 0.800564706325531,
      "learning_rate": 8.093306288032456e-05,
      "loss": 1.5157,
      "step": 1010
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.7705591917037964,
      "learning_rate": 8.073022312373225e-05,
      "loss": 1.4415,
      "step": 1020
    },
    {
      "epoch": 0.1648,
      "grad_norm": 0.854865312576294,
      "learning_rate": 8.052738336713996e-05,
      "loss": 1.6235,
      "step": 1030
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.6734772324562073,
      "learning_rate": 8.032454361054767e-05,
      "loss": 1.541,
      "step": 1040
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.8943230509757996,
      "learning_rate": 8.012170385395537e-05,
      "loss": 1.5294,
      "step": 1050
    },
    {
      "epoch": 0.1696,
      "grad_norm": 1.0136455297470093,
      "learning_rate": 7.991886409736308e-05,
      "loss": 1.5255,
      "step": 1060
    },
    {
      "epoch": 0.1712,
      "grad_norm": 0.6639547944068909,
      "learning_rate": 7.97160243407708e-05,
      "loss": 1.5911,
      "step": 1070
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.8482313752174377,
      "learning_rate": 7.95131845841785e-05,
      "loss": 1.4628,
      "step": 1080
    },
    {
      "epoch": 0.1744,
      "grad_norm": 0.941949725151062,
      "learning_rate": 7.931034482758621e-05,
      "loss": 1.538,
      "step": 1090
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.8564821481704712,
      "learning_rate": 7.910750507099391e-05,
      "loss": 1.5709,
      "step": 1100
    },
    {
      "epoch": 0.1776,
      "grad_norm": 0.7663846611976624,
      "learning_rate": 7.890466531440162e-05,
      "loss": 1.4841,
      "step": 1110
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.8091522455215454,
      "learning_rate": 7.870182555780933e-05,
      "loss": 1.5359,
      "step": 1120
    },
    {
      "epoch": 0.1808,
      "grad_norm": 0.8865818381309509,
      "learning_rate": 7.849898580121704e-05,
      "loss": 1.5385,
      "step": 1130
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.6862669587135315,
      "learning_rate": 7.829614604462476e-05,
      "loss": 1.3872,
      "step": 1140
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.84327232837677,
      "learning_rate": 7.809330628803245e-05,
      "loss": 1.3533,
      "step": 1150
    },
    {
      "epoch": 0.1856,
      "grad_norm": 1.1154327392578125,
      "learning_rate": 7.789046653144016e-05,
      "loss": 1.4714,
      "step": 1160
    },
    {
      "epoch": 0.1872,
      "grad_norm": 0.6653951406478882,
      "learning_rate": 7.768762677484787e-05,
      "loss": 1.5765,
      "step": 1170
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.8485225439071655,
      "learning_rate": 7.748478701825558e-05,
      "loss": 1.5679,
      "step": 1180
    },
    {
      "epoch": 0.1904,
      "grad_norm": 0.8555701375007629,
      "learning_rate": 7.72819472616633e-05,
      "loss": 1.5695,
      "step": 1190
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.6411374807357788,
      "learning_rate": 7.7079107505071e-05,
      "loss": 1.4445,
      "step": 1200
    },
    {
      "epoch": 0.1936,
      "grad_norm": 0.7546342611312866,
      "learning_rate": 7.68762677484787e-05,
      "loss": 1.513,
      "step": 1210
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.7529492974281311,
      "learning_rate": 7.667342799188641e-05,
      "loss": 1.5546,
      "step": 1220
    },
    {
      "epoch": 0.1968,
      "grad_norm": 0.8925682902336121,
      "learning_rate": 7.647058823529411e-05,
      "loss": 1.5897,
      "step": 1230
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.8105003833770752,
      "learning_rate": 7.626774847870182e-05,
      "loss": 1.4026,
      "step": 1240
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8478330373764038,
      "learning_rate": 7.606490872210955e-05,
      "loss": 1.5367,
      "step": 1250
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.7117589712142944,
      "learning_rate": 7.586206896551724e-05,
      "loss": 1.5096,
      "step": 1260
    },
    {
      "epoch": 0.2032,
      "grad_norm": 0.6756378412246704,
      "learning_rate": 7.565922920892495e-05,
      "loss": 1.4363,
      "step": 1270
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.8317671418190002,
      "learning_rate": 7.545638945233266e-05,
      "loss": 1.4908,
      "step": 1280
    },
    {
      "epoch": 0.2064,
      "grad_norm": 0.8347654342651367,
      "learning_rate": 7.525354969574036e-05,
      "loss": 1.4809,
      "step": 1290
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.8407132029533386,
      "learning_rate": 7.505070993914807e-05,
      "loss": 1.3965,
      "step": 1300
    },
    {
      "epoch": 0.2096,
      "grad_norm": 0.9293540120124817,
      "learning_rate": 7.484787018255578e-05,
      "loss": 1.5601,
      "step": 1310
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.888976514339447,
      "learning_rate": 7.46450304259635e-05,
      "loss": 1.3249,
      "step": 1320
    },
    {
      "epoch": 0.2128,
      "grad_norm": 1.1678154468536377,
      "learning_rate": 7.44421906693712e-05,
      "loss": 1.5059,
      "step": 1330
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.828194260597229,
      "learning_rate": 7.42393509127789e-05,
      "loss": 1.4796,
      "step": 1340
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.6538078784942627,
      "learning_rate": 7.403651115618661e-05,
      "loss": 1.5022,
      "step": 1350
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.6212038397789001,
      "learning_rate": 7.383367139959432e-05,
      "loss": 1.5556,
      "step": 1360
    },
    {
      "epoch": 0.2192,
      "grad_norm": 0.8193343877792358,
      "learning_rate": 7.363083164300203e-05,
      "loss": 1.4565,
      "step": 1370
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.6584758162498474,
      "learning_rate": 7.342799188640974e-05,
      "loss": 1.549,
      "step": 1380
    },
    {
      "epoch": 0.2224,
      "grad_norm": 0.7993935346603394,
      "learning_rate": 7.322515212981744e-05,
      "loss": 1.6376,
      "step": 1390
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.6824424266815186,
      "learning_rate": 7.302231237322515e-05,
      "loss": 1.5205,
      "step": 1400
    },
    {
      "epoch": 0.2256,
      "grad_norm": 1.070913553237915,
      "learning_rate": 7.281947261663286e-05,
      "loss": 1.4221,
      "step": 1410
    },
    {
      "epoch": 0.2272,
      "grad_norm": 1.1080129146575928,
      "learning_rate": 7.261663286004057e-05,
      "loss": 1.5592,
      "step": 1420
    },
    {
      "epoch": 0.2288,
      "grad_norm": 0.8505323529243469,
      "learning_rate": 7.241379310344828e-05,
      "loss": 1.4511,
      "step": 1430
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.9794740676879883,
      "learning_rate": 7.2210953346856e-05,
      "loss": 1.5121,
      "step": 1440
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.6628521084785461,
      "learning_rate": 7.200811359026369e-05,
      "loss": 1.5905,
      "step": 1450
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.8294353485107422,
      "learning_rate": 7.18052738336714e-05,
      "loss": 1.6115,
      "step": 1460
    },
    {
      "epoch": 0.2352,
      "grad_norm": 1.0685625076293945,
      "learning_rate": 7.16024340770791e-05,
      "loss": 1.4263,
      "step": 1470
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.8424131870269775,
      "learning_rate": 7.139959432048681e-05,
      "loss": 1.5122,
      "step": 1480
    },
    {
      "epoch": 0.2384,
      "grad_norm": 0.9048033356666565,
      "learning_rate": 7.119675456389454e-05,
      "loss": 1.4798,
      "step": 1490
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.7316602468490601,
      "learning_rate": 7.099391480730223e-05,
      "loss": 1.4046,
      "step": 1500
    },
    {
      "epoch": 0.2416,
      "grad_norm": 0.650810956954956,
      "learning_rate": 7.079107505070994e-05,
      "loss": 1.6366,
      "step": 1510
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.6859937310218811,
      "learning_rate": 7.058823529411765e-05,
      "loss": 1.5001,
      "step": 1520
    },
    {
      "epoch": 0.2448,
      "grad_norm": 0.7956719398498535,
      "learning_rate": 7.038539553752535e-05,
      "loss": 1.4611,
      "step": 1530
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.8426589965820312,
      "learning_rate": 7.018255578093306e-05,
      "loss": 1.3004,
      "step": 1540
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.8136724829673767,
      "learning_rate": 6.997971602434077e-05,
      "loss": 1.4873,
      "step": 1550
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.8124368786811829,
      "learning_rate": 6.977687626774848e-05,
      "loss": 1.5005,
      "step": 1560
    },
    {
      "epoch": 0.2512,
      "grad_norm": 0.8657099604606628,
      "learning_rate": 6.95740365111562e-05,
      "loss": 1.5456,
      "step": 1570
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.8967484831809998,
      "learning_rate": 6.937119675456389e-05,
      "loss": 1.4589,
      "step": 1580
    },
    {
      "epoch": 0.2544,
      "grad_norm": 0.7720628976821899,
      "learning_rate": 6.91683569979716e-05,
      "loss": 1.4424,
      "step": 1590
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.8142846822738647,
      "learning_rate": 6.896551724137931e-05,
      "loss": 1.4176,
      "step": 1600
    },
    {
      "epoch": 0.2576,
      "grad_norm": 0.723961353302002,
      "learning_rate": 6.876267748478702e-05,
      "loss": 1.4433,
      "step": 1610
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.8454954028129578,
      "learning_rate": 6.855983772819473e-05,
      "loss": 1.579,
      "step": 1620
    },
    {
      "epoch": 0.2608,
      "grad_norm": 0.8000863194465637,
      "learning_rate": 6.835699797160243e-05,
      "loss": 1.3919,
      "step": 1630
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.8179385662078857,
      "learning_rate": 6.815415821501014e-05,
      "loss": 1.4501,
      "step": 1640
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.9470431208610535,
      "learning_rate": 6.795131845841785e-05,
      "loss": 1.3945,
      "step": 1650
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.7068872451782227,
      "learning_rate": 6.774847870182556e-05,
      "loss": 1.4742,
      "step": 1660
    },
    {
      "epoch": 0.2672,
      "grad_norm": 0.7054203748703003,
      "learning_rate": 6.754563894523327e-05,
      "loss": 1.4438,
      "step": 1670
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.7306079864501953,
      "learning_rate": 6.734279918864099e-05,
      "loss": 1.4841,
      "step": 1680
    },
    {
      "epoch": 0.2704,
      "grad_norm": 0.9410521388053894,
      "learning_rate": 6.713995943204868e-05,
      "loss": 1.4806,
      "step": 1690
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.8054413199424744,
      "learning_rate": 6.69371196754564e-05,
      "loss": 1.5897,
      "step": 1700
    },
    {
      "epoch": 0.2736,
      "grad_norm": 0.8597100973129272,
      "learning_rate": 6.673427991886409e-05,
      "loss": 1.5797,
      "step": 1710
    },
    {
      "epoch": 0.2752,
      "grad_norm": 1.0193672180175781,
      "learning_rate": 6.65314401622718e-05,
      "loss": 1.4948,
      "step": 1720
    },
    {
      "epoch": 0.2768,
      "grad_norm": 0.6713398694992065,
      "learning_rate": 6.632860040567953e-05,
      "loss": 1.4876,
      "step": 1730
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.7806057333946228,
      "learning_rate": 6.612576064908722e-05,
      "loss": 1.3604,
      "step": 1740
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.7411827445030212,
      "learning_rate": 6.592292089249493e-05,
      "loss": 1.4223,
      "step": 1750
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.747585654258728,
      "learning_rate": 6.572008113590264e-05,
      "loss": 1.4759,
      "step": 1760
    },
    {
      "epoch": 0.2832,
      "grad_norm": 0.655139148235321,
      "learning_rate": 6.551724137931034e-05,
      "loss": 1.4646,
      "step": 1770
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.851161003112793,
      "learning_rate": 6.531440162271805e-05,
      "loss": 1.524,
      "step": 1780
    },
    {
      "epoch": 0.2864,
      "grad_norm": 0.8720593452453613,
      "learning_rate": 6.511156186612576e-05,
      "loss": 1.47,
      "step": 1790
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.7257354855537415,
      "learning_rate": 6.490872210953347e-05,
      "loss": 1.4698,
      "step": 1800
    },
    {
      "epoch": 0.2896,
      "grad_norm": 0.8432925939559937,
      "learning_rate": 6.470588235294118e-05,
      "loss": 1.5128,
      "step": 1810
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.696212649345398,
      "learning_rate": 6.450304259634888e-05,
      "loss": 1.4325,
      "step": 1820
    },
    {
      "epoch": 0.2928,
      "grad_norm": 0.709725558757782,
      "learning_rate": 6.430020283975659e-05,
      "loss": 1.4612,
      "step": 1830
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.8701031804084778,
      "learning_rate": 6.40973630831643e-05,
      "loss": 1.5741,
      "step": 1840
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.7509334087371826,
      "learning_rate": 6.389452332657201e-05,
      "loss": 1.3882,
      "step": 1850
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.8448138236999512,
      "learning_rate": 6.369168356997972e-05,
      "loss": 1.4554,
      "step": 1860
    },
    {
      "epoch": 0.2992,
      "grad_norm": 0.829753041267395,
      "learning_rate": 6.348884381338742e-05,
      "loss": 1.517,
      "step": 1870
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.8275175094604492,
      "learning_rate": 6.328600405679513e-05,
      "loss": 1.5145,
      "step": 1880
    },
    {
      "epoch": 0.3024,
      "grad_norm": 0.7375383377075195,
      "learning_rate": 6.308316430020284e-05,
      "loss": 1.4347,
      "step": 1890
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.7672603130340576,
      "learning_rate": 6.288032454361054e-05,
      "loss": 1.3533,
      "step": 1900
    },
    {
      "epoch": 0.3056,
      "grad_norm": 0.9198083281517029,
      "learning_rate": 6.267748478701826e-05,
      "loss": 1.5153,
      "step": 1910
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.9903436303138733,
      "learning_rate": 6.247464503042598e-05,
      "loss": 1.4345,
      "step": 1920
    },
    {
      "epoch": 0.3088,
      "grad_norm": 1.1620358228683472,
      "learning_rate": 6.227180527383367e-05,
      "loss": 1.5443,
      "step": 1930
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.8703617453575134,
      "learning_rate": 6.206896551724138e-05,
      "loss": 1.3112,
      "step": 1940
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.771286129951477,
      "learning_rate": 6.186612576064908e-05,
      "loss": 1.4153,
      "step": 1950
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.8943174481391907,
      "learning_rate": 6.166328600405679e-05,
      "loss": 1.5657,
      "step": 1960
    },
    {
      "epoch": 0.3152,
      "grad_norm": 0.7589307427406311,
      "learning_rate": 6.146044624746452e-05,
      "loss": 1.4126,
      "step": 1970
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.6583797335624695,
      "learning_rate": 6.125760649087221e-05,
      "loss": 1.4217,
      "step": 1980
    },
    {
      "epoch": 0.3184,
      "grad_norm": 0.8261314630508423,
      "learning_rate": 6.105476673427992e-05,
      "loss": 1.5038,
      "step": 1990
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.768459677696228,
      "learning_rate": 6.0851926977687634e-05,
      "loss": 1.5839,
      "step": 2000
    },
    {
      "epoch": 0.3216,
      "grad_norm": 0.7238166332244873,
      "learning_rate": 6.064908722109534e-05,
      "loss": 1.5098,
      "step": 2010
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.847366988658905,
      "learning_rate": 6.044624746450305e-05,
      "loss": 1.3801,
      "step": 2020
    },
    {
      "epoch": 0.3248,
      "grad_norm": 0.8706203103065491,
      "learning_rate": 6.0243407707910746e-05,
      "loss": 1.4568,
      "step": 2030
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.6880550980567932,
      "learning_rate": 6.0040567951318463e-05,
      "loss": 1.4095,
      "step": 2040
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.7416784763336182,
      "learning_rate": 5.9837728194726174e-05,
      "loss": 1.4224,
      "step": 2050
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.778136134147644,
      "learning_rate": 5.963488843813387e-05,
      "loss": 1.3978,
      "step": 2060
    },
    {
      "epoch": 0.3312,
      "grad_norm": 1.0044938325881958,
      "learning_rate": 5.943204868154159e-05,
      "loss": 1.2894,
      "step": 2070
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.9396066069602966,
      "learning_rate": 5.92292089249493e-05,
      "loss": 1.4427,
      "step": 2080
    },
    {
      "epoch": 0.3344,
      "grad_norm": 0.7602688670158386,
      "learning_rate": 5.9026369168357e-05,
      "loss": 1.5022,
      "step": 2090
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.7073438763618469,
      "learning_rate": 5.882352941176471e-05,
      "loss": 1.4401,
      "step": 2100
    },
    {
      "epoch": 0.3376,
      "grad_norm": 0.8761054873466492,
      "learning_rate": 5.862068965517241e-05,
      "loss": 1.5012,
      "step": 2110
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.7597482800483704,
      "learning_rate": 5.841784989858012e-05,
      "loss": 1.5398,
      "step": 2120
    },
    {
      "epoch": 0.3408,
      "grad_norm": 0.7038214206695557,
      "learning_rate": 5.821501014198783e-05,
      "loss": 1.5085,
      "step": 2130
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.7399268746376038,
      "learning_rate": 5.801217038539554e-05,
      "loss": 1.5109,
      "step": 2140
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.6264271140098572,
      "learning_rate": 5.780933062880325e-05,
      "loss": 1.4252,
      "step": 2150
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.6571562886238098,
      "learning_rate": 5.760649087221096e-05,
      "loss": 1.4404,
      "step": 2160
    },
    {
      "epoch": 0.3472,
      "grad_norm": 0.7995081543922424,
      "learning_rate": 5.740365111561866e-05,
      "loss": 1.3708,
      "step": 2170
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.7455433011054993,
      "learning_rate": 5.720081135902637e-05,
      "loss": 1.3199,
      "step": 2180
    },
    {
      "epoch": 0.3504,
      "grad_norm": 0.8443031907081604,
      "learning_rate": 5.699797160243408e-05,
      "loss": 1.3461,
      "step": 2190
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.7456616759300232,
      "learning_rate": 5.679513184584179e-05,
      "loss": 1.3873,
      "step": 2200
    },
    {
      "epoch": 0.3536,
      "grad_norm": 0.8237265348434448,
      "learning_rate": 5.65922920892495e-05,
      "loss": 1.4215,
      "step": 2210
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.6858148574829102,
      "learning_rate": 5.63894523326572e-05,
      "loss": 1.453,
      "step": 2220
    },
    {
      "epoch": 0.3568,
      "grad_norm": 0.8578169941902161,
      "learning_rate": 5.618661257606491e-05,
      "loss": 1.4299,
      "step": 2230
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.7621243000030518,
      "learning_rate": 5.5983772819472624e-05,
      "loss": 1.5106,
      "step": 2240
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.9632871150970459,
      "learning_rate": 5.578093306288033e-05,
      "loss": 1.399,
      "step": 2250
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.8085991144180298,
      "learning_rate": 5.557809330628804e-05,
      "loss": 1.3633,
      "step": 2260
    },
    {
      "epoch": 0.3632,
      "grad_norm": 0.6866801977157593,
      "learning_rate": 5.5375253549695736e-05,
      "loss": 1.3779,
      "step": 2270
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.8746480345726013,
      "learning_rate": 5.517241379310345e-05,
      "loss": 1.4511,
      "step": 2280
    },
    {
      "epoch": 0.3664,
      "grad_norm": 0.7714369893074036,
      "learning_rate": 5.4969574036511164e-05,
      "loss": 1.4124,
      "step": 2290
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.9127957224845886,
      "learning_rate": 5.476673427991886e-05,
      "loss": 1.3662,
      "step": 2300
    },
    {
      "epoch": 0.3696,
      "grad_norm": 0.8726272583007812,
      "learning_rate": 5.456389452332657e-05,
      "loss": 1.4015,
      "step": 2310
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.762836217880249,
      "learning_rate": 5.436105476673429e-05,
      "loss": 1.4259,
      "step": 2320
    },
    {
      "epoch": 0.3728,
      "grad_norm": 0.8870751261711121,
      "learning_rate": 5.4158215010141987e-05,
      "loss": 1.3831,
      "step": 2330
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.8370562195777893,
      "learning_rate": 5.39553752535497e-05,
      "loss": 1.4319,
      "step": 2340
    },
    {
      "epoch": 0.376,
      "grad_norm": 1.0303740501403809,
      "learning_rate": 5.37525354969574e-05,
      "loss": 1.4275,
      "step": 2350
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.5940431952476501,
      "learning_rate": 5.354969574036511e-05,
      "loss": 1.3863,
      "step": 2360
    },
    {
      "epoch": 0.3792,
      "grad_norm": 0.7848506569862366,
      "learning_rate": 5.334685598377282e-05,
      "loss": 1.5844,
      "step": 2370
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.6541472673416138,
      "learning_rate": 5.314401622718053e-05,
      "loss": 1.3225,
      "step": 2380
    },
    {
      "epoch": 0.3824,
      "grad_norm": 0.8245867490768433,
      "learning_rate": 5.294117647058824e-05,
      "loss": 1.4911,
      "step": 2390
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.6516074538230896,
      "learning_rate": 5.273833671399595e-05,
      "loss": 1.4401,
      "step": 2400
    },
    {
      "epoch": 0.3856,
      "grad_norm": 0.9805840849876404,
      "learning_rate": 5.253549695740365e-05,
      "loss": 1.4857,
      "step": 2410
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.7679899334907532,
      "learning_rate": 5.233265720081136e-05,
      "loss": 1.3695,
      "step": 2420
    },
    {
      "epoch": 0.3888,
      "grad_norm": 0.8954936861991882,
      "learning_rate": 5.212981744421907e-05,
      "loss": 1.3894,
      "step": 2430
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.7677492499351501,
      "learning_rate": 5.192697768762678e-05,
      "loss": 1.4339,
      "step": 2440
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.7712604403495789,
      "learning_rate": 5.172413793103449e-05,
      "loss": 1.3207,
      "step": 2450
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.7585605382919312,
      "learning_rate": 5.152129817444219e-05,
      "loss": 1.3693,
      "step": 2460
    },
    {
      "epoch": 0.3952,
      "grad_norm": 1.1001477241516113,
      "learning_rate": 5.13184584178499e-05,
      "loss": 1.3812,
      "step": 2470
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.8667353987693787,
      "learning_rate": 5.1115618661257614e-05,
      "loss": 1.4472,
      "step": 2480
    },
    {
      "epoch": 0.3984,
      "grad_norm": 0.7773622870445251,
      "learning_rate": 5.091277890466532e-05,
      "loss": 1.3703,
      "step": 2490
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.9364688992500305,
      "learning_rate": 5.070993914807303e-05,
      "loss": 1.3569,
      "step": 2500
    },
    {
      "epoch": 0.4016,
      "grad_norm": 0.6922894716262817,
      "learning_rate": 5.0507099391480726e-05,
      "loss": 1.3934,
      "step": 2510
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.6994293332099915,
      "learning_rate": 5.030425963488844e-05,
      "loss": 1.3941,
      "step": 2520
    },
    {
      "epoch": 0.4048,
      "grad_norm": 0.703568160533905,
      "learning_rate": 5.0101419878296154e-05,
      "loss": 1.3753,
      "step": 2530
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.7280287146568298,
      "learning_rate": 4.989858012170386e-05,
      "loss": 1.4547,
      "step": 2540
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.769101083278656,
      "learning_rate": 4.969574036511156e-05,
      "loss": 1.4568,
      "step": 2550
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.8144255876541138,
      "learning_rate": 4.949290060851927e-05,
      "loss": 1.4932,
      "step": 2560
    },
    {
      "epoch": 0.4112,
      "grad_norm": 0.7549088001251221,
      "learning_rate": 4.9290060851926976e-05,
      "loss": 1.4626,
      "step": 2570
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.8042471408843994,
      "learning_rate": 4.908722109533469e-05,
      "loss": 1.3664,
      "step": 2580
    },
    {
      "epoch": 0.4144,
      "grad_norm": 0.8683868646621704,
      "learning_rate": 4.88843813387424e-05,
      "loss": 1.3343,
      "step": 2590
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.8438317179679871,
      "learning_rate": 4.86815415821501e-05,
      "loss": 1.4551,
      "step": 2600
    },
    {
      "epoch": 0.4176,
      "grad_norm": 0.913829505443573,
      "learning_rate": 4.847870182555781e-05,
      "loss": 1.4605,
      "step": 2610
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.7295091152191162,
      "learning_rate": 4.827586206896552e-05,
      "loss": 1.4109,
      "step": 2620
    },
    {
      "epoch": 0.4208,
      "grad_norm": 0.8623864650726318,
      "learning_rate": 4.807302231237323e-05,
      "loss": 1.295,
      "step": 2630
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.6889348030090332,
      "learning_rate": 4.787018255578094e-05,
      "loss": 1.3165,
      "step": 2640
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.9166956543922424,
      "learning_rate": 4.766734279918864e-05,
      "loss": 1.4605,
      "step": 2650
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.7856289744377136,
      "learning_rate": 4.746450304259635e-05,
      "loss": 1.4245,
      "step": 2660
    },
    {
      "epoch": 0.4272,
      "grad_norm": 0.8294436931610107,
      "learning_rate": 4.726166328600406e-05,
      "loss": 1.4116,
      "step": 2670
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.8435599207878113,
      "learning_rate": 4.705882352941177e-05,
      "loss": 1.4129,
      "step": 2680
    },
    {
      "epoch": 0.4304,
      "grad_norm": 0.6988587975502014,
      "learning_rate": 4.685598377281947e-05,
      "loss": 1.4092,
      "step": 2690
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.7235763669013977,
      "learning_rate": 4.665314401622718e-05,
      "loss": 1.3857,
      "step": 2700
    },
    {
      "epoch": 0.4336,
      "grad_norm": 0.8404995799064636,
      "learning_rate": 4.645030425963489e-05,
      "loss": 1.3666,
      "step": 2710
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.8587724566459656,
      "learning_rate": 4.62474645030426e-05,
      "loss": 1.3673,
      "step": 2720
    },
    {
      "epoch": 0.4368,
      "grad_norm": 0.7824047803878784,
      "learning_rate": 4.604462474645031e-05,
      "loss": 1.4784,
      "step": 2730
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.9805149435997009,
      "learning_rate": 4.584178498985802e-05,
      "loss": 1.3985,
      "step": 2740
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.7506240606307983,
      "learning_rate": 4.563894523326572e-05,
      "loss": 1.399,
      "step": 2750
    },
    {
      "epoch": 0.4416,
      "grad_norm": 1.1338064670562744,
      "learning_rate": 4.543610547667343e-05,
      "loss": 1.3585,
      "step": 2760
    },
    {
      "epoch": 0.4432,
      "grad_norm": 0.8251412510871887,
      "learning_rate": 4.523326572008114e-05,
      "loss": 1.5158,
      "step": 2770
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.8259724974632263,
      "learning_rate": 4.503042596348885e-05,
      "loss": 1.4496,
      "step": 2780
    },
    {
      "epoch": 0.4464,
      "grad_norm": 0.9131335020065308,
      "learning_rate": 4.482758620689655e-05,
      "loss": 1.2838,
      "step": 2790
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.8390583992004395,
      "learning_rate": 4.462474645030426e-05,
      "loss": 1.3406,
      "step": 2800
    },
    {
      "epoch": 0.4496,
      "grad_norm": 0.9235684275627136,
      "learning_rate": 4.4421906693711966e-05,
      "loss": 1.359,
      "step": 2810
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.8435768485069275,
      "learning_rate": 4.421906693711968e-05,
      "loss": 1.3435,
      "step": 2820
    },
    {
      "epoch": 0.4528,
      "grad_norm": 0.6606183648109436,
      "learning_rate": 4.401622718052739e-05,
      "loss": 1.3997,
      "step": 2830
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.7467994093894958,
      "learning_rate": 4.381338742393509e-05,
      "loss": 1.4417,
      "step": 2840
    },
    {
      "epoch": 0.456,
      "grad_norm": 1.224295973777771,
      "learning_rate": 4.36105476673428e-05,
      "loss": 1.5001,
      "step": 2850
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.8111449480056763,
      "learning_rate": 4.340770791075051e-05,
      "loss": 1.3596,
      "step": 2860
    },
    {
      "epoch": 0.4592,
      "grad_norm": 0.8036226630210876,
      "learning_rate": 4.320486815415822e-05,
      "loss": 1.4546,
      "step": 2870
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.816628634929657,
      "learning_rate": 4.300202839756592e-05,
      "loss": 1.398,
      "step": 2880
    },
    {
      "epoch": 0.4624,
      "grad_norm": 0.939947247505188,
      "learning_rate": 4.279918864097363e-05,
      "loss": 1.4399,
      "step": 2890
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.8091351985931396,
      "learning_rate": 4.259634888438134e-05,
      "loss": 1.4296,
      "step": 2900
    },
    {
      "epoch": 0.4656,
      "grad_norm": 0.9602811336517334,
      "learning_rate": 4.2393509127789046e-05,
      "loss": 1.4441,
      "step": 2910
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.7727035284042358,
      "learning_rate": 4.219066937119676e-05,
      "loss": 1.4771,
      "step": 2920
    },
    {
      "epoch": 0.4688,
      "grad_norm": 0.7776862978935242,
      "learning_rate": 4.198782961460446e-05,
      "loss": 1.4367,
      "step": 2930
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.9107199907302856,
      "learning_rate": 4.178498985801217e-05,
      "loss": 1.285,
      "step": 2940
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.7945104837417603,
      "learning_rate": 4.158215010141988e-05,
      "loss": 1.435,
      "step": 2950
    },
    {
      "epoch": 0.4736,
      "grad_norm": 1.391619324684143,
      "learning_rate": 4.1379310344827587e-05,
      "loss": 1.524,
      "step": 2960
    },
    {
      "epoch": 0.4752,
      "grad_norm": 0.8388264179229736,
      "learning_rate": 4.11764705882353e-05,
      "loss": 1.4508,
      "step": 2970
    },
    {
      "epoch": 0.4768,
      "grad_norm": 1.0040127038955688,
      "learning_rate": 4.097363083164301e-05,
      "loss": 1.3741,
      "step": 2980
    },
    {
      "epoch": 0.4784,
      "grad_norm": 0.8224083185195923,
      "learning_rate": 4.077079107505071e-05,
      "loss": 1.5105,
      "step": 2990
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8134840726852417,
      "learning_rate": 4.0567951318458416e-05,
      "loss": 1.3409,
      "step": 3000
    },
    {
      "epoch": 0.4816,
      "grad_norm": 0.7531696557998657,
      "learning_rate": 4.036511156186613e-05,
      "loss": 1.4369,
      "step": 3010
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.8255640268325806,
      "learning_rate": 4.016227180527384e-05,
      "loss": 1.4782,
      "step": 3020
    },
    {
      "epoch": 0.4848,
      "grad_norm": 0.8592851161956787,
      "learning_rate": 3.995943204868154e-05,
      "loss": 1.4689,
      "step": 3030
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.839590847492218,
      "learning_rate": 3.975659229208925e-05,
      "loss": 1.4,
      "step": 3040
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.8546445965766907,
      "learning_rate": 3.9553752535496956e-05,
      "loss": 1.4078,
      "step": 3050
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.8112866878509521,
      "learning_rate": 3.935091277890467e-05,
      "loss": 1.4288,
      "step": 3060
    },
    {
      "epoch": 0.4912,
      "grad_norm": 1.23469877243042,
      "learning_rate": 3.914807302231238e-05,
      "loss": 1.4295,
      "step": 3070
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.7116204500198364,
      "learning_rate": 3.894523326572008e-05,
      "loss": 1.3944,
      "step": 3080
    },
    {
      "epoch": 0.4944,
      "grad_norm": 0.8031020760536194,
      "learning_rate": 3.874239350912779e-05,
      "loss": 1.3761,
      "step": 3090
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.7796714901924133,
      "learning_rate": 3.85395537525355e-05,
      "loss": 1.4383,
      "step": 3100
    },
    {
      "epoch": 0.4976,
      "grad_norm": 0.6973742842674255,
      "learning_rate": 3.833671399594321e-05,
      "loss": 1.3507,
      "step": 3110
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.9106045961380005,
      "learning_rate": 3.813387423935091e-05,
      "loss": 1.3073,
      "step": 3120
    },
    {
      "epoch": 0.5008,
      "grad_norm": 0.8738226294517517,
      "learning_rate": 3.793103448275862e-05,
      "loss": 1.435,
      "step": 3130
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.8685261607170105,
      "learning_rate": 3.772819472616633e-05,
      "loss": 1.3957,
      "step": 3140
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.7489401698112488,
      "learning_rate": 3.7525354969574036e-05,
      "loss": 1.3978,
      "step": 3150
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.931246280670166,
      "learning_rate": 3.732251521298175e-05,
      "loss": 1.4185,
      "step": 3160
    },
    {
      "epoch": 0.5072,
      "grad_norm": 0.6278898119926453,
      "learning_rate": 3.711967545638945e-05,
      "loss": 1.3093,
      "step": 3170
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.7450694441795349,
      "learning_rate": 3.691683569979716e-05,
      "loss": 1.2888,
      "step": 3180
    },
    {
      "epoch": 0.5104,
      "grad_norm": 0.9010635614395142,
      "learning_rate": 3.671399594320487e-05,
      "loss": 1.4189,
      "step": 3190
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.7903779745101929,
      "learning_rate": 3.6511156186612576e-05,
      "loss": 1.3984,
      "step": 3200
    },
    {
      "epoch": 0.5136,
      "grad_norm": 0.8521105647087097,
      "learning_rate": 3.630831643002029e-05,
      "loss": 1.372,
      "step": 3210
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.832203209400177,
      "learning_rate": 3.6105476673428e-05,
      "loss": 1.3513,
      "step": 3220
    },
    {
      "epoch": 0.5168,
      "grad_norm": 0.6941149830818176,
      "learning_rate": 3.59026369168357e-05,
      "loss": 1.4557,
      "step": 3230
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.7686852812767029,
      "learning_rate": 3.5699797160243406e-05,
      "loss": 1.3021,
      "step": 3240
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.823742687702179,
      "learning_rate": 3.5496957403651116e-05,
      "loss": 1.4165,
      "step": 3250
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.7812134623527527,
      "learning_rate": 3.529411764705883e-05,
      "loss": 1.452,
      "step": 3260
    },
    {
      "epoch": 0.5232,
      "grad_norm": 0.9503807425498962,
      "learning_rate": 3.509127789046653e-05,
      "loss": 1.3479,
      "step": 3270
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.7697770595550537,
      "learning_rate": 3.488843813387424e-05,
      "loss": 1.3111,
      "step": 3280
    },
    {
      "epoch": 0.5264,
      "grad_norm": 1.1085420846939087,
      "learning_rate": 3.4685598377281946e-05,
      "loss": 1.2394,
      "step": 3290
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.9299151301383972,
      "learning_rate": 3.4482758620689657e-05,
      "loss": 1.2936,
      "step": 3300
    },
    {
      "epoch": 0.5296,
      "grad_norm": 0.7949966192245483,
      "learning_rate": 3.427991886409737e-05,
      "loss": 1.4256,
      "step": 3310
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.8013599514961243,
      "learning_rate": 3.407707910750507e-05,
      "loss": 1.4491,
      "step": 3320
    },
    {
      "epoch": 0.5328,
      "grad_norm": 1.1205172538757324,
      "learning_rate": 3.387423935091278e-05,
      "loss": 1.2401,
      "step": 3330
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.7110530734062195,
      "learning_rate": 3.367139959432049e-05,
      "loss": 1.4416,
      "step": 3340
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.8773007988929749,
      "learning_rate": 3.34685598377282e-05,
      "loss": 1.3766,
      "step": 3350
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.9566447734832764,
      "learning_rate": 3.32657200811359e-05,
      "loss": 1.3417,
      "step": 3360
    },
    {
      "epoch": 0.5392,
      "grad_norm": 0.8582170605659485,
      "learning_rate": 3.306288032454361e-05,
      "loss": 1.3405,
      "step": 3370
    },
    {
      "epoch": 0.5408,
      "grad_norm": 1.08076810836792,
      "learning_rate": 3.286004056795132e-05,
      "loss": 1.3765,
      "step": 3380
    },
    {
      "epoch": 0.5424,
      "grad_norm": 0.8801867961883545,
      "learning_rate": 3.2657200811359026e-05,
      "loss": 1.4233,
      "step": 3390
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.9401204586029053,
      "learning_rate": 3.245436105476674e-05,
      "loss": 1.2817,
      "step": 3400
    },
    {
      "epoch": 0.5456,
      "grad_norm": 0.7613691091537476,
      "learning_rate": 3.225152129817444e-05,
      "loss": 1.3568,
      "step": 3410
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.7864488959312439,
      "learning_rate": 3.204868154158215e-05,
      "loss": 1.3435,
      "step": 3420
    },
    {
      "epoch": 0.5488,
      "grad_norm": 0.8781312108039856,
      "learning_rate": 3.184584178498986e-05,
      "loss": 1.4176,
      "step": 3430
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.9100823998451233,
      "learning_rate": 3.1643002028397566e-05,
      "loss": 1.3533,
      "step": 3440
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.832099437713623,
      "learning_rate": 3.144016227180527e-05,
      "loss": 1.4135,
      "step": 3450
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.7773131728172302,
      "learning_rate": 3.123732251521299e-05,
      "loss": 1.3946,
      "step": 3460
    },
    {
      "epoch": 0.5552,
      "grad_norm": 0.9809824228286743,
      "learning_rate": 3.103448275862069e-05,
      "loss": 1.2571,
      "step": 3470
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.8351847529411316,
      "learning_rate": 3.0831643002028396e-05,
      "loss": 1.5343,
      "step": 3480
    },
    {
      "epoch": 0.5584,
      "grad_norm": 0.8704239130020142,
      "learning_rate": 3.0628803245436106e-05,
      "loss": 1.4628,
      "step": 3490
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.747197151184082,
      "learning_rate": 3.0425963488843817e-05,
      "loss": 1.341,
      "step": 3500
    },
    {
      "epoch": 0.5616,
      "grad_norm": 0.8803126811981201,
      "learning_rate": 3.0223123732251524e-05,
      "loss": 1.3711,
      "step": 3510
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.9751145839691162,
      "learning_rate": 3.0020283975659232e-05,
      "loss": 1.2958,
      "step": 3520
    },
    {
      "epoch": 0.5648,
      "grad_norm": 0.7865713238716125,
      "learning_rate": 2.9817444219066936e-05,
      "loss": 1.3419,
      "step": 3530
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.8873676657676697,
      "learning_rate": 2.961460446247465e-05,
      "loss": 1.3827,
      "step": 3540
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.8832893371582031,
      "learning_rate": 2.9411764705882354e-05,
      "loss": 1.3512,
      "step": 3550
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.6538761854171753,
      "learning_rate": 2.920892494929006e-05,
      "loss": 1.3941,
      "step": 3560
    },
    {
      "epoch": 0.5712,
      "grad_norm": 1.031559705734253,
      "learning_rate": 2.900608519269777e-05,
      "loss": 1.3678,
      "step": 3570
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.9533334970474243,
      "learning_rate": 2.880324543610548e-05,
      "loss": 1.3249,
      "step": 3580
    },
    {
      "epoch": 0.5744,
      "grad_norm": 1.0496972799301147,
      "learning_rate": 2.8600405679513187e-05,
      "loss": 1.343,
      "step": 3590
    },
    {
      "epoch": 0.576,
      "grad_norm": 1.0601290464401245,
      "learning_rate": 2.8397565922920894e-05,
      "loss": 1.3382,
      "step": 3600
    },
    {
      "epoch": 0.5776,
      "grad_norm": 0.8638586401939392,
      "learning_rate": 2.81947261663286e-05,
      "loss": 1.3433,
      "step": 3610
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.7777538299560547,
      "learning_rate": 2.7991886409736312e-05,
      "loss": 1.3665,
      "step": 3620
    },
    {
      "epoch": 0.5808,
      "grad_norm": 0.6789288520812988,
      "learning_rate": 2.778904665314402e-05,
      "loss": 1.3623,
      "step": 3630
    },
    {
      "epoch": 0.5824,
      "grad_norm": 1.0239415168762207,
      "learning_rate": 2.7586206896551727e-05,
      "loss": 1.4301,
      "step": 3640
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.9509468078613281,
      "learning_rate": 2.738336713995943e-05,
      "loss": 1.4437,
      "step": 3650
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.9083086252212524,
      "learning_rate": 2.7180527383367145e-05,
      "loss": 1.3129,
      "step": 3660
    },
    {
      "epoch": 0.5872,
      "grad_norm": 0.7706483006477356,
      "learning_rate": 2.697768762677485e-05,
      "loss": 1.4301,
      "step": 3670
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.8252948522567749,
      "learning_rate": 2.6774847870182556e-05,
      "loss": 1.4042,
      "step": 3680
    },
    {
      "epoch": 0.5904,
      "grad_norm": 0.8689036965370178,
      "learning_rate": 2.6572008113590263e-05,
      "loss": 1.1999,
      "step": 3690
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.7418636679649353,
      "learning_rate": 2.6369168356997974e-05,
      "loss": 1.4241,
      "step": 3700
    },
    {
      "epoch": 0.5936,
      "grad_norm": 0.832951545715332,
      "learning_rate": 2.616632860040568e-05,
      "loss": 1.3442,
      "step": 3710
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.996060311794281,
      "learning_rate": 2.596348884381339e-05,
      "loss": 1.3034,
      "step": 3720
    },
    {
      "epoch": 0.5968,
      "grad_norm": 0.705885112285614,
      "learning_rate": 2.5760649087221096e-05,
      "loss": 1.4364,
      "step": 3730
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.7900152206420898,
      "learning_rate": 2.5557809330628807e-05,
      "loss": 1.3484,
      "step": 3740
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8224239349365234,
      "learning_rate": 2.5354969574036514e-05,
      "loss": 1.4556,
      "step": 3750
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.8629553914070129,
      "learning_rate": 2.515212981744422e-05,
      "loss": 1.3952,
      "step": 3760
    },
    {
      "epoch": 0.6032,
      "grad_norm": 0.9885224103927612,
      "learning_rate": 2.494929006085193e-05,
      "loss": 1.3182,
      "step": 3770
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.8529279232025146,
      "learning_rate": 2.4746450304259636e-05,
      "loss": 1.3986,
      "step": 3780
    },
    {
      "epoch": 0.6064,
      "grad_norm": 0.7550370693206787,
      "learning_rate": 2.4543610547667344e-05,
      "loss": 1.3796,
      "step": 3790
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.8512799739837646,
      "learning_rate": 2.434077079107505e-05,
      "loss": 1.3888,
      "step": 3800
    },
    {
      "epoch": 0.6096,
      "grad_norm": 0.706964910030365,
      "learning_rate": 2.413793103448276e-05,
      "loss": 1.3735,
      "step": 3810
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.7493021488189697,
      "learning_rate": 2.393509127789047e-05,
      "loss": 1.4348,
      "step": 3820
    },
    {
      "epoch": 0.6128,
      "grad_norm": 0.9412333369255066,
      "learning_rate": 2.3732251521298176e-05,
      "loss": 1.4326,
      "step": 3830
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.6976621747016907,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 1.2963,
      "step": 3840
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.8718717098236084,
      "learning_rate": 2.332657200811359e-05,
      "loss": 1.403,
      "step": 3850
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.7933664321899414,
      "learning_rate": 2.31237322515213e-05,
      "loss": 1.3301,
      "step": 3860
    },
    {
      "epoch": 0.6192,
      "grad_norm": 0.9271290302276611,
      "learning_rate": 2.292089249492901e-05,
      "loss": 1.252,
      "step": 3870
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.7640563249588013,
      "learning_rate": 2.2718052738336716e-05,
      "loss": 1.3344,
      "step": 3880
    },
    {
      "epoch": 0.6224,
      "grad_norm": 0.735242486000061,
      "learning_rate": 2.2515212981744424e-05,
      "loss": 1.2704,
      "step": 3890
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.8058167695999146,
      "learning_rate": 2.231237322515213e-05,
      "loss": 1.3281,
      "step": 3900
    },
    {
      "epoch": 0.6256,
      "grad_norm": 0.9089419841766357,
      "learning_rate": 2.210953346855984e-05,
      "loss": 1.3272,
      "step": 3910
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.8395382761955261,
      "learning_rate": 2.1906693711967546e-05,
      "loss": 1.4408,
      "step": 3920
    },
    {
      "epoch": 0.6288,
      "grad_norm": 0.9245477914810181,
      "learning_rate": 2.1703853955375257e-05,
      "loss": 1.362,
      "step": 3930
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.9677096009254456,
      "learning_rate": 2.150101419878296e-05,
      "loss": 1.3991,
      "step": 3940
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.9897236824035645,
      "learning_rate": 2.129817444219067e-05,
      "loss": 1.5116,
      "step": 3950
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.9850999712944031,
      "learning_rate": 2.109533468559838e-05,
      "loss": 1.3876,
      "step": 3960
    },
    {
      "epoch": 0.6352,
      "grad_norm": 0.7112181782722473,
      "learning_rate": 2.0892494929006086e-05,
      "loss": 1.4135,
      "step": 3970
    },
    {
      "epoch": 0.6368,
      "grad_norm": 0.6867976784706116,
      "learning_rate": 2.0689655172413793e-05,
      "loss": 1.3143,
      "step": 3980
    },
    {
      "epoch": 0.6384,
      "grad_norm": 1.0180323123931885,
      "learning_rate": 2.0486815415821504e-05,
      "loss": 1.3705,
      "step": 3990
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.1014171838760376,
      "learning_rate": 2.0283975659229208e-05,
      "loss": 1.3141,
      "step": 4000
    },
    {
      "epoch": 0.6416,
      "grad_norm": 0.9209902286529541,
      "learning_rate": 2.008113590263692e-05,
      "loss": 1.2576,
      "step": 4010
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.8500963449478149,
      "learning_rate": 1.9878296146044626e-05,
      "loss": 1.4116,
      "step": 4020
    },
    {
      "epoch": 0.6448,
      "grad_norm": 0.9721909165382385,
      "learning_rate": 1.9675456389452333e-05,
      "loss": 1.3289,
      "step": 4030
    },
    {
      "epoch": 0.6464,
      "grad_norm": 1.317718267440796,
      "learning_rate": 1.947261663286004e-05,
      "loss": 1.3778,
      "step": 4040
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.9101319313049316,
      "learning_rate": 1.926977687626775e-05,
      "loss": 1.426,
      "step": 4050
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.8741719126701355,
      "learning_rate": 1.9066937119675455e-05,
      "loss": 1.4806,
      "step": 4060
    },
    {
      "epoch": 0.6512,
      "grad_norm": 0.885848343372345,
      "learning_rate": 1.8864097363083166e-05,
      "loss": 1.3996,
      "step": 4070
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.7545533180236816,
      "learning_rate": 1.8661257606490873e-05,
      "loss": 1.4227,
      "step": 4080
    },
    {
      "epoch": 0.6544,
      "grad_norm": 0.9819992184638977,
      "learning_rate": 1.845841784989858e-05,
      "loss": 1.4271,
      "step": 4090
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.8249715566635132,
      "learning_rate": 1.8255578093306288e-05,
      "loss": 1.3331,
      "step": 4100
    },
    {
      "epoch": 0.6576,
      "grad_norm": 0.8639522790908813,
      "learning_rate": 1.8052738336714e-05,
      "loss": 1.3474,
      "step": 4110
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.7227104306221008,
      "learning_rate": 1.7849898580121703e-05,
      "loss": 1.3276,
      "step": 4120
    },
    {
      "epoch": 0.6608,
      "grad_norm": 0.9825816750526428,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 1.3308,
      "step": 4130
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.7120647430419922,
      "learning_rate": 1.744421906693712e-05,
      "loss": 1.3679,
      "step": 4140
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.6898847222328186,
      "learning_rate": 1.7241379310344828e-05,
      "loss": 1.4823,
      "step": 4150
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.9716871380805969,
      "learning_rate": 1.7038539553752536e-05,
      "loss": 1.3503,
      "step": 4160
    },
    {
      "epoch": 0.6672,
      "grad_norm": 0.8497072458267212,
      "learning_rate": 1.6835699797160246e-05,
      "loss": 1.3294,
      "step": 4170
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.9010389447212219,
      "learning_rate": 1.663286004056795e-05,
      "loss": 1.3643,
      "step": 4180
    },
    {
      "epoch": 0.6704,
      "grad_norm": 0.7756496071815491,
      "learning_rate": 1.643002028397566e-05,
      "loss": 1.3912,
      "step": 4190
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.7958990931510925,
      "learning_rate": 1.622718052738337e-05,
      "loss": 1.2344,
      "step": 4200
    },
    {
      "epoch": 0.6736,
      "grad_norm": 1.5075416564941406,
      "learning_rate": 1.6024340770791076e-05,
      "loss": 1.3564,
      "step": 4210
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.9622378945350647,
      "learning_rate": 1.5821501014198783e-05,
      "loss": 1.4722,
      "step": 4220
    },
    {
      "epoch": 0.6768,
      "grad_norm": 0.6588869690895081,
      "learning_rate": 1.5618661257606494e-05,
      "loss": 1.2721,
      "step": 4230
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.9116512537002563,
      "learning_rate": 1.5415821501014198e-05,
      "loss": 1.3334,
      "step": 4240
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9403899908065796,
      "learning_rate": 1.5212981744421909e-05,
      "loss": 1.4443,
      "step": 4250
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.9178708791732788,
      "learning_rate": 1.5010141987829616e-05,
      "loss": 1.3642,
      "step": 4260
    },
    {
      "epoch": 0.6832,
      "grad_norm": 0.9384081959724426,
      "learning_rate": 1.4807302231237325e-05,
      "loss": 1.3698,
      "step": 4270
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.7728821039199829,
      "learning_rate": 1.460446247464503e-05,
      "loss": 1.4612,
      "step": 4280
    },
    {
      "epoch": 0.6864,
      "grad_norm": 0.8059030175209045,
      "learning_rate": 1.440162271805274e-05,
      "loss": 1.41,
      "step": 4290
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.7898443937301636,
      "learning_rate": 1.4198782961460447e-05,
      "loss": 1.3698,
      "step": 4300
    },
    {
      "epoch": 0.6896,
      "grad_norm": 0.8634099960327148,
      "learning_rate": 1.3995943204868156e-05,
      "loss": 1.3239,
      "step": 4310
    },
    {
      "epoch": 0.6912,
      "grad_norm": 1.0027539730072021,
      "learning_rate": 1.3793103448275863e-05,
      "loss": 1.3668,
      "step": 4320
    },
    {
      "epoch": 0.6928,
      "grad_norm": 0.8978163003921509,
      "learning_rate": 1.3590263691683572e-05,
      "loss": 1.3609,
      "step": 4330
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.9273055791854858,
      "learning_rate": 1.3387423935091278e-05,
      "loss": 1.3224,
      "step": 4340
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.6951127052307129,
      "learning_rate": 1.3184584178498987e-05,
      "loss": 1.3958,
      "step": 4350
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.8663493990898132,
      "learning_rate": 1.2981744421906694e-05,
      "loss": 1.2714,
      "step": 4360
    },
    {
      "epoch": 0.6992,
      "grad_norm": 0.8381744027137756,
      "learning_rate": 1.2778904665314403e-05,
      "loss": 1.2578,
      "step": 4370
    },
    {
      "epoch": 0.7008,
      "grad_norm": 1.1157745122909546,
      "learning_rate": 1.257606490872211e-05,
      "loss": 1.284,
      "step": 4380
    },
    {
      "epoch": 0.7024,
      "grad_norm": 0.8827899098396301,
      "learning_rate": 1.2373225152129818e-05,
      "loss": 1.4336,
      "step": 4390
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.7563768625259399,
      "learning_rate": 1.2170385395537525e-05,
      "loss": 1.3222,
      "step": 4400
    },
    {
      "epoch": 0.7056,
      "grad_norm": 0.9869970083236694,
      "learning_rate": 1.1967545638945234e-05,
      "loss": 1.3535,
      "step": 4410
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.8248376846313477,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 1.3953,
      "step": 4420
    },
    {
      "epoch": 0.7088,
      "grad_norm": 0.8114254474639893,
      "learning_rate": 1.156186612576065e-05,
      "loss": 1.3259,
      "step": 4430
    },
    {
      "epoch": 0.7104,
      "grad_norm": 1.0196243524551392,
      "learning_rate": 1.1359026369168358e-05,
      "loss": 1.5373,
      "step": 4440
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.8844570517539978,
      "learning_rate": 1.1156186612576066e-05,
      "loss": 1.4636,
      "step": 4450
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.9430680871009827,
      "learning_rate": 1.0953346855983773e-05,
      "loss": 1.3663,
      "step": 4460
    },
    {
      "epoch": 0.7152,
      "grad_norm": 0.9466381072998047,
      "learning_rate": 1.075050709939148e-05,
      "loss": 1.3144,
      "step": 4470
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.8997886180877686,
      "learning_rate": 1.054766734279919e-05,
      "loss": 1.4141,
      "step": 4480
    },
    {
      "epoch": 0.7184,
      "grad_norm": 0.9555508494377136,
      "learning_rate": 1.0344827586206897e-05,
      "loss": 1.4292,
      "step": 4490
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8645280003547668,
      "learning_rate": 1.0141987829614604e-05,
      "loss": 1.4092,
      "step": 4500
    },
    {
      "epoch": 0.7216,
      "grad_norm": 0.8445830941200256,
      "learning_rate": 9.939148073022313e-06,
      "loss": 1.3546,
      "step": 4510
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.7512080669403076,
      "learning_rate": 9.73630831643002e-06,
      "loss": 1.4346,
      "step": 4520
    },
    {
      "epoch": 0.7248,
      "grad_norm": 0.8196505308151245,
      "learning_rate": 9.533468559837728e-06,
      "loss": 1.354,
      "step": 4530
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.8481506109237671,
      "learning_rate": 9.330628803245437e-06,
      "loss": 1.3504,
      "step": 4540
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.9044469594955444,
      "learning_rate": 9.127789046653144e-06,
      "loss": 1.3297,
      "step": 4550
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.8796373009681702,
      "learning_rate": 8.924949290060851e-06,
      "loss": 1.3237,
      "step": 4560
    },
    {
      "epoch": 0.7312,
      "grad_norm": 0.7506686449050903,
      "learning_rate": 8.72210953346856e-06,
      "loss": 1.3449,
      "step": 4570
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.9704193472862244,
      "learning_rate": 8.519269776876268e-06,
      "loss": 1.4063,
      "step": 4580
    },
    {
      "epoch": 0.7344,
      "grad_norm": 0.8338668346405029,
      "learning_rate": 8.316430020283975e-06,
      "loss": 1.3435,
      "step": 4590
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.9024910926818848,
      "learning_rate": 8.113590263691684e-06,
      "loss": 1.3664,
      "step": 4600
    },
    {
      "epoch": 0.7376,
      "grad_norm": 0.9343035221099854,
      "learning_rate": 7.910750507099392e-06,
      "loss": 1.3153,
      "step": 4610
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.76399827003479,
      "learning_rate": 7.707910750507099e-06,
      "loss": 1.2957,
      "step": 4620
    },
    {
      "epoch": 0.7408,
      "grad_norm": 0.8459974527359009,
      "learning_rate": 7.505070993914808e-06,
      "loss": 1.3934,
      "step": 4630
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.7022688984870911,
      "learning_rate": 7.302231237322515e-06,
      "loss": 1.3167,
      "step": 4640
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.6942615509033203,
      "learning_rate": 7.0993914807302235e-06,
      "loss": 1.4671,
      "step": 4650
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.8787364959716797,
      "learning_rate": 6.896551724137932e-06,
      "loss": 1.4055,
      "step": 4660
    },
    {
      "epoch": 0.7472,
      "grad_norm": 0.8846457600593567,
      "learning_rate": 6.693711967545639e-06,
      "loss": 1.3936,
      "step": 4670
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.8653333187103271,
      "learning_rate": 6.490872210953347e-06,
      "loss": 1.4805,
      "step": 4680
    },
    {
      "epoch": 0.7504,
      "grad_norm": 0.8352001905441284,
      "learning_rate": 6.288032454361055e-06,
      "loss": 1.3985,
      "step": 4690
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.8867529630661011,
      "learning_rate": 6.085192697768763e-06,
      "loss": 1.4465,
      "step": 4700
    },
    {
      "epoch": 0.7536,
      "grad_norm": 0.6923931241035461,
      "learning_rate": 5.882352941176471e-06,
      "loss": 1.333,
      "step": 4710
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.9098489880561829,
      "learning_rate": 5.679513184584179e-06,
      "loss": 1.3858,
      "step": 4720
    },
    {
      "epoch": 0.7568,
      "grad_norm": 0.899730920791626,
      "learning_rate": 5.4766734279918865e-06,
      "loss": 1.3489,
      "step": 4730
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.924389660358429,
      "learning_rate": 5.273833671399595e-06,
      "loss": 1.3702,
      "step": 4740
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9712046980857849,
      "learning_rate": 5.070993914807302e-06,
      "loss": 1.356,
      "step": 4750
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.9079488515853882,
      "learning_rate": 4.86815415821501e-06,
      "loss": 1.4066,
      "step": 4760
    },
    {
      "epoch": 0.7632,
      "grad_norm": 0.7201410531997681,
      "learning_rate": 4.665314401622718e-06,
      "loss": 1.3654,
      "step": 4770
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.8959711194038391,
      "learning_rate": 4.462474645030426e-06,
      "loss": 1.3248,
      "step": 4780
    },
    {
      "epoch": 0.7664,
      "grad_norm": 1.0216981172561646,
      "learning_rate": 4.259634888438134e-06,
      "loss": 1.4433,
      "step": 4790
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.8937973976135254,
      "learning_rate": 4.056795131845842e-06,
      "loss": 1.4326,
      "step": 4800
    },
    {
      "epoch": 0.7696,
      "grad_norm": 0.8216485381126404,
      "learning_rate": 3.8539553752535494e-06,
      "loss": 1.4409,
      "step": 4810
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.9484663605690002,
      "learning_rate": 3.6511156186612576e-06,
      "loss": 1.3283,
      "step": 4820
    },
    {
      "epoch": 0.7728,
      "grad_norm": 0.7381030917167664,
      "learning_rate": 3.448275862068966e-06,
      "loss": 1.4376,
      "step": 4830
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.798279881477356,
      "learning_rate": 3.2454361054766736e-06,
      "loss": 1.3134,
      "step": 4840
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.9430242776870728,
      "learning_rate": 3.0425963488843814e-06,
      "loss": 1.3852,
      "step": 4850
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.7687798142433167,
      "learning_rate": 2.8397565922920896e-06,
      "loss": 1.283,
      "step": 4860
    },
    {
      "epoch": 0.7792,
      "grad_norm": 0.8667353391647339,
      "learning_rate": 2.6369168356997973e-06,
      "loss": 1.3975,
      "step": 4870
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.7604361772537231,
      "learning_rate": 2.434077079107505e-06,
      "loss": 1.4119,
      "step": 4880
    },
    {
      "epoch": 0.7824,
      "grad_norm": 0.8409517407417297,
      "learning_rate": 2.231237322515213e-06,
      "loss": 1.4006,
      "step": 4890
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.7949575185775757,
      "learning_rate": 2.028397565922921e-06,
      "loss": 1.3408,
      "step": 4900
    },
    {
      "epoch": 0.7856,
      "grad_norm": 0.9525842070579529,
      "learning_rate": 1.8255578093306288e-06,
      "loss": 1.3606,
      "step": 4910
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.9600664973258972,
      "learning_rate": 1.6227180527383368e-06,
      "loss": 1.1382,
      "step": 4920
    },
    {
      "epoch": 0.7888,
      "grad_norm": 0.8406458497047424,
      "learning_rate": 1.4198782961460448e-06,
      "loss": 1.3669,
      "step": 4930
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.8937062621116638,
      "learning_rate": 1.2170385395537525e-06,
      "loss": 1.4203,
      "step": 4940
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.8134217262268066,
      "learning_rate": 1.0141987829614605e-06,
      "loss": 1.2862,
      "step": 4950
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.9232685565948486,
      "learning_rate": 8.113590263691684e-07,
      "loss": 1.3666,
      "step": 4960
    },
    {
      "epoch": 0.7952,
      "grad_norm": 0.7853943705558777,
      "learning_rate": 6.085192697768763e-07,
      "loss": 1.4673,
      "step": 4970
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.8831775784492493,
      "learning_rate": 4.056795131845842e-07,
      "loss": 1.3468,
      "step": 4980
    },
    {
      "epoch": 0.7984,
      "grad_norm": 1.4679898023605347,
      "learning_rate": 2.028397565922921e-07,
      "loss": 1.3509,
      "step": 4990
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8490538597106934,
      "learning_rate": 0.0,
      "loss": 1.3766,
      "step": 5000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.157765325651548e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}